diff --git a/feed_weekly.xml b/feed_weekly.xml index 27fac6d..77407ad 100644 --- a/feed_weekly.xml +++ b/feed_weekly.xml @@ -7,741 +7,741 @@ http://www.rssboard.org/rss-specification python-feedgen en - Mon, 13 Oct 2025 00:11:48 +0000 + Mon, 20 Oct 2025 00:12:04 +0000 - OrtSAE: Orthogonal Sparse Autoencoders Uncover Atomic Features - https://arxiv.org/abs/2509.22033 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.22033.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Anton Korznikov, Andrey Galichin, Alexey Dontsov, Oleg Rogov, Elena Tutubalina, Ivan Oseledets</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> Sparse autoencoders (SAEs) are a technique for sparse decomposition of neural network activations into human-interpretable features. However, current SAEs suffer from feature absorption, where specialized features capture instances of general features creating representation holes, and feature composition, where independent features merge into composite representations. In this work, we introduce Orthogonal SAE (OrtSAE), a novel approach aimed to mitigate these issues by enforcing orthogonality between the learned features. By implementing a new training procedure that penalizes high pairwise cosine similarity between SAE features, OrtSAE promotes the development of disentangled features while scaling linearly with the SAE size, avoiding significant computational overhead. We train OrtSAE across different models and layers and compare it with other methods. We find that OrtSAE discovers 9% more distinct features, reduces feature absorption (by 65%) and composition (by 15%), improves performance on spurious correlation removal (+6%), and achieves on-par performance for other downstream tasks compared to traditional SAEs.</p> - https://arxiv.org/abs/2509.22033 - Fri, 26 Sep 2025 08:10:52 +0000 + Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity + https://arxiv.org/abs/2510.01171 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01171.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiayi Zhang, Simon Yu, Derek Chong, Anthony Sicilia, Michael R. Tomz, Christopher D. Manning, Weiyan Shi</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Post-training alignment often reduces LLM diversity, leading to a phenomenon known as mode collapse. Unlike prior work that attributes this effect to algorithmic limitations, we identify a fundamental, pervasive data-level driver: typicality bias in preference data, whereby annotators systematically favor familiar text as a result of well-established findings in cognitive psychology. We formalize this bias theoretically, verify it on preference datasets empirically, and show that it plays a central role in mode collapse. Motivated by this analysis, we introduce Verbalized Sampling, a simple, training-free prompting strategy to circumvent mode collapse. VS prompts the model to verbalize a probability distribution over a set of responses (e.g., ``Generate 5 jokes about coffee and their corresponding probabilities''). Comprehensive experiments show that VS significantly improves performance across creative writing (poems, stories, jokes), dialogue simulation, open-ended QA, and synthetic data generation, without sacrificing factual accuracy and safety. For instance, in creative writing, VS increases diversity by 1.6-2.1x over direct prompting. We further observe an emergent trend that more capable models benefit more from VS. In sum, our work provides a new data-centric perspective on mode collapse and a practical inference-time remedy that helps unlock pre-trained generative diversity.</p> + https://arxiv.org/abs/2510.01171 + Wed, 01 Oct 2025 17:55:37 +0000 - Meta-Awareness Enhances Reasoning Models: Self-Alignment Reinforcement Learning - https://arxiv.org/abs/2510.03259 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03259.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yoonjeon Kim, Doohyuk Jang, Eunho Yang</p><p><b>Upvotes:</b> 44</p><p><b>Summary:</b> Recent studies on reasoning models explore the meta-awareness of language models, the ability to know how to think by itself. We argue that large reasoning models lack this meta-awareness property by proving severe misalignment between true rollouts and predicted meta information. We posit that aligning meta-prediction with true rollouts will lead to significant performance gains. To verify this hypothesis, we design a training pipeline that boosts Meta-Awareness via Self-Alignment (MASA), and prove that enhanced meta-awareness directly translates to improved accuracy. Unlike existing meta-cognitive reasoning models, our method does not require external training sources but leverages self-generated signals to train meta-awareness. Moreover, our method enables efficient training by i) filtering out zero-variance prompts that are either trivial or unsolvable and ii) cutting off lengthy rollouts when they are unlikely to lead to correct answers. The results are inspiring: our strategy yields significant improvements in both accuracy and training efficiency on in-domain tasks and shows strong generalization to out-of-domain benchmarks. More specifically, our method can speed up GRPO training by over 1.28x to reach the same performance, and achieve a 19.3% gain in accuracy on AIME25, and a 6.2 % average gain over six mathematics benchmarks. Training with meta-cognitive guidance enhances out-of-domain generalization, giving a 3.87 % boost on GPQA-Diamond and a 2.08 % overall accuracy gain across 13 benchmarks spanning logical, scientific, and coding domains.</p> - https://arxiv.org/abs/2510.03259 - Fri, 26 Sep 2025 14:05:48 +0000 + TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling + https://arxiv.org/abs/2510.04533 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04533.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hyunmin Cho, Donghoon Ahn, Susung Hong, Jee Eun Kim, Seungryong Kim, Kyong Hwan Jin</p><p><b>Upvotes:</b> 46</p><p><b>Summary:</b> Recent diffusion models achieve the state-of-the-art performance in image generation, but often suffer from semantic inconsistencies or hallucinations. While various inference-time guidance methods can enhance generation, they often operate indirectly by relying on external signals or architectural modifications, which introduces additional computational overhead. In this paper, we propose Tangential Amplifying Guidance (TAG), a more efficient and direct guidance method that operates solely on trajectory signals without modifying the underlying diffusion model. TAG leverages an intermediate sample as a projection basis and amplifies the tangential components of the estimated scores with respect to this basis to correct the sampling trajectory. We formalize this guidance process by leveraging a first-order Taylor expansion, which demonstrates that amplifying the tangential component steers the state toward higher-probability regions, thereby reducing inconsistencies and enhancing sample quality. TAG is a plug-and-play, architecture-agnostic module that improves diffusion sampling fidelity with minimal computational addition, offering a new perspective on diffusion guidance.</p> + https://arxiv.org/abs/2510.04533 + Mon, 06 Oct 2025 06:53:29 +0000 - Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data - https://arxiv.org/abs/2510.03264 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03264.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Syeda Nahida Akter, Shrimai Prabhumoye, Eric Nyberg, Mostofa Patwary, Mohammad Shoeybi, Yejin Choi, Bryan Catanzaro</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> The prevailing paradigm for enhancing the reasoning abilities of LLMs revolves around post-training on high-quality, reasoning-intensive data. While emerging literature suggests that reasoning data is increasingly incorporated also during the mid-training stage-a practice that is relatively more proprietary and less openly characterized-the role of such data in pretraining remains unclear. In particular, due to the opaqueness of pretraining corpora in most frontier models, the effect of reasoning data introduced at different phases of pre- and/or post-training is relatively less reported in the scientific literature. This raises several important questions: Is adding reasoning data earlier during pretraining any better than introducing it during post-training? Could earlier inclusion risk overfitting and harm generalization, or instead establish durable foundations that later fine-tuning cannot recover? We conduct the first systematic study of how reasoning data-varying in scale, diversity, and quality-affects LLM performance when introduced at different stages of training. We find that front-loading reasoning data into pretraining is critical (19% avg gain), establishing foundational capabilities that cannot be fully replicated by later-stage SFT, even with more data. We uncover an asymmetric principle for optimal data allocation: pretraining benefits most from broad diversity in reasoning patterns (11% avg gain), while SFT is more sensitive to data quality (15% avg gain). We show that high-quality pretraining data has latent effects, activated only after SFT, and that naively scaling SFT data can be detrimental, washing away the benefits of early reasoning injection. Our results challenge the conventional separation of language modeling and reasoning, providing a principled guide for strategically allocating data across the entire training pipeline to build more capable models.</p> - https://arxiv.org/abs/2510.03264 - Fri, 26 Sep 2025 20:08:51 +0000 + Making Mathematical Reasoning Adaptive + https://arxiv.org/abs/2510.04617 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04617.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhejian Lai, Xiang Geng, Zhijun Wang, Yang Bai, Jiahuan Li, Rongxiang Weng, Jingang Wang, Xuezhi Cao, Xunliang Cai, Shujian Huang</p><p><b>Upvotes:</b> 22</p><p><b>Summary:</b> Mathematical reasoning is a primary indicator of large language models (LLMs) intelligence. However, existing LLMs exhibit failures of robustness and generalization. This paper attributes these deficiencies to spurious reasoning, i.e., producing answers from superficial features. To address this challenge, we propose the AdaR framework to enable adaptive reasoning, wherein models rely on problem-solving logic to produce answers. AdaR synthesizes logically equivalent queries by varying variable values, and trains models with RLVR on these data to penalize spurious logic while encouraging adaptive logic. To improve data quality, we extract the problem-solving logic from the original query and generate the corresponding answer by code execution, then apply a sanity check. Experimental results demonstrate that AdaR improves robustness and generalization, achieving substantial improvement in mathematical reasoning while maintaining high data efficiency. Analysis indicates that data synthesis and RLVR function in a coordinated manner to enable adaptive reasoning in LLMs. Subsequent analyses derive key design insights into the effect of critical factors and the applicability to instruct LLMs. Our project is available at https://github.com/LaiZhejian/AdaR</p> + https://arxiv.org/abs/2510.04617 + Mon, 06 Oct 2025 09:30:05 +0000 - CoDA: Coding LM via Diffusion Adaptation - https://arxiv.org/abs/2510.03270 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03270.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haolin Chen, Shiyu Wang, Can Qin, Bo Pang, Zuxin Liu, Jielin Qiu, Jianguo Zhang, Yingbo Zhou, Zeyuan Chen, Ran Xu, Shelby Heinecke, Silvio Savarese, Caiming Xiong, Huan Wang, Weiran Yao</p><p><b>Upvotes:</b> 39</p><p><b>Summary:</b> Diffusion language models promise bidirectional context and infilling capabilities that autoregressive coders lack, yet practical systems remain heavyweight. We introduce CoDA, a 1.7B-parameter diffusion coder trained on TPU with a fully open-source training pipeline. CoDA pairs large-scale diffusion pre-training with code-centric mid-training and instruction tuning, enabling confidence-guided sampling that keeps inference latency competitive. On Humaneval, MBPP, and EvalPlus, CoDA-1.7B-Instruct matches or surpasses diffusion models up to 7B parameters. Our release includes model checkpoints, evaluation harnesses, and TPU training pipelines to accelerate research on lightweight diffusion-based coding assistants.</p> - https://arxiv.org/abs/2510.03270 - Sat, 27 Sep 2025 05:41:55 +0000 + ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMs + https://arxiv.org/abs/2510.04767 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04767.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wonjun Kang, Kevin Galim, Seunghyuk Oh, Minjae Lee, Yuchen Zeng, Shuibai Zhang, Coleman Hooper, Yuezhou Hu, Hyung Il Koo, Nam Ik Cho, Kangwook Lee</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> While most autoregressive LLMs are constrained to one-by-one decoding, diffusion LLMs (dLLMs) have attracted growing interest for their potential to dramatically accelerate inference through parallel decoding. Despite this promise, the conditional independence assumption in dLLMs causes parallel decoding to ignore token dependencies, inevitably degrading generation quality when these dependencies are strong. However, existing works largely overlook these inherent challenges, and evaluations on standard benchmarks (e.g., math and coding) are not sufficient to capture the quality degradation caused by parallel decoding. To address this gap, we first provide an information-theoretic analysis of parallel decoding. We then conduct case studies on analytically tractable synthetic list operations from both data distribution and decoding strategy perspectives, offering quantitative insights that highlight the fundamental limitations of parallel decoding. Building on these insights, we propose ParallelBench, the first benchmark specifically designed for dLLMs, featuring realistic tasks that are trivial for humans and autoregressive LLMs yet exceptionally challenging for dLLMs under parallel decoding. Using ParallelBench, we systematically analyze both dLLMs and autoregressive LLMs, revealing that: (i) dLLMs under parallel decoding can suffer dramatic quality degradation in real-world scenarios, and (ii) current parallel decoding strategies struggle to adapt their degree of parallelism based on task difficulty, thus failing to achieve meaningful speedup without compromising quality. Our findings underscore the pressing need for innovative decoding methods that can overcome the current speed-quality trade-off. We release our benchmark to help accelerate the development of truly efficient dLLMs.</p> + https://arxiv.org/abs/2510.04767 + Mon, 06 Oct 2025 12:41:31 +0000 - Bridging the Gap Between Promise and Performance for Microscaling FP4 Quantization - https://arxiv.org/abs/2509.23202 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.23202.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Vage Egiazarian, Roberto L. Castro, Denis Kuznedelev, Andrei Panferov, Eldar Kurtic, Shubhra Pandit, Alexandre Marques, Mark Kurtz, Saleh Ashkboos, Torsten Hoefler, Dan Alistarh</p><p><b>Upvotes:</b> 25</p><p><b>Summary:</b> The recent hardware-accelerated microscaling 4-bit floating-point formats such as MXFP4 and NVFP4, supported on NVIDIA and AMD GPUs, promise to revolutionize large language model (LLM) inference. Yet, their practical benefits remain unproven. We present the first comprehensive study of MXFP4 and NVFP4 for post-training quantization, revealing gaps between their promise and real-world performance. Our analysis shows that state-of-the-art methods struggle with FP4, due to two key issues: (1) NVFP4's small group size provably neutralizes traditional outlier mitigation techniques; (2) MXFP4's power-of-two scale quantization severely degrades accuracy due to high induced error. To bridge this gap, we introduce Micro-Rotated-GPTQ (MR-GPTQ), a variant of the classic GPTQ quantization algorithm that tailors the quantization process to FP4's unique properties, by using block-wise Hadamard transforms and format-specific optimizations. We support our proposal with a set of high-performance GPU kernels that enable the MR-GPTQ format with negligible overhead, by rotation fusion into the weights, and fast online computation of the activations. This leads to speedups vs. FP16 of up to 3.6x layer-wise, and 2.2x end-to-end on NVIDIA B200, and of 6x layer-wise and 4x end-to-end on RTX5090. Our extensive empirical evaluation demonstrates that MR-GPTQ matches or outperforms state-of-the-art accuracy, significantly boosting MXFP4, to the point where it nears that of NVFP4. We conclude that, while FP4 is not an automatic upgrade over INT4, format-specialized methods like MR-GPTQ can unlock a new frontier of accuracy-performance trade-offs.</p> - https://arxiv.org/abs/2509.23202 - Sat, 27 Sep 2025 09:22:21 +0000 + When Models Lie, We Learn: Multilingual Span-Level Hallucination Detection with PsiloQA + https://arxiv.org/abs/2510.04849 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04849.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Elisei Rykov, Kseniia Petrushina, Maksim Savkin, Valerii Olisov, Artem Vazhentsev, Kseniia Titova, Alexander Panchenko, Vasily Konovalov, Julia Belikova</p><p><b>Upvotes:</b> 100</p><p><b>Summary:</b> Hallucination detection remains a fundamental challenge for the safe and reliable deployment of large language models (LLMs), especially in applications requiring factual accuracy. Existing hallucination benchmarks often operate at the sequence level and are limited to English, lacking the fine-grained, multilingual supervision needed for a comprehensive evaluation. In this work, we introduce PsiloQA, a large-scale, multilingual dataset annotated with span-level hallucinations across 14 languages. PsiloQA is constructed through an automated three-stage pipeline: generating question-answer pairs from Wikipedia using GPT-4o, eliciting potentially hallucinated answers from diverse LLMs in a no-context setting, and automatically annotating hallucinated spans using GPT-4o by comparing against golden answers and retrieved context. We evaluate a wide range of hallucination detection methods -- including uncertainty quantification, LLM-based tagging, and fine-tuned encoder models -- and show that encoder-based models achieve the strongest performance across languages. Furthermore, PsiloQA demonstrates effective cross-lingual generalization and supports robust knowledge transfer to other benchmarks, all while being significantly more cost-efficient than human-annotated datasets. Our dataset and results advance the development of scalable, fine-grained hallucination detection in multilingual settings.</p> + https://arxiv.org/abs/2510.04849 + Mon, 06 Oct 2025 14:36:30 +0000 - CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding - https://arxiv.org/abs/2509.23379 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.23379.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xi Zhang, Zaiqiao Meng, Jake Lever, Edmond S. L. Ho</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Multimodal large language models (MLLMs) have recently achieved remarkable progress in radiology by integrating visual perception with natural language understanding. However, they often generate clinically unsupported descriptions, known as medical hallucinations, which pose serious risks in medical applications that demand accuracy and image-grounded outputs. Through empirical analysis, we find that prompt-induced hallucinations remain prevalent in radiology MLLMs, largely due to over-sensitivity to clinical sections. To address this, we introduce Clinical Contrastive Cecoding (CCD), a training-free and retrieval-free inference framework that integrates structured clinical signals from task-specific radiology expert models. CCD introduces a dual-stage contrastive mechanism to refine token-level logits during generation, thereby enhancing clinical fidelity without modifying the base MLLM. Experiments on three datasets and multiple models demonstrate that CCD consistently improves overall performance on radiology report generation (RRG). On the MIMIC-CXR dataset, it yields up to a 17% improvement in RadGraph-F1 when applied to state-of-the-art RRG models. Our approach provides a lightweight and generalisable solution for mitigating medical hallucinations, effectively bridging expert models and MLLMs in radiology.</p> - https://arxiv.org/abs/2509.23379 - Sat, 27 Sep 2025 16:01:09 +0000 + D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI + https://arxiv.org/abs/2510.05684 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05684.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Suwhan Choi, Jaeyoon Jung, Haebin Seong, Minchan Kim, Minyeong Kim, Yongjun Cho, Yoonshik Kim, Yubeen Park, Youngjae Yu, Yunsung Lee</p><p><b>Upvotes:</b> 127</p><p><b>Summary:</b> Large language models leverage internet-scale text data, yet embodied AI remains constrained by the prohibitive costs of physical trajectory collection. Desktop environments -- particularly gaming -- offer a compelling alternative: they provide rich sensorimotor interactions at scale while maintaining the structured observation-action coupling essential for embodied learning. We present D2E (Desktop to Embodied AI), a framework that demonstrates desktop interactions can serve as an effective pretraining substrate for robotics embodied AI tasks. Unlike prior work that remained domain-specific (e.g., VPT for Minecraft) or kept data proprietary (e.g., SIMA), D2E establishes a complete pipeline from scalable desktop data collection to verified transfer in embodied domains. Our framework comprises three components: (1) the OWA Toolkit that unifies diverse desktop interactions into a standardized format with 152x compression, (2) the Generalist-IDM that achieves strong zero-shot generalization across unseen games through timestamp-based event prediction, enabling internet-scale pseudo-labeling, and (3) VAPT that transfers desktop-pretrained representations to physical manipulation and navigation. Using 1.3K+ hours of data (259 hours of human demonstrations, and 1K+ hours of pseudo-labeled gameplay), we achieve a total of 96.6% success rate on LIBERO manipulation and 83.3% on CANVAS navigation benchmarks. This validates that sensorimotor primitives in digital interactions exhibit sufficient invariance to transfer meaningfully to physical embodied tasks, establishing desktop pretraining as a practical paradigm for robotics. We will make all our work public, including the OWA toolkit, datasets of human-collected and pseudo-labeled, and VAPT-trained models available at https://worv-ai.github.io/d2e/</p> + https://arxiv.org/abs/2510.05684 + Tue, 07 Oct 2025 08:40:33 +0000 - From What to Why: A Multi-Agent System for Evidence-based Chemical Reaction Condition Reasoning - https://arxiv.org/abs/2509.23768 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.23768.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Cheng Yang, Jiaxuan Lu, Haiyuan Wan, Junchi Yu, Feiwei Qin</p><p><b>Upvotes:</b> 46</p><p><b>Summary:</b> The chemical reaction recommendation is to select proper reaction condition parameters for chemical reactions, which is pivotal to accelerating chemical science. With the rapid development of large language models (LLMs), there is growing interest in leveraging their reasoning and planning capabilities for reaction condition recommendation. Despite their success, existing methods rarely explain the rationale behind the recommended reaction conditions, limiting their utility in high-stakes scientific workflows. In this work, we propose ChemMAS, a multi-agent system that reframes condition prediction as an evidence-based reasoning task. ChemMAS decomposes the task into mechanistic grounding, multi-channel recall, constraint-aware agentic debate, and rationale aggregation. Each decision is backed by interpretable justifications grounded in chemical knowledge and retrieved precedents. Experiments show that ChemMAS achieves 20-35% gains over domain-specific baselines and outperforms general-purpose LLMs by 10-15% in Top-1 accuracy, while offering falsifiable, human-trustable rationales, which establishes a new paradigm for explainable AI in scientific discovery.</p> - https://arxiv.org/abs/2509.23768 - Sun, 28 Sep 2025 09:34:35 +0000 + Webscale-RL: Automated Data Pipeline for Scaling RL Data to Pretraining Levels + https://arxiv.org/abs/2510.06499 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06499.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhepeng Cen, Haolin Chen, Shiyu Wang, Zuxin Liu, Zhiwei Liu, Ding Zhao, Silvio Savarese, Caiming Xiong, Huan Wang, Weiran Yao</p><p><b>Upvotes:</b> 31</p><p><b>Summary:</b> Large Language Models (LLMs) have achieved remarkable success through imitation learning on vast text corpora, but this paradigm creates a training-generation gap and limits robust reasoning. Reinforcement learning (RL) offers a more data-efficient solution capable of bridging this gap, yet its application has been constrained by a critical data bottleneck: existing RL datasets are orders of magnitude smaller and less diverse than web-scale pre-training corpora. To address this, we introduce the Webscale-RL pipeline, a scalable data engine that systematically converts large-scale pre-training documents into millions of diverse, verifiable question-answer pairs for RL. Using this pipeline, we construct the Webscale-RL dataset, containing 1.2 million examples across more than 9 domains. Our experiments show that the model trained on this dataset significantly outperforms continual pretraining and strong data refinement baselines across a suite of benchmarks. Notably, RL training with our dataset proves substantially more efficient, achieving the performance of continual pre-training with up to 100times fewer tokens. Our work presents a viable path toward scaling RL to pre-training levels, enabling more capable and efficient language models.</p> + https://arxiv.org/abs/2510.06499 + Tue, 07 Oct 2025 22:30:59 +0000 - MemMamba: Rethinking Memory Patterns in State Space Model - https://arxiv.org/abs/2510.03279 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03279.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Youjin Wang, Yangjingyi Chen, Jiahao Yan, Jiaxuan Lu, Xiao Sun</p><p><b>Upvotes:</b> 63</p><p><b>Summary:</b> With the explosive growth of data, long-sequence modeling has become increasingly important in tasks such as natural language processing and bioinformatics. However, existing methods face inherent trade-offs between efficiency and memory. Recurrent neural networks suffer from gradient vanishing and explosion, making them hard to scale. Transformers can model global dependencies but are constrained by quadratic complexity. Recently, selective state-space models such as Mamba have demonstrated high efficiency with O(n) time and O(1) recurrent inference, yet their long-range memory decays exponentially. In this work, we conduct mathematical derivations and information-theoretic analysis to systematically uncover the memory decay mechanism of Mamba, answering a fundamental question: what is the nature of Mamba's long-range memory and how does it retain information? To quantify key information loss, we further introduce horizontal-vertical memory fidelity metrics that capture degradation both within and across layers. Inspired by how humans distill and retain salient information when reading long documents, we propose MemMamba, a novel architectural framework that integrates state summarization mechanism together with cross-layer and cross-token attention, which alleviates long-range forgetting while preserving linear complexity. MemMamba achieves significant improvements over existing Mamba variants and Transformers on long-sequence benchmarks such as PG19 and Passkey Retrieval, while delivering a 48% speedup in inference efficiency. Both theoretical analysis and empirical results demonstrate that MemMamba achieves a breakthrough in the complexity-memory trade-off, offering a new paradigm for ultra-long sequence modeling.</p> - https://arxiv.org/abs/2510.03279 - Sun, 28 Sep 2025 14:40:58 +0000 + CVD-STORM: Cross-View Video Diffusion with Spatial-Temporal Reconstruction Model for Autonomous Driving + https://arxiv.org/abs/2510.07944 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07944.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Tianrui Zhang, Yichen Liu, Zilin Guo, Yuxin Guo, Jingcheng Ni, Chenjing Ding, Dan Xu, Lewei Lu, Zehuan Wu</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Generative models have been widely applied to world modeling for environment simulation and future state prediction. With advancements in autonomous driving, there is a growing demand not only for high-fidelity video generation under various controls, but also for producing diverse and meaningful information such as depth estimation. To address this, we propose CVD-STORM, a cross-view video diffusion model utilizing a spatial-temporal reconstruction Variational Autoencoder (VAE) that generates long-term, multi-view videos with 4D reconstruction capabilities under various control inputs. Our approach first fine-tunes the VAE with an auxiliary 4D reconstruction task, enhancing its ability to encode 3D structures and temporal dynamics. Subsequently, we integrate this VAE into the video diffusion process to significantly improve generation quality. Experimental results demonstrate that our model achieves substantial improvements in both FID and FVD metrics. Additionally, the jointly-trained Gaussian Splatting Decoder effectively reconstructs dynamic scenes, providing valuable geometric information for comprehensive scene understanding.</p> + https://arxiv.org/abs/2510.07944 + Thu, 09 Oct 2025 08:41:58 +0000 - Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval and Synthesis for SLMs - https://arxiv.org/abs/2509.24107 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.24107.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shreyas Singh, Kunal Singh, Pradeep Moturi</p><p><b>Upvotes:</b> 66</p><p><b>Summary:</b> Tool-integrated reasoning has emerged as a key focus for enabling agentic applications. Among these, DeepResearch Agents have gained significant attention for their strong performance on complex, open-ended information-seeking tasks. We introduce Fathom-DeepResearch, an agentic system composed of two specialized models. The first is Fathom-Search-4B, a DeepSearch model trained from Qwen3-4B and optimized for evidence-based investigation through live web search and targeted webpage querying. Its training combines three advances: (i) DUETQA, a 5K-sample dataset generated via multi-agent self-play that enforces strict web-search dependence and heterogeneous source grounding; (ii) RAPO, a zero-overhead extension of GRPO that stabilizes multi-turn Reinforcement Learning with Verifiable Rewards through curriculum pruning, reward-aware advantage scaling, and per-prompt replay buffers; and (iii) a steerable step-level reward that classifies each tool call by cognitive behavior and marginal utility, enabling explicit control over search trajectory breadth, depth, and horizon. These improvements enable reliable extension of tool-calling beyond 20 calls when warranted. The second is Fathom-Synthesizer-4B, trained from Qwen3-4B, which converts multi-turn DeepSearch traces into structured, citation-dense DeepResearch Reports for comprehensive synthesis. Evaluated on DeepSearch benchmarks (SimpleQA, FRAMES, WebWalker, Seal0, MuSiQue) and DeepResearch-Bench, the system achieves state-of-the-art performance in the open-weights category while demonstrating strong generalization to diverse reasoning tasks including HLE, AIME-25, GPQA-Diamond, and MedQA.</p> - https://arxiv.org/abs/2509.24107 - Sun, 28 Sep 2025 22:58:11 +0000 + DISCO: Diversifying Sample Condensation for Efficient Model Evaluation + https://arxiv.org/abs/2510.07959 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07959.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Alexander Rubinstein, Benjamin Raible, Martin Gubri, Seong Joon Oh</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Evaluating modern machine learning models has become prohibitively expensive. Benchmarks such as LMMs-Eval and HELM demand thousands of GPU hours per model. Costly evaluation reduces inclusivity, slows the cycle of innovation, and worsens environmental impact. The typical approach follows two steps. First, select an anchor subset of data. Second, train a mapping from the accuracy on this subset to the final test result. The drawback is that anchor selection depends on clustering, which can be complex and sensitive to design choices. We argue that promoting diversity among samples is not essential; what matters is to select samples that maximise diversity in model responses. Our method, Diversifying Sample Condensation (DISCO), selects the top-k samples with the greatest model disagreements. This uses greedy, sample-wise statistics rather than global clustering. The approach is conceptually simpler. From a theoretical view, inter-model disagreement provides an information-theoretically optimal rule for such greedy selection. DISCO shows empirical gains over prior methods, achieving state-of-the-art results in performance prediction across MMLU, Hellaswag, Winogrande, and ARC. Code is available here: https://github.com/arubique/disco-public.</p> + https://arxiv.org/abs/2510.07959 + Thu, 09 Oct 2025 08:53:59 +0000 - Reinforcement Mid-Training - https://arxiv.org/abs/2509.24375 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.24375.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yijun Tian, Shaoyu Chen, Zhichao Xu, Yawei Wang, Jinhe Bi, Peng Han, Wei Wang</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> The development of state-of-the-art large language models is commonly understood as a two-stage process involving pre-training and post-training. We point out the need for an additional intermediate stage called reinforcement mid-training with potential for strong performance gains. In this paper, we formally define the problem and identify three key challenges: (1) inefficient training due to excessive reasoning steps, (2) disregard of the imbalanced token entropy distribution, and (3) underutilization of token information. To address these challenges, we propose RMT, a framework for efficient, adaptive, and unified reinforcement mid-training with various innovative components. In particular, we first introduce a dynamic token budget mechanism that constrains unnecessary reasoning steps and mitigates model overthinking. Next, we design a curriculum-based adaptive sampling method that fosters a progressive learning trajectory from easy to hard tokens. Finally, we present a dual training strategy that combines reinforcement learning with next-token prediction, ensuring targeted learning on key tokens and full exploitation of all token information. Extensive experiments demonstrate the superiority of RMT over state-of-the-art methods, achieving up to +64.91% performance improvement with only 21% of the reasoning length in language modeling. We also show that checkpoints obtained after reinforcement mid-training can benefit the subsequent post-training, yielding up to +18.76% improvement in the mathematical domain.</p> - https://arxiv.org/abs/2509.24375 - Mon, 29 Sep 2025 07:21:24 +0000 + R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth? + https://arxiv.org/abs/2510.08189 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08189.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yi Lu, Jianing Wang, Linsen Guo, Wei He, Hongyin Tang, Tao Gui, Xuanjing Huang, Xuezhi Cao, Wei Wang, Xunliang Cai</p><p><b>Upvotes:</b> 24</p><p><b>Summary:</b> Recent trends in test-time scaling for reasoning models (e.g., OpenAI o1, DeepSeek-R1) have led to remarkable improvements through long Chain-of-Thought (CoT). However, existing benchmarks mainly focus on immediate, single-horizon tasks, failing to adequately evaluate models' ability to understand and respond to complex, long-horizon scenarios. To address this incomplete evaluation of Large Reasoning Models (LRMs), we propose R-HORIZON, a method designed to stimulate long-horizon reasoning behaviors in LRMs through query composition. Based on R-HORIZON, we construct a long-horizon reasoning benchmark, comprising complex multi-step reasoning tasks with interdependent problems that span long reasoning horizons. Through comprehensive evaluation of LRMs using the R-HORIZON benchmark, we find that even the most advanced LRMs suffer significant performance degradation. Our analysis reveals that LRMs exhibit limited effective reasoning length and struggle to allocate thinking budget across multiple problems appropriately. Recognizing these limitations, we use R-HORIZON to construct long-horizon reasoning data for reinforcement learning with verified rewards (RLVR). Compared to training with single-horizon data, RLVR with R-HORIZON not only substantially improves performance on the multi-horizon reasoning tasks, but also promotes accuracy on standard reasoning tasks, with an increase of 7.5 on AIME2024. These results position R-HORIZON as a scalable, controllable, and low-cost paradigm for enhancing and evaluating the long-horizon reasoning capabilities of LRMs.</p> + https://arxiv.org/abs/2510.08189 + Thu, 09 Oct 2025 13:16:22 +0000 - Free Lunch Alignment of Text-to-Image Diffusion Models without Preference Image Pairs - https://arxiv.org/abs/2509.25771 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.25771.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jia Jun Cheng Xian, Muchen Li, Haotian Yang, Xin Tao, Pengfei Wan, Leonid Sigal, Renjie Liao</p><p><b>Upvotes:</b> 10</p><p><b>Summary:</b> Recent advances in diffusion-based text-to-image (T2I) models have led to remarkable success in generating high-quality images from textual prompts. However, ensuring accurate alignment between the text and the generated image remains a significant challenge for state-of-the-art diffusion models. To address this, existing studies employ reinforcement learning with human feedback (RLHF) to align T2I outputs with human preferences. These methods, however, either rely directly on paired image preference data or require a learned reward function, both of which depend heavily on costly, high-quality human annotations and thus face scalability limitations. In this work, we introduce Text Preference Optimization (TPO), a framework that enables "free-lunch" alignment of T2I models, achieving alignment without the need for paired image preference data. TPO works by training the model to prefer matched prompts over mismatched prompts, which are constructed by perturbing original captions using a large language model. Our framework is general and compatible with existing preference-based algorithms. We extend both DPO and KTO to our setting, resulting in TDPO and TKTO. Quantitative and qualitative evaluations across multiple benchmarks show that our methods consistently outperform their original counterparts, delivering better human preference scores and improved text-to-image alignment. Our Open-source code is available at https://github.com/DSL-Lab/T2I-Free-Lunch-Alignment.</p> - https://arxiv.org/abs/2509.25771 - Tue, 30 Sep 2025 04:32:34 +0000 + ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shaping + https://arxiv.org/abs/2510.08457 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08457.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shuang Chen, Yue Guo, Yimeng Ye, Shijue Huang, Wenbo Hu, Haoxi Li, Manyuan Zhang, Jiayu Chen, Song Guo, Nanyun Peng</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> Recent advances in multimodal large reasoning models (MLRMs) have substantially improved their ability to solve complex textual and visual tasks. However, these models tend to overthink on simple problems, producing unnecessarily lengthy reasoning traces, while under-exploring on challenging ones, leading to missed solutions. To address this imbalance, we propose ARES, a unified open-source framework for adaptive reasoning that dynamically allocates exploration effort based on task difficulty. Our approach is motivated by two key empirical findings: (i) while single-token entropy is noisy, high window-entropy (HWE) tokens (token-level entropies averaged under a sliding window) can reliably capture reasoning-critical moments; and (ii) reducing HWE usage benefits easy problems, while increasing it is essential for solving hard ones. Building on these insights, ARES introduces a two-stage training pipeline. In the Adaptive Cold-Start stage, we curate multimodal and textual data paired with reasoning traces of length proportional to problem difficulty, equipping the model with initial difficulty awareness. In the second stage, we develop Adaptive Entropy Policy Optimization (AEPO), which uses HWE tokens as exploration triggers to decide when to explore, and a hierarchical entropy reward with dynamic KL control to decide how much to explore. Extensive experiments demonstrate that ARES achieves superior performance and reasoning efficiency across diverse mathematical, logical, and multimodal benchmarks, while closing the gap to leading commercial systems under significantly lower inference costs.</p> + https://arxiv.org/abs/2510.08457 + Thu, 09 Oct 2025 17:03:28 +0000 - Fast-dLLM v2: Efficient Block-Diffusion LLM - https://arxiv.org/abs/2509.26328 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.26328.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chengyue Wu, Hao Zhang, Shuchen Xue, Shizhe Diao, Yonggan Fu, Zhijian Liu, Pavlo Molchanov, Ping Luo, Song Han, Enze Xie</p><p><b>Upvotes:</b> 45</p><p><b>Summary:</b> Autoregressive (AR) large language models (LLMs) have achieved remarkable performance across a wide range of natural language tasks, yet their inherent sequential decoding limits inference efficiency. In this work, we propose Fast-dLLM v2, a carefully designed block diffusion language model (dLLM) that efficiently adapts pretrained AR models into dLLMs for parallel text generation, requiring only approximately 1B tokens of fine-tuning. This represents a 500x reduction in training data compared to full-attention diffusion LLMs such as Dream (580B tokens), while preserving the original model's performance. Our approach introduces a novel training recipe that combines a block diffusion mechanism with a complementary attention mask, enabling blockwise bidirectional context modeling without sacrificing AR training objectives. To further accelerate decoding, we design a hierarchical caching mechanism: a block-level cache that stores historical context representations across blocks, and a sub-block cache that enables efficient parallel generation within partially decoded blocks. Coupled with our parallel decoding pipeline, Fast-dLLM v2 achieves up to 2.5x speedup over standard AR decoding without compromising generation quality. Extensive experiments across diverse benchmarks demonstrate that Fast-dLLM v2 matches or surpasses AR baselines in accuracy, while delivering state-of-the-art efficiency among dLLMs - marking a significant step toward the practical deployment of fast and accurate LLMs. Code and model will be publicly released.</p> - https://arxiv.org/abs/2509.26328 - Tue, 30 Sep 2025 14:40:18 +0000 + Which Heads Matter for Reasoning? RL-Guided KV Cache Compression + https://arxiv.org/abs/2510.08525 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08525.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wenjie Du, Li Jiang, Keda Tao, Xue Liu, Huan Wang</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> Reasoning large language models exhibit complex reasoning behaviors through the extended chain-of-thought generation, creating unprecedented Key-Value (KV) cache overhead during the decoding phase. Existing KV cache compression methods underperform on reasoning models: token-dropping methods break reasoning integrity by discarding critical information, while head-reallocating methods mistakenly compress reasoning-critical heads since they are designed for retrieval tasks, resulting in significant performance degradation as compression rates increase. We hypothesize that KV heads exhibit functional heterogeneity in reasoning models-some heads are critical for chain-of-thought consistency while others are compressible. To validate and exploit this insight, we propose RLKV, a novel reasoning-critical head identification framework, which uses reinforcement learning to directly optimize the relationship between each head's cache usage and reasoning quality. As RLKV produces rewards from actual generated samples during training, it naturally identifies heads relevant to reasoning behaviors. We then allocate full KV cache to these heads while applying compressed constant KV cache to others for efficient inference. Our experiments reveal that only a small fraction of attention heads is essential for reasoning, enabling our KV compression approach to outperform baseline methods while achieving 20-50% cache reduction with near lossless performance compared to uncompressed results.</p> + https://arxiv.org/abs/2510.08525 + Thu, 09 Oct 2025 17:50:00 +0000 - Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents - https://arxiv.org/abs/2509.26354 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.26354.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shuai Shao, Qihan Ren, Chen Qian, Boyi Wei, Dadi Guo, Jingyi Yang, Xinhao Song, Linfeng Zhang, Weinan Zhang, Dongrui Liu, Jing Shao</p><p><b>Upvotes:</b> 16</p><p><b>Summary:</b> Advances in Large Language Models (LLMs) have enabled a new class of self-evolving agents that autonomously improve through interaction with the environment, demonstrating strong capabilities. However, self-evolution also introduces novel risks overlooked by current safety research. In this work, we study the case where an agent's self-evolution deviates in unintended ways, leading to undesirable or even harmful outcomes. We refer to this as Misevolution. To provide a systematic investigation, we evaluate misevolution along four key evolutionary pathways: model, memory, tool, and workflow. Our empirical findings reveal that misevolution is a widespread risk, affecting agents built even on top-tier LLMs (e.g., Gemini-2.5-Pro). Different emergent risks are observed in the self-evolutionary process, such as the degradation of safety alignment after memory accumulation, or the unintended introduction of vulnerabilities in tool creation and reuse. To our knowledge, this is the first study to systematically conceptualize misevolution and provide empirical evidence of its occurrence, highlighting an urgent need for new safety paradigms for self-evolving agents. Finally, we discuss potential mitigation strategies to inspire further research on building safer and more trustworthy self-evolving agents. Our code and data are available at https://github.com/ShaoShuai0605/Misevolution . Warning: this paper includes examples that may be offensive or harmful in nature.</p> - https://arxiv.org/abs/2509.26354 - Tue, 30 Sep 2025 14:55:55 +0000 + Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation + https://arxiv.org/abs/2510.08673 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08673.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kang Liao, Size Wu, Zhonghua Wu, Linyi Jin, Chao Wang, Yikai Wang, Fei Wang, Wei Li, Chen Change Loy</p><p><b>Upvotes:</b> 112</p><p><b>Summary:</b> Camera-centric understanding and generation are two cornerstones of spatial intelligence, yet they are typically studied in isolation. We present Puffin, a unified camera-centric multimodal model that extends spatial awareness along the camera dimension. Puffin integrates language regression and diffusion-based generation to interpret and create scenes from arbitrary viewpoints. To bridge the modality gap between cameras and vision-language, we introduce a novel paradigm that treats camera as language, enabling thinking with camera. This guides the model to align spatially grounded visual cues with photographic terminology while reasoning across geometric context. Puffin is trained on Puffin-4M, a large-scale dataset of 4 million vision-language-camera triplets. We incorporate both global camera parameters and pixel-wise camera maps, yielding flexible and reliable spatial generation. Experiments demonstrate Puffin superior performance over specialized models for camera-centric generation and understanding. With instruction tuning, Puffin generalizes to diverse cross-view tasks such as spatial imagination, world exploration, and photography guidance. We will release the code, models, dataset pipeline, and benchmark to advance multimodal spatial intelligence research.</p> + https://arxiv.org/abs/2510.08673 + Thu, 09 Oct 2025 17:59:29 +0000 - Game-Time: Evaluating Temporal Dynamics in Spoken Language Models - https://arxiv.org/abs/2509.26388 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.26388.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kai-Wei Chang, En-Pei Hu, Chun-Yi Kuan, Wenze Ren, Wei-Chih Chen, Guan-Ting Lin, Yu Tsao, Shao-Hua Sun, Hung-yi Lee, James Glass</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> Conversational Spoken Language Models (SLMs) are emerging as a promising paradigm for real-time speech interaction. However, their capacity of temporal dynamics, including the ability to manage timing, tempo and simultaneous speaking, remains a critical and unevaluated challenge for conversational fluency. To address this gap, we introduce the Game-Time Benchmark, a framework to systematically assess these temporal capabilities. Inspired by how humans learn a language through language activities, Game-Time consists of basic instruction-following tasks and advanced tasks with temporal constraints, such as tempo adherence and synchronized responses. Our evaluation of diverse SLM architectures reveals a clear performance disparity: while state-of-the-art models handle basic tasks well, many contemporary systems still struggle with fundamental instruction-following. More critically, nearly all models degrade substantially under temporal constraints, exposing persistent weaknesses in time awareness and full-duplex interaction. The Game-Time Benchmark provides a foundation for guiding future research toward more temporally-aware conversational AI. Demos and datasets are available on our project website https://ga642381.github.io/Game-Time.</p> - https://arxiv.org/abs/2509.26388 - Tue, 30 Sep 2025 15:23:39 +0000 + Don't Waste Mistakes: Leveraging Negative RL-Groups via Confidence Reweighting + https://arxiv.org/abs/2510.08696 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08696.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yunzhen Feng, Parag Jain, Anthony Hartshorn, Yaqi Duan, Julia Kempe</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> Reinforcement learning with verifiable rewards (RLVR) has become a standard recipe for improving large language models (LLMs) on reasoning tasks, with Group Relative Policy Optimization (GRPO) widely used in practice. Yet GRPO wastes substantial compute on negative groups: groups in which no sampled response is correct yield zero advantage and thus no gradient. We ask whether negative groups can be leveraged without extra supervision. Starting from a maximum-likelihood (MLE) objective in reward modeling, we show that the MLE gradient is equivalent to a policy gradient for a modified value function. This value function adds a confidence-weighted penalty on incorrect responses, imposing larger penalties on more confident mistakes. We refer to this as Likelihood Estimation with Negative Samples (LENS). LENS modifies GRPO to assign non-zero, confidence-dependent rewards to incorrect generations, making negative groups informative and converting previously wasted samples into useful gradient updates. On the MATH benchmark with Llama-3.1-8B and Qwen-2.5-3B, the proposed variant consistently outperforms GRPO baseline, with significant gains on harder items. These results demonstrate a principled and practical way to "rescue" negative groups, improving efficiency and performance in RLVR.</p> + https://arxiv.org/abs/2510.08696 + Thu, 09 Oct 2025 18:01:44 +0000 - Judging with Confidence: Calibrating Autoraters to Preference Distributions - https://arxiv.org/abs/2510.00263 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.00263.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhuohang Li, Xiaowei Li, Chengyu Huang, Guowang Li, Katayoon Goshvadi, Bo Dai, Dale Schuurmans, Paul Zhou, Hamid Palangi, Yiwen Song, Palash Goyal, Murat Kantarcioglu, Bradley A. Malin, Yuan Xue</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> The alignment of large language models (LLMs) with human values increasingly relies on using other LLMs as automated judges, or ``autoraters''. However, their reliability is limited by a foundational issue: they are trained on discrete preference labels, forcing a single ground truth onto tasks that are often subjective, ambiguous, or nuanced. We argue that a reliable autorater must learn to model the full distribution of preferences defined by a target population. In this paper, we propose a general framework for calibrating probabilistic autoraters to any given preference distribution. We formalize the problem and present two learning methods tailored to different data conditions: 1) a direct supervised fine-tuning for dense, probabilistic labels, and 2) a reinforcement learning approach for sparse, binary labels. Our empirical results show that finetuning autoraters with a distribution-matching objective leads to verbalized probability predictions that are better aligned with the target preference distribution, with improved calibration and significantly lower positional bias, all while preserving performance on objective tasks.</p> - https://arxiv.org/abs/2510.00263 - Tue, 30 Sep 2025 20:36:41 +0000 + BigCodeArena: Unveiling More Reliable Human Preferences in Code Generation via Execution + https://arxiv.org/abs/2510.08697 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08697.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Terry Yue Zhuo, Xiaolong Jin, Hange Liu, Juyong Jiang, Tianyang Liu, Chen Gong, Bhupesh Bishnoi, Vaisakhi Mishra, Marek Suppa, Noah Ziems, Saiteja Utpala, Ming Xu, Guangyu Song, Kaixin Li, Yuhan Cao, Bo Liu, Zheng Liu, Sabina Abdurakhmanova, Wenhao Yu, Mengzhao Jia, Jihan Yao, Kenneth Hamilton, Kumar Shridhar, Minh Chien Vu, Dingmin Wang, Jiawei Liu, Zijian Wang, Qian Liu, Binyuan Hui, Meg Risdal, Ahsen Khaliq, Atin Sood, Zhenchang Xing, Wasi Uddin Ahmad, John Grundy, David Lo, Banghua Zhu, Xiaoning Du, Torsten Scholak, Leandro von Werra</p><p><b>Upvotes:</b> 28</p><p><b>Summary:</b> Crowdsourced model evaluation platforms, such as Chatbot Arena, enable real-time evaluation from human perspectives to assess the quality of model responses. In the coding domain, manually examining the quality of LLM-generated content is extremely challenging, as it requires understanding long chunks of raw code and deliberately simulating code execution. To this end, we introduce BigCodeArena, an open human evaluation platform for code generation backed by a comprehensive and on-the-fly execution environment. Built on top of Chatbot Arena, BigCodeArena enables the execution of LLM-generated code and allows humans to interact with the execution process and outcomes. We collected over 14,000 raw code-centric conversation sessions across 10 widely used LLMs, spanning 10 languages and 8 types of execution environments. Among these conversations, we identified more than 4,700 multi-turn samples with pairwise human preferences. Further analysis uncovers underexplored preferences of LLMs in fine-grained domains characterized by tasks, languages, and frameworks. To systematically examine code understanding and generation capabilities of frontier LLMs, we curated two benchmarks based on the collected data, namely BigCodeReward and AutoCodeArena. For BigCodeReward, we post-processed the 4,700 conversations and evaluated the consistency between reward models and human preferences. The evaluation shows that most LLMs have superior performance in judging coding preferences when the execution results are available. Inspired by these findings, we propose AutoCodeArena, an automatic Elo rating benchmark designed to assess the coding quality of LLMs without human involvement. We find that proprietary LLMs like GPT-5, Claude-Sonnet-4, and Claude-Opus-4 still lead in code generation performance among recent emerging models.</p> + https://arxiv.org/abs/2510.08697 + Thu, 09 Oct 2025 18:01:47 +0000 - MOSS-Speech: Towards True Speech-to-Speech Models Without Text Guidance - https://arxiv.org/abs/2510.00499 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.00499.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xingjian Zhao, Zhe Xu, Qinyuan Cheng, Zhaoye Fei, Luozhijie Jin, Yang Wang, Hanfu Chen, Yaozhou Jiang, Qinghui Gao, Ke Chen, Ruixiao Li, Mingshu Chen, Ruiming Wang, Wenbo Zhang, Yiyang Zhang, Donghua Yu, Yang Gao, Xiaogui Yang, Yitian Gong, Yuanfan Xu, Yaqian Zhou, Xuanjing Huang, Xipeng Qiu</p><p><b>Upvotes:</b> 18</p><p><b>Summary:</b> Spoken dialogue systems often rely on cascaded pipelines that transcribe, process, and resynthesize speech. While effective, this design discards paralinguistic cues and limits expressivity. Recent end-to-end methods reduce latency and better preserve these cues, yet still rely on text intermediates, creating a fundamental bottleneck. We present MOSS-Speech, a true speech-to-speech large language model that directly understands and generates speech without relying on text guidance. Our approach combines a modality-based layer-splitting architecture with a frozen pre-training strategy, preserving the reasoning and knowledge of pretrained text LLMs while adding native speech capabilities. Experiments show that our model achieves state-of-the-art results in spoken question answering and delivers comparable speech-to-speech performance relative to existing text-guided systems, while still maintaining competitive text performance. By narrowing the gap between text-guided and direct speech generation, our work establishes a new paradigm for expressive and efficient end-to-end speech interaction.</p> - https://arxiv.org/abs/2510.00499 - Wed, 01 Oct 2025 04:32:37 +0000 + BEAR: Benchmarking and Enhancing Multimodal Language Models for Atomic Embodied Capabilities + https://arxiv.org/abs/2510.08759 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08759.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yu Qi, Haibo Zhao, Ziyu Guo, Siyuan Ma, Ziyan Chen, Yaokun Han, Renrui Zhang, Zitiantao Lin, Shiji Xin, Yijian Huang, Kai Cheng, Peiheng Wang, Jiazheng Liu, Jiayi Zhang, Yizhe Zhu, Wenqing Wang, Yiran Qin, Xupeng Zhu, Haojie Huang, Lawson L. S. Wong</p><p><b>Upvotes:</b> 44</p><p><b>Summary:</b> Embodied capabilities refer to a suite of fundamental abilities for an agent to perceive, comprehend, and interact with the physical world. While multimodal large language models (MLLMs) show promise as embodied agents, a thorough and systematic evaluation of their embodied capabilities remains underexplored, as existing benchmarks primarily focus on specific domains such as planning or spatial understanding. To bridge this gap, we introduce BEAR, a comprehensive and fine-grained benchmark that evaluates MLLMs on atomic embodied capabilities. BEAR comprises 4,469 interleaved image-video-text entries across 14 domains in 6 categories, including tasks from low-level pointing, trajectory understanding, spatial reasoning, to high-level planning. Extensive evaluation results of 20 representative MLLMs reveal their persistent limitations across all domains of embodied capabilities. To tackle the shortfall, we propose BEAR-Agent, a multimodal conversable agent that integrates pretrained vision models to strengthen MLLM perception, 3D understanding, and planning capabilities. It substantially enhances MLLM performance across diverse embodied capabilities on BEAR, yielding a 9.12% absolute gain and a relative improvement of 17.5% on GPT-5. Furthermore, our experiments indicate that improving MLLM embodied capabilities can benefit embodied tasks in simulated environments. Project website: https://bear-official66.github.io/</p> + https://arxiv.org/abs/2510.08759 + Thu, 09 Oct 2025 19:18:36 +0000 - Efficient Multi-modal Large Language Models via Progressive Consistency Distillation - https://arxiv.org/abs/2510.00515 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.00515.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zichen Wen, Shaobo Wang, Yufa Zhou, Junyuan Zhang, Qintong Zhang, Yifeng Gao, Zhaorun Chen, Bin Wang, Weijia Li, Conghui He, Linfeng Zhang</p><p><b>Upvotes:</b> 38</p><p><b>Summary:</b> Visual tokens consume substantial computational resources in multi-modal large models (MLLMs), significantly compromising their efficiency. Recent works have attempted to improve efficiency by compressing visual tokens during training, either through modifications to model components or by introducing additional parameters. However, they often overlook the increased learning difficulty caused by such compression, as the model's parameter space struggles to quickly adapt to the substantial perturbations in the feature space induced by token compression. In this work, we propose to develop Efficient MLLMs via Progressive Consistency Distillation (EPIC), a progressive learning framework. Specifically, by decomposing the feature space perturbations introduced by token compression along the token-wise and layer-wise dimensions, we introduce token consistency distillation and layer consistency distillation, respectively, aiming to reduce the training difficulty by leveraging guidance from a teacher model and following a progressive learning trajectory. Extensive experiments demonstrate the superior effectiveness, robustness, and generalization capabilities of our proposed framework.</p> - https://arxiv.org/abs/2510.00515 - Wed, 01 Oct 2025 04:56:40 +0000 + FinAuditing: A Financial Taxonomy-Structured Multi-Document Benchmark for Evaluating LLMs + https://arxiv.org/abs/2510.08886 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08886.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yan Wang, Keyi Wang, Shanshan Yang, Jaisal Patel, Jeff Zhao, Fengran Mo, Xueqing Peng, Lingfei Qian, Jimin Huang, Guojun Xiong, Xiao-Yang Liu, Jian-Yun Nie</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> The complexity of the Generally Accepted Accounting Principles (GAAP) and the hierarchical structure of eXtensible Business Reporting Language (XBRL) filings make financial auditing increasingly difficult to automate and verify. While large language models (LLMs) have demonstrated strong capabilities in unstructured text understanding, their ability to reason over structured, interdependent, and taxonomy-driven financial documents remains largely unexplored. To fill this gap, we introduce FinAuditing, the first taxonomy-aligned, structure-aware, multi-document benchmark for evaluating LLMs on financial auditing tasks. Built from real US-GAAP-compliant XBRL filings, FinAuditing defines three complementary subtasks, FinSM for semantic consistency, FinRE for relational consistency, and FinMR for numerical consistency, each targeting a distinct aspect of structured auditing reasoning. We further propose a unified evaluation framework integrating retrieval, classification, and reasoning metrics across these subtasks. Extensive zero-shot experiments on 13 state-of-the-art LLMs reveal that current models perform inconsistently across semantic, relational, and mathematical dimensions, with accuracy drops of up to 60-90% when reasoning over hierarchical multi-document structures. Our findings expose the systematic limitations of modern LLMs in taxonomy-grounded financial reasoning and establish FinAuditing as a foundation for developing trustworthy, structure-aware, and regulation-aligned financial intelligence systems. The benchmark dataset is available at Hugging Face.</p> + https://arxiv.org/abs/2510.08886 + Fri, 10 Oct 2025 00:41:55 +0000 - Large Reasoning Models Learn Better Alignment from Flawed Thinking - https://arxiv.org/abs/2510.00938 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.00938.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> ShengYun Peng, Eric Smith, Ivan Evtimov, Song Jiang, Pin-Yu Chen, Hongyuan Zhan, Haozhu Wang, Duen Horng Chau, Mahesh Pasupuleti, Jianfeng Chi</p><p><b>Upvotes:</b> 52</p><p><b>Summary:</b> Large reasoning models (LRMs) "think" by generating structured chain-of-thought (CoT) before producing a final answer, yet they still lack the ability to reason critically about safety alignment and are easily biased when a flawed premise is injected into their thought process. We propose RECAP (Robust Safety Alignment via Counter-Aligned Prefilling), a principled reinforcement learning (RL) method for post-training that explicitly teaches models to override flawed reasoning trajectories and reroute to safe and helpful responses. RECAP trains on a mixture of synthetically generated counter-aligned CoT prefills and standard prompts, requires no additional training cost or modifications beyond vanilla reinforcement learning from human feedback (RLHF), and substantially improves safety and jailbreak robustness, reduces overrefusal, and preserves core reasoning capability -- all while maintaining inference token budget. Extensive analysis shows that RECAP-trained models engage in self-reflection more frequently and remain robust under adaptive attacks, preserving safety even after repeated attempts to override their reasoning.</p> - https://arxiv.org/abs/2510.00938 - Wed, 01 Oct 2025 14:15:43 +0000 + On Epistemic Uncertainty of Visual Tokens for Object Hallucinations in Large Vision-Language Models + https://arxiv.org/abs/2510.09008 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09008.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hoigi Seo, Dong Un Kang, Hyunjin Cho, Joohoon Lee, Se Young Chun</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Large vision-language models (LVLMs), which integrate a vision encoder (VE) with a large language model, have achieved remarkable success across various tasks. However, there are still crucial challenges in LVLMs such as object hallucination, generating descriptions of objects that are not in the input image. Here, we argue that uncertain visual tokens within the VE is a key factor that contributes to object hallucination. Our statistical analysis found that there are positive correlations between visual tokens with high epistemic uncertainty and the occurrence of hallucinations. Furthermore, we show theoretically and empirically that visual tokens in early VE layers that exhibit large representation deviations under small adversarial perturbations indicate high epistemic uncertainty. Based on these findings, we propose a simple yet effective strategy to mitigate object hallucination by modifying the VE only. Our method comprises a proxy method with adversarial perturbations for identifying uncertain visual tokens efficiently and a method to mask these uncertain visual tokens during the self-attention process in the middle layers of the VE, suppressing their influence on visual encoding and thus alleviating hallucinations. Extensive experiments show that our method significantly reduces object hallucinations in LVLMs and can synergistically work with other prior arts.</p> + https://arxiv.org/abs/2510.09008 + Fri, 10 Oct 2025 05:12:52 +0000 - Compose Your Policies! Improving Diffusion-based or Flow-based Robot Policies via Test-time Distribution-level Composition - https://arxiv.org/abs/2510.01068 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01068.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiahang Cao, Yize Huang, Hanzhong Guo, Rui Zhang, Mu Nan, Weijian Mai, Jiaxu Wang, Hao Cheng, Jingkai Sun, Gang Han, Wen Zhao, Qiang Zhang, Yijie Guo, Qihao Zheng, Chunfeng Song, Xiao Li, Ping Luo, Andrew F. Luo</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> Diffusion-based models for robotic control, including vision-language-action (VLA) and vision-action (VA) policies, have demonstrated significant capabilities. Yet their advancement is constrained by the high cost of acquiring large-scale interaction datasets. This work introduces an alternative paradigm for enhancing policy performance without additional model training. Perhaps surprisingly, we demonstrate that the composed policies can exceed the performance of either parent policy. Our contribution is threefold. First, we establish a theoretical foundation showing that the convex composition of distributional scores from multiple diffusion models can yield a superior one-step functional objective compared to any individual score. A Gr\"onwall-type bound is then used to show that this single-step improvement propagates through entire generation trajectories, leading to systemic performance gains. Second, motivated by these results, we propose General Policy Composition (GPC), a training-free method that enhances performance by combining the distributional scores of multiple pre-trained policies via a convex combination and test-time search. GPC is versatile, allowing for the plug-and-play composition of heterogeneous policies, including VA and VLA models, as well as those based on diffusion or flow-matching, irrespective of their input visual modalities. Third, we provide extensive empirical validation. Experiments on Robomimic, PushT, and RoboTwin benchmarks, alongside real-world robotic evaluations, confirm that GPC consistently improves performance and adaptability across a diverse set of tasks. Further analysis of alternative composition operators and weighting strategies offers insights into the mechanisms underlying the success of GPC. These results establish GPC as a simple yet effective method for improving control performance by leveraging existing policies.</p> - https://arxiv.org/abs/2510.01068 - Wed, 01 Oct 2025 16:05:53 +0000 + Large Language Models Do NOT Really Know What They Don't Know + https://arxiv.org/abs/2510.09033 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09033.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chi Seng Cheang, Hou Pong Chan, Wenxuan Zhang, Yang Deng</p><p><b>Upvotes:</b> 16</p><p><b>Summary:</b> Recent work suggests that large language models (LLMs) encode factuality signals in their internal representations, such as hidden states, attention weights, or token probabilities, implying that LLMs may "know what they don't know". However, LLMs can also produce factual errors by relying on shortcuts or spurious associations. These error are driven by the same training objective that encourage correct predictions, raising the question of whether internal computations can reliably distinguish between factual and hallucinated outputs. In this work, we conduct a mechanistic analysis of how LLMs internally process factual queries by comparing two types of hallucinations based on their reliance on subject information. We find that when hallucinations are associated with subject knowledge, LLMs employ the same internal recall process as for correct responses, leading to overlapping and indistinguishable hidden-state geometries. In contrast, hallucinations detached from subject knowledge produce distinct, clustered representations that make them detectable. These findings reveal a fundamental limitation: LLMs do not encode truthfulness in their internal states but only patterns of knowledge recall, demonstrating that "LLMs don't really know what they don't know".</p> + https://arxiv.org/abs/2510.09033 + Fri, 10 Oct 2025 06:09:04 +0000 - Apriel-1.5-15b-Thinker - https://arxiv.org/abs/2510.01141 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01141.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shruthan Radhakrishna, Aman Tiwari, Aanjaneya Shukla, Masoud Hashemi, Rishabh Maheshwary, Shiva Krishna Reddy Malay, Jash Mehta, Pulkit Pattnaik, Saloni Mittal, Khalil Slimi, Kelechi Ogueji, Akintunde Oladipo, Soham Parikh, Oluwanifemi Bamgbose, Toby Liang, Ahmed Masry, Khyati Mahajan, Sai Rajeswar Mudumba, Vikas Yadav, Sathwik Tejaswi Madhusudhan, Torsten Scholak, Sagar Davasam, Srinivas Sunkara, Nicholas Chapados</p><p><b>Upvotes:</b> 105</p><p><b>Summary:</b> We present Apriel-1.5-15B-Thinker, a 15-billion parameter open-weights multimodal reasoning model that achieves frontier-level performance through training design rather than sheer scale. Starting from Pixtral-12B, we apply a progressive three-stage methodology: (1) depth upscaling to expand reasoning capacity without pretraining from scratch, (2) staged continual pre-training that first develops foundational text and vision understanding, then enhances visual reasoning through targeted synthetic data generation addressing spatial structure, compositional understanding, and fine-grained perception, and (3) high-quality text-only supervised fine-tuning on curated instruction-response pairs with explicit reasoning traces spanning mathematics, coding, science, and tool use. Notably, our model achieves competitive results without reinforcement learning or preference optimization, isolating the contribution of our data-centric continual pre-training approach. On the Artificial Analysis Intelligence Index, Apriel-1.5-15B-Thinker attains a score of 52, matching DeepSeek-R1-0528 despite requiring significantly fewer computational resources. Across ten image benchmarks, its performance is on average within five points of Gemini-2.5-Flash and Claude Sonnet-3.7, a key achievement for a model operating within single-GPU deployment constraints. Our results demonstrate that thoughtful mid-training 2 design can close substantial capability gaps without massive scale, making frontier-level multimodal reasoning accessible to organizations with limited infrastructure. We release the model checkpoint, all training recipes, and evaluation protocols under the MIT license to to advance open-source research.</p> - https://arxiv.org/abs/2510.01141 - Wed, 01 Oct 2025 17:29:35 +0000 + DITING: A Multi-Agent Evaluation Framework for Benchmarking Web Novel Translation + https://arxiv.org/abs/2510.09116 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09116.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Enze Zhang, Jiaying Wang, Mengxi Xiao, Jifei Liu, Ziyan Kuang, Rui Dong, Eric Dong, Sophia Ananiadou, Min Peng, Qianqian Xie</p><p><b>Upvotes:</b> 94</p><p><b>Summary:</b> Large language models (LLMs) have substantially advanced machine translation (MT), yet their effectiveness in translating web novels remains unclear. Existing benchmarks rely on surface-level metrics that fail to capture the distinctive traits of this genre. To address these gaps, we introduce DITING, the first comprehensive evaluation framework for web novel translation, assessing narrative and cultural fidelity across six dimensions: idiom translation, lexical ambiguity, terminology localization, tense consistency, zero-pronoun resolution, and cultural safety, supported by over 18K expert-annotated Chinese-English sentence pairs. We further propose AgentEval, a reasoning-driven multi-agent evaluation framework that simulates expert deliberation to assess translation quality beyond lexical overlap, achieving the highest correlation with human judgments among seven tested automatic metrics. To enable metric comparison, we develop MetricAlign, a meta-evaluation dataset of 300 sentence pairs annotated with error labels and scalar quality scores. Comprehensive evaluation of fourteen open, closed, and commercial models reveals that Chinese-trained LLMs surpass larger foreign counterparts, and that DeepSeek-V3 delivers the most faithful and stylistically coherent translations. Our work establishes a new paradigm for exploring LLM-based web novel translation and provides public resources to advance future research.</p> + https://arxiv.org/abs/2510.09116 + Fri, 10 Oct 2025 08:10:10 +0000 - Prosperity before Collapse: How Far Can Off-Policy RL Reach with Stale Data on LLMs? - https://arxiv.org/abs/2510.01161 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01161.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haizhong Zheng, Jiawei Zhao, Bedi Chen</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> Reinforcement learning has been central to recent advances in large language model reasoning, but most algorithms rely on on-policy training that demands fresh rollouts at every update, limiting efficiency and scalability. Asynchronous RL systems alleviate this by decoupling rollout generation from training, yet their effectiveness hinges on tolerating large staleness in rollout data, a setting where existing methods either degrade in performance or collapse. We revisit this challenge and uncover a prosperity-before-collapse phenomenon: stale data can be as informative as on-policy data if exploited properly. Building on this insight, we introduce M2PO (Second-Moment Trust Policy Optimization), which constrains the second moment of importance weights to suppress only extreme outliers while preserving informative updates. Notably, M2PO sharply reduces the fraction of clipped tokens under high staleness (from 1.22% to 0.06% over training), precisely masking high-variance tokens while maintaining stable optimization. Extensive evaluation across six models (from 1.7B to 32B) and eight benchmarks shows that M2PO delivers stable off-policy training even with data stale by at least 256 model updates and matches on-policy performance.</p> - https://arxiv.org/abs/2510.01161 - Wed, 01 Oct 2025 17:48:23 +0000 + Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs + https://arxiv.org/abs/2510.09201 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09201.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yumin Choi, Dongki Kim, Jinheon Baek, Sung Ju Hwang</p><p><b>Upvotes:</b> 45</p><p><b>Summary:</b> Large Language Models (LLMs) have shown remarkable success, and their multimodal expansions (MLLMs) further unlock capabilities spanning images, videos, and other modalities beyond text. However, despite this shift, prompt optimization approaches, designed to reduce the burden of manual prompt crafting while maximizing performance, remain confined to text, ultimately limiting the full potential of MLLMs. Motivated by this gap, we introduce the new problem of multimodal prompt optimization, which expands the prior definition of prompt optimization to the multimodal space defined by the pairs of textual and non-textual prompts. To tackle this problem, we then propose the Multimodal Prompt Optimizer (MPO), a unified framework that not only performs the joint optimization of multimodal prompts through alignment-preserving updates but also guides the selection process of candidate prompts by leveraging earlier evaluations as priors in a Bayesian-based selection strategy. Through extensive experiments across diverse modalities that go beyond text, such as images, videos, and even molecules, we demonstrate that MPO outperforms leading text-only optimization methods, establishing multimodal prompt optimization as a crucial step to realizing the potential of MLLMs.</p> + https://arxiv.org/abs/2510.09201 + Fri, 10 Oct 2025 09:41:25 +0000 - OpenTSLM: Time-Series Language Models for Reasoning over Multivariate Medical Text- and Time-Series Data - https://arxiv.org/abs/2510.02410 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.02410.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Patrick Langer, Thomas Kaar, Max Rosenblattl, Maxwell A. Xu, Winnie Chow, Martin Maritsch, Aradhana Verma, Brian Han, Daniel Seung Kim, Henry Chubb, Scott Ceresnak, Aydin Zahedivash, Alexander Tarlochan Singh Sandhu, Fatima Rodriguez, Daniel McDuff, Elgar Fleisch, Oliver Aalami, Filipe Barata, Paul Schmiedmayer</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> LLMs have emerged as powerful tools for interpreting multimodal data. In medicine, they hold particular promise for synthesizing large volumes of clinical information into actionable insights and digital health applications. Yet, a major limitation remains their inability to handle time series. To overcome this gap, we present OpenTSLM, a family of Time Series Language Models (TSLMs) created by integrating time series as a native modality to pretrained LLMs, enabling reasoning over multiple time series of any length. We investigate two architectures for OpenTSLM. The first, OpenTSLM-SoftPrompt, models time series implicitly by concatenating learnable time series tokens with text tokens via soft prompting. Although parameter-efficient, we hypothesize that explicit time series modeling scales better and outperforms implicit approaches. We thus introduce OpenTSLM-Flamingo, which integrates time series with text via cross-attention. We benchmark both variants against baselines that treat time series as text tokens or plots, across a suite of text-time-series Chain-of-Thought (CoT) reasoning tasks. We introduce three datasets: HAR-CoT, Sleep-CoT, and ECG-QA-CoT. Across all, OpenTSLM models outperform baselines, reaching 69.9 F1 in sleep staging and 65.4 in HAR, compared to 9.05 and 52.2 for finetuned text-only models. Notably, even 1B-parameter OpenTSLM models surpass GPT-4o (15.47 and 2.95). OpenTSLM-Flamingo matches OpenTSLM-SoftPrompt in performance and outperforms on longer sequences, while maintaining stable memory requirements. By contrast, SoftPrompt grows exponentially in memory with sequence length, requiring around 110 GB compared to 40 GB VRAM when training on ECG-QA with LLaMA-3B. Expert reviews by clinicians find strong reasoning capabilities exhibited by OpenTSLMs on ECG-QA. To facilitate further research, we provide all code, datasets, and models open-source.</p> - https://arxiv.org/abs/2510.02410 - Thu, 02 Oct 2025 09:58:23 +0000 + Spotlight on Token Perception for Multimodal Reinforcement Learning + https://arxiv.org/abs/2510.09285 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09285.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Siyuan Huang, Xiaoye Qu, Yafu Li, Yun Luo, Zefeng He, Daizong Liu, Yu Cheng</p><p><b>Upvotes:</b> 35</p><p><b>Summary:</b> While Reinforcement Learning with Verifiable Rewards (RLVR) has advanced the reasoning capabilities of Large Vision-Language Models (LVLMs), most existing methods in multimodal reasoning neglect the critical role of visual perception within the RLVR optimization process. In this paper, we undertake a pioneering exploration of multimodal RLVR through the novel perspective of token perception, which measures the visual dependency of each generated token. With a granular analysis of Chain-of-Thought (CoT) processes, we uncover two key insights: first, token perception in a rollout trajectory is sparsely distributed, where only a small fraction of tokens have high visual dependency for visually-grounded reasoning; second, different trajectories exhibit significant divergence in their overall visual dependency. Based on these observations, we propose Visually-Perceptive Policy Optimization (VPPO), a novel policy gradient algorithm that explicitly leverages token perception to refine the learning signal. Specifically, VPPO achieves this through a dual mechanism: it reweights a trajectory's advantage by its overall visual dependency, and focuses policy updates exclusively on perceptually pivotal tokens. On a comprehensive suite of eight perception and reasoning benchmarks, VPPO demonstrates substantial gains over leading open-source RL-tuned models, with its effectiveness consistently validated across 7B and 32B model scales. Our findings not only establish a new token-level perceptual perspective for analyzing multimodal RLVR but also present a novel and effective optimization strategy to significantly enhance the multimodal reasoning capabilities of LVLMs.</p> + https://arxiv.org/abs/2510.09285 + Fri, 10 Oct 2025 11:25:33 +0000 - Patch-as-Decodable-Token: Towards Unified Multi-Modal Vision Tasks in MLLMs - https://arxiv.org/abs/2510.01954 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01954.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yongyi Su, Haojie Zhang, Shijie Li, Nanqing Liu, Jingyi Liao, Junyi Pan, Yuan Liu, Xiaofen Xing, Chong Sun, Chen Li, Nancy F. Chen, Shuicheng Yan, Xulei Yang, Xun Xu</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Multimodal large language models (MLLMs) have advanced rapidly in recent years. However, existing approaches for vision tasks often rely on indirect representations, such as generating coordinates as text for detection, which limits performance and prevents dense prediction tasks like segmentation. To overcome these challenges, we introduce Patch-as-Decodable Token (PaDT), a unified paradigm that enables MLLMs to directly generate both textual and diverse visual outputs. Central to PaDT are Visual Reference Tokens (VRTs), derived from visual patch embeddings of query images and interleaved seamlessly with LLM's output textual tokens. A lightweight decoder then transforms LLM's outputs into detection, segmentation, and grounding predictions. Unlike prior methods, PaDT processes VRTs independently at each forward pass and dynamically expands the embedding table, thus improving localization and differentiation among similar objects. We further tailor a training strategy for PaDT by randomly selecting VRTs for supervised fine-tuning and introducing a robust per-token cross-entropy loss. Our empirical studies across four visual perception and understanding tasks suggest PaDT consistently achieving state-of-the-art performance, even compared with significantly larger MLLM models. The code is available at https://github.com/Gorilla-Lab-SCUT/PaDT.</p> - https://arxiv.org/abs/2510.01954 - Thu, 02 Oct 2025 12:23:57 +0000 + KORMo: Korean Open Reasoning Model for Everyone + https://arxiv.org/abs/2510.09426 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09426.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Minjun Kim, Hyeonseok Lim, Hangyeol Yoo, Inho Won, Seungwoo Song, Minkyung Cho, Junhun Yuk, Changsu Choi, Dongjae Shin, Huige Lee, Hoyun Song, Alice Oh, Kyungtae Lim</p><p><b>Upvotes:</b> 62</p><p><b>Summary:</b> This work presents the first large-scale investigation into constructing a fully open bilingual large language model (LLM) for a non-English language, specifically Korean, trained predominantly on synthetic data. We introduce KORMo-10B, a 10.8B-parameter model trained from scratch on a Korean-English corpus in which 68.74% of the Korean portion is synthetic. Through systematic experimentation, we demonstrate that synthetic data, when carefully curated with balanced linguistic coverage and diverse instruction styles, does not cause instability or degradation during large-scale pretraining. Furthermore, the model achieves performance comparable to that of contemporary open-weight multilingual baselines across a wide range of reasoning, knowledge, and instruction-following benchmarks. Our experiments reveal two key findings: (1) synthetic data can reliably sustain long-horizon pretraining without model collapse, and (2) bilingual instruction tuning enables near-native reasoning and discourse coherence in Korean. By fully releasing all components including data, code, training recipes, and logs, this work establishes a transparent framework for developing synthetic data-driven fully open models (FOMs) in low-resource settings and sets a reproducible precedent for future multilingual LLM research.</p> + https://arxiv.org/abs/2510.09426 + Fri, 10 Oct 2025 14:31:25 +0000 - Self-Improvement in Multimodal Large Language Models: A Survey - https://arxiv.org/abs/2510.02665 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.02665.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shijian Deng, Kai Wang, Tianyu Yang, Harsh Singh, Yapeng Tian</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Recent advancements in self-improvement for Large Language Models (LLMs) have efficiently enhanced model capabilities without significantly increasing costs, particularly in terms of human effort. While this area is still relatively young, its extension to the multimodal domain holds immense potential for leveraging diverse data sources and developing more general self-improving models. This survey is the first to provide a comprehensive overview of self-improvement in Multimodal LLMs (MLLMs). We provide a structured overview of the current literature and discuss methods from three perspectives: 1) data collection, 2) data organization, and 3) model optimization, to facilitate the further development of self-improvement in MLLMs. We also include commonly used evaluations and downstream applications. Finally, we conclude by outlining open challenges and future research directions.</p> - https://arxiv.org/abs/2510.02665 - Fri, 03 Oct 2025 01:48:26 +0000 + SPG: Sandwiched Policy Gradient for Masked Diffusion Language Models + https://arxiv.org/abs/2510.09541 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09541.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chenyu Wang, Paria Rashidinejad, DiJia Su, Song Jiang, Sid Wang, Siyan Zhao, Cai Zhou, Shannon Zejiang Shen, Feiyu Chen, Tommi Jaakkola, Yuandong Tian, Bo Liu</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Diffusion large language models (dLLMs) are emerging as an efficient alternative to autoregressive models due to their ability to decode multiple tokens in parallel. However, aligning dLLMs with human preferences or task-specific rewards via reinforcement learning (RL) is challenging because their intractable log-likelihood precludes the direct application of standard policy gradient methods. While prior work uses surrogates like the evidence lower bound (ELBO), these one-sided approximations can introduce significant policy gradient bias. To address this, we propose the Sandwiched Policy Gradient (SPG) that leverages both an upper and a lower bound of the true log-likelihood. Experiments show that SPG significantly outperforms baselines based on ELBO or one-step estimation. Specifically, SPG improves the accuracy over state-of-the-art RL methods for dLLMs by 3.6% in GSM8K, 2.6% in MATH500, 18.4% in Countdown and 27.0% in Sudoku.</p> + https://arxiv.org/abs/2510.09541 + Fri, 10 Oct 2025 16:52:25 +0000 - Self-Reflective Generation at Test Time - https://arxiv.org/abs/2510.02919 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.02919.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jian Mu, Qixin Zhang, Zhiyong Wang, Menglin Yang, Shuang Qiu, Chengwei Qin, Zhongxiang Dai, Yao Shu</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Large language models (LLMs) increasingly solve complex reasoning tasks via long chain-of-thought, but their forward-only autoregressive generation process is fragile; early token errors can cascade, which creates a clear need for self-reflection mechanisms. However, existing self-reflection either performs revisions over full drafts or learns self-correction via expensive training, both fundamentally reactive and inefficient. To address this, we propose Self-Reflective Generation at Test Time (SRGen), a lightweight test-time framework that reflects before generating at uncertain points. During token generation, SRGen utilizes dynamic entropy thresholding to identify high-uncertainty tokens. For each identified token, it trains a specific corrective vector, which fully exploits the already generated context for a self-reflective generation to correct the token probability distribution. By retrospectively analyzing the partial output, this self-reflection enables more trustworthy decisions, thereby significantly reducing the probability of errors at highly uncertain points. Evaluated on challenging mathematical reasoning benchmarks and a diverse set of LLMs, SRGen can consistently strengthen model reasoning: improvements in single-pass quality also translate into stronger self-consistency voting. Especially, on AIME2024 with DeepSeek-R1-Distill-Qwen-7B, SRGen yields absolute improvements of +12.0% on Pass@1 and +13.3% on Cons@5. Moreover, our findings position SRGen as a plug-and-play method that integrates reflection into the generation process for reliable LLM reasoning, achieving consistent gains with bounded overhead and broad composability with other training-time (e.g., RLHF) and test-time (e.g., SLOT) techniques.</p> - https://arxiv.org/abs/2510.02919 - Fri, 03 Oct 2025 11:46:04 +0000 + AutoPR: Let's Automate Your Academic Promotion! + https://arxiv.org/abs/2510.09558 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09558.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qiguang Chen, Zheng Yan, Mingda Yang, Libo Qin, Yixin Yuan, Hanjing Li, Jinhao Liu, Yiyan Ji, Dengyun Peng, Jiannan Guan, Mengkang Hu, Yantao Du, Wanxiang Che</p><p><b>Upvotes:</b> 48</p><p><b>Summary:</b> As the volume of peer-reviewed research surges, scholars increasingly rely on social platforms for discovery, while authors invest considerable effort in promoting their work to ensure visibility and citations. To streamline this process and reduce the reliance on human effort, we introduce Automatic Promotion (AutoPR), a novel task that transforms research papers into accurate, engaging, and timely public content. To enable rigorous evaluation, we release PRBench, a multimodal benchmark that links 512 peer-reviewed articles to high-quality promotional posts, assessing systems along three axes: Fidelity (accuracy and tone), Engagement (audience targeting and appeal), and Alignment (timing and channel optimization). We also introduce PRAgent, a multi-agent framework that automates AutoPR in three stages: content extraction with multimodal preparation, collaborative synthesis for polished outputs, and platform-specific adaptation to optimize norms, tone, and tagging for maximum reach. When compared to direct LLM pipelines on PRBench, PRAgent demonstrates substantial improvements, including a 604% increase in total watch time, a 438% rise in likes, and at least a 2.9x boost in overall engagement. Ablation studies show that platform modeling and targeted promotion contribute the most to these gains. Our results position AutoPR as a tractable, measurable research problem and provide a roadmap for scalable, impactful automated scholarly communication.</p> + https://arxiv.org/abs/2510.09558 + Fri, 10 Oct 2025 17:08:36 +0000 - Taming Text-to-Sounding Video Generation via Advanced Modality Condition and Interaction - https://arxiv.org/abs/2510.03117 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03117.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kaisi Guan, Xihua Wang, Zhengfeng Lai, Xin Cheng, Peng Zhang, XiaoJiang Liu, Ruihua Song, Meng Cao</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> This study focuses on a challenging yet promising task, Text-to-Sounding-Video (T2SV) generation, which aims to generate a video with synchronized audio from text conditions, meanwhile ensuring both modalities are aligned with text. Despite progress in joint audio-video training, two critical challenges still remain unaddressed: (1) a single, shared text caption where the text for video is equal to the text for audio often creates modal interference, confusing the pretrained backbones, and (2) the optimal mechanism for cross-modal feature interaction remains unclear. To address these challenges, we first propose the Hierarchical Visual-Grounded Captioning (HVGC) framework that generates pairs of disentangled captions, a video caption, and an audio caption, eliminating interference at the conditioning stage. Based on HVGC, we further introduce BridgeDiT, a novel dual-tower diffusion transformer, which employs a Dual CrossAttention (DCA) mechanism that acts as a robust ``bridge" to enable a symmetric, bidirectional exchange of information, achieving both semantic and temporal synchronization. Extensive experiments on three benchmark datasets, supported by human evaluations, demonstrate that our method achieves state-of-the-art results on most metrics. Comprehensive ablation studies further validate the effectiveness of our contributions, offering key insights for the future T2SV task. All the codes and checkpoints will be publicly released.</p> - https://arxiv.org/abs/2510.03117 - Fri, 03 Oct 2025 15:43:56 +0000 + SpaceVista: All-Scale Visual Spatial Reasoning from mm to km + https://arxiv.org/abs/2510.09606 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09606.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Peiwen Sun, Shiqiang Lang, Dongming Wu, Yi Ding, Kaituo Feng, Huadai Liu, Zhen Ye, Rui Liu, Yun-Hui Liu, Jianan Wang, Xiangyu Yue</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> With the current surge in spatial reasoning explorations, researchers have made significant progress in understanding indoor scenes, but still struggle with diverse applications such as robotics and autonomous driving. This paper aims to advance all-scale spatial reasoning across diverse scenarios by tackling two key challenges: 1) the heavy reliance on indoor 3D scans and labor-intensive manual annotations for dataset curation; 2) the absence of effective all-scale scene modeling, which often leads to overfitting to individual scenes. In this paper, we introduce a holistic solution that integrates a structured spatial reasoning knowledge system, scale-aware modeling, and a progressive training paradigm, as the first attempt to broaden the all-scale spatial intelligence of MLLMs to the best of our knowledge. Using a task-specific, specialist-driven automated pipeline, we curate over 38K video scenes across 5 spatial scales to create SpaceVista-1M, a dataset comprising approximately 1M spatial QA pairs spanning 19 diverse task types. While specialist models can inject useful domain knowledge, they are not reliable for evaluation. We then build an all-scale benchmark with precise annotations by manually recording, retrieving, and assembling video-based data. However, naive training with SpaceVista-1M often yields suboptimal results due to the potential knowledge conflict. Accordingly, we introduce SpaceVista-7B, a spatial reasoning model that accepts dense inputs beyond semantics and uses scale as an anchor for scale-aware experts and progressive rewards. Finally, extensive evaluations across 5 benchmarks, including our SpaceVista-Bench, demonstrate competitive performance, showcasing strong generalization across all scales and scenarios. Our dataset, model, and benchmark will be released on https://peiwensun2000.github.io/mm2km .</p> + https://arxiv.org/abs/2510.09606 + Fri, 10 Oct 2025 17:59:46 +0000 - CoDA: Agentic Systems for Collaborative Data Visualization - https://arxiv.org/abs/2510.03194 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03194.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zichen Chen, Jiefeng Chen, Sercan Ö. Arik, Misha Sra, Tomas Pfister, Jinsung Yoon</p><p><b>Upvotes:</b> 27</p><p><b>Summary:</b> Deep research has revolutionized data analysis, yet data scientists still devote substantial time to manually crafting visualizations, highlighting the need for robust automation from natural language queries. However, current systems struggle with complex datasets containing multiple files and iterative refinement. Existing approaches, including simple single- or multi-agent systems, often oversimplify the task, focusing on initial query parsing while failing to robustly manage data complexity, code errors, or final visualization quality. In this paper, we reframe this challenge as a collaborative multi-agent problem. We introduce CoDA, a multi-agent system that employs specialized LLM agents for metadata analysis, task planning, code generation, and self-reflection. We formalize this pipeline, demonstrating how metadata-focused analysis bypasses token limits and quality-driven refinement ensures robustness. Extensive evaluations show CoDA achieves substantial gains in the overall score, outperforming competitive baselines by up to 41.5%. This work demonstrates that the future of visualization automation lies not in isolated code generation but in integrated, collaborative agentic workflows.</p> - https://arxiv.org/abs/2510.03194 - Fri, 03 Oct 2025 17:30:16 +0000 + StreamingVLM: Real-Time Understanding for Infinite Video Streams + https://arxiv.org/abs/2510.09608 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09608.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ruyi Xu, Guangxuan Xiao, Yukang Chen, Liuning He, Kelly Peng, Yao Lu, Song Han</p><p><b>Upvotes:</b> 42</p><p><b>Summary:</b> Vision-language models (VLMs) could power real-time assistants and autonomous agents, but they face a critical challenge: understanding near-infinite video streams without escalating latency and memory usage. Processing entire videos with full attention leads to quadratic computational costs and poor performance on long videos. Meanwhile, simple sliding window methods are also flawed, as they either break coherence or suffer from high latency due to redundant recomputation. In this paper, we introduce StreamingVLM, a model designed for real-time, stable understanding of infinite visual input. Our approach is a unified framework that aligns training with streaming inference. During inference, we maintain a compact KV cache by reusing states of attention sinks, a short window of recent vision tokens, and a long window of recent text tokens. This streaming ability is instilled via a simple supervised fine-tuning (SFT) strategy that applies full attention on short, overlapped video chunks, which effectively mimics the inference-time attention pattern without training on prohibitively long contexts. For evaluation, we build Inf-Streams-Eval, a new benchmark with videos averaging over two hours that requires dense, per-second alignment between frames and text. On Inf-Streams-Eval, StreamingVLM achieves a 66.18% win rate against GPT-4O mini and maintains stable, real-time performance at up to 8 FPS on a single NVIDIA H100. Notably, our SFT strategy also enhances general VQA abilities without any VQA-specific fine-tuning, improving performance on LongVideoBench by +4.30 and OVOBench Realtime by +5.96. Code is available at https://github.com/mit-han-lab/streaming-vlm.</p> + https://arxiv.org/abs/2510.09608 + Fri, 10 Oct 2025 17:59:58 +0000 - Cache-to-Cache: Direct Semantic Communication Between Large Language Models - https://arxiv.org/abs/2510.03215 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03215.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Tianyu Fu, Zihan Min, Hanling Zhang, Jichao Yan, Guohao Dai, Wanli Ouyang, Yu Wang</p><p><b>Upvotes:</b> 83</p><p><b>Summary:</b> Multi-LLM systems harness the complementary strengths of diverse Large Language Models, achieving performance and efficiency gains unattainable by a single model. In existing designs, LLMs communicate through text, forcing internal representations to be transformed into output token sequences. This process both loses rich semantic information and incurs token-by-token generation latency. Motivated by these limitations, we ask: Can LLMs communicate beyond text? Oracle experiments show that enriching the KV-Cache semantics can improve response quality without increasing cache size, supporting KV-Cache as an effective medium for inter-model communication. Thus, we propose Cache-to-Cache (C2C), a new paradigm for direct semantic communication between LLMs. C2C uses a neural network to project and fuse the source model's KV-cache with that of the target model to enable direct semantic transfer. A learnable gating mechanism selects the target layers that benefit from cache communication. Compared with text communication, C2C utilizes the deep, specialized semantics from both models, while avoiding explicit intermediate text generation. Experiments show that C2C achieves 8.5-10.5% higher average accuracy than individual models. It further outperforms the text communication paradigm by approximately 3.0-5.0%, while delivering an average 2.0x speedup in latency. Our code is available at https://github.com/thu-nics/C2C.</p> - https://arxiv.org/abs/2510.03215 - Fri, 03 Oct 2025 17:52:32 +0000 + Building a Foundational Guardrail for General Agentic Systems via Synthetic Data + https://arxiv.org/abs/2510.09781 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.09781.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yue Huang, Hang Hua, Yujun Zhou, Pengcheng Jing, Manish Nagireddy, Inkit Padhi, Greta Dolcetti, Zhangchen Xu, Subhajit Chaudhury, Ambrish Rawat, Liubov Nedoshivina, Pin-Yu Chen, Prasanna Sattigeri, Xiangliang Zhang</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> While LLM agents can plan multi-step tasks, intervening at the planning stage-before any action is executed-is often the safest way to prevent harm, since certain risks can lead to severe consequences once carried out. However, existing guardrails mostly operate post-execution, which is difficult to scale and leaves little room for controllable supervision at the plan level. To address this challenge, we highlight three critical gaps in current research: data gap, model gap, and evaluation gap. To close the data gap, we introduce AuraGen, a controllable engine that (i) synthesizes benign trajectories, (ii) injects category-labeled risks with calibrated difficulty, and (iii) filters outputs via an automated reward model, producing large and reliable corpora for pre-execution safety. To close the guardian model gap, we propose a foundational guardrail Safiron, combining a cross-planner adapter with a compact guardian model. The adapter unifies different input formats, while Safiron flags risky cases, assigns risk types, and generates rationales; trained in two stages with a broadly explored data recipe, Safiron achieves robust transfer across settings. To close the evaluation gap, we release Pre-Exec Bench, a realistic benchmark covering diverse tools and branching trajectories, which measures detection, fine-grained categorization, explanation, and cross-planner generalization in human-verified scenarios. Extensive experiments demonstrate consistent gains of the proposed guardrail over strong baselines on Pre-Exec Bench, and ablations further distill actionable practices, providing a practical template for safer agentic systems.</p> + https://arxiv.org/abs/2510.09781 + Fri, 10 Oct 2025 18:42:32 +0000 - Low-probability Tokens Sustain Exploration in Reinforcement Learning with Verifiable Reward - https://arxiv.org/abs/2510.03222 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03222.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Guanhua Huang, Tingqiang Xu, Mingze Wang, Qi Yi, Xue Gong, Siheng Li, Ruibin Xiong, Kejiao Li, Yuhao Jiang, Bo Zhou</p><p><b>Upvotes:</b> 37</p><p><b>Summary:</b> Reinforcement Learning with Verifiable Rewards (RLVR) has propelled Large Language Models in complex reasoning, yet its scalability is often hindered by a training bottleneck where performance plateaus as policy entropy collapses, signaling a loss of exploration. Previous methods typically address this by maintaining high policy entropy, yet the precise mechanisms that govern meaningful exploration have remained underexplored. Our analysis suggests that an unselective focus on entropy risks amplifying irrelevant tokens and destabilizing training. This paper investigates the exploration dynamics within RLVR and identifies a key issue: the gradual elimination of valuable low-probability exploratory tokens, which we term \textit{reasoning sparks}. We find that while abundant in pre-trained models, these sparks are systematically extinguished during RLVR due to over-penalization, leading to a degeneracy in exploration. To address this, we introduce Low-probability Regularization (Lp-Reg). Its core mechanism regularizes the policy towards a heuristic proxy distribution. This proxy is constructed by filtering out presumed noise tokens and re-normalizing the distribution over the remaining candidates. The result is a less-noisy proxy where the probability of reasoning sparks is amplified, which then serves as a soft regularization target to shield these valuable tokens from elimination via KL divergence. Experiments show that Lp-Reg enables stable on-policy training for around 1,000 steps, a regime where baseline entropy-control methods collapse. This sustained exploration leads to state-of-the-art performance, achieving a 60.17% average accuracy on five math benchmarks, an improvement of 2.66% over prior methods. Code is available at https://github.com/CarlanLark/Lp-Reg.</p> - https://arxiv.org/abs/2510.03222 - Fri, 03 Oct 2025 17:56:13 +0000 + SwarmSys: Decentralized Swarm-Inspired Agents for Scalable and Adaptive Reasoning + https://arxiv.org/abs/2510.10047 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10047.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ruohao Li, Hongjun Liu, Leyi Zhao, Zisu Li, Jiawei Li, Jiajun Jiang, Linning Xu, Chen Zhao, Mingming Fan, Chen Liang</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> Large language model (LLM) agents have shown remarkable reasoning abilities. However, existing multi-agent frameworks often rely on fixed roles or centralized control, limiting scalability and adaptability in long-horizon reasoning. We introduce SwarmSys, a closed-loop framework for distributed multi-agent reasoning inspired by swarm intelligence. Coordination in SwarmSys emerges through iterative interactions among three specialized roles, Explorers, Workers, and Validators, that continuously cycle through exploration, exploitation, and validation. To enable scalable and adaptive collaboration, we integrate adaptive agent and event profiles, embedding-based probabilistic matching, and a pheromone-inspired reinforcement mechanism, supporting dynamic task allocation and self-organizing convergence without global supervision. Across symbolic reasoning, research synthesis, and scientific programming tasks, SwarmSys consistently outperforms baselines, improving both accuracy and reasoning stability. These findings highlight swarm-inspired coordination as a promising paradigm for scalable, robust, and adaptive multi-agent reasoning, suggesting that coordination scaling may rival model scaling in advancing LLM intelligence.</p> + https://arxiv.org/abs/2510.10047 + Sat, 11 Oct 2025 06:28:22 +0000 - OneFlow: Concurrent Mixed-Modal and Interleaved Generation with Edit Flows - https://arxiv.org/abs/2510.03506 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03506.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> John Nguyen, Marton Havasi, Tariq Berrada, Luke Zettlemoyer, Ricky T. Q. Chen</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> We present OneFlow, the first non-autoregressive multimodal model that enables variable-length and concurrent mixed-modal generation. Unlike autoregressive models that enforce rigid causal ordering between text and image generation, OneFlow combines an insertion-based Edit Flow for discrete text tokens with Flow Matching for image latents. OneFlow enables concurrent text-image synthesis with hierarchical sampling that prioritizes content over grammar. Through controlled experiments across model sizes from 1B to 8B, we demonstrate that OneFlow outperforms autoregressive baselines on both generation and understanding tasks while using up to 50% fewer training FLOPs. OneFlow surpasses both autoregressive and diffusion-based approaches while unlocking new capabilities for concurrent generation, iterative refinement, and natural reasoning-like generation.</p> - https://arxiv.org/abs/2510.03506 - Fri, 03 Oct 2025 20:40:30 +0000 + Don't Just Fine-tune the Agent, Tune the Environment + https://arxiv.org/abs/2510.10197 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10197.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Siyuan Lu, Zechuan Wang, Hongxuan Zhang, Qintong Wu, Leilei Gan, Chenyi Zhuang, Jinjie Gu, Tao Lin</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> Large Language Model (LLM) agents show great promise for complex, multi-turn tool-use tasks, but their development is often hampered by the extreme scarcity of high-quality training data. Supervised fine-tuning (SFT) on synthetic data leads to overfitting, whereas standard reinforcement learning (RL) struggles with a critical cold-start problem and training instability. To address these challenges, we introduce Environment Tuning, a novel training paradigm that enables agents to learn complex behaviors directly from problem instances without relying on pre-collected expert trajectories. Environment Tuning orchestrates this learning process through a structured curriculum, actionable environment augmentation that provides corrective feedback, and fine-grained progress rewards to ensure stable and efficient exploration. Using only 400 problem instances from Berkeley Function-Calling Leaderboard (BFCL) benchmark, our method not only achieves competitive in-distribution performance against strong baselines but also demonstrates superior out-of-distribution generalization, overcoming the performance collapse common to SFT-based approaches. Our work presents a paradigm shift from supervised fine-tuning on static trajectories to dynamic, environment-based exploration, paving the way for training more robust and data-efficient agents.</p> + https://arxiv.org/abs/2510.10197 + Sat, 11 Oct 2025 12:35:15 +0000 - Fine-Tuning on Noisy Instructions: Effects on Generalization and Performance - https://arxiv.org/abs/2510.03528 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03528.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ahmed Alajrami, Xingwei Tan, Nikolaos Aletras</p><p><b>Upvotes:</b> 15</p><p><b>Summary:</b> Instruction-tuning plays a vital role in enhancing the task-solving abilities of large language models (LLMs), improving their usability in generating helpful responses on various tasks. However, previous work has demonstrated that they are sensitive to minor variations in instruction phrasing. In this paper, we explore whether introducing perturbations in instruction-tuning data can enhance LLMs' resistance against noisy instructions. We focus on how instruction-tuning with perturbations, such as removing stop words or shuffling words, affects LLMs' performance on the original and perturbed versions of widely-used benchmarks (MMLU, BBH, GSM8K). We further assess learning dynamics and potential shifts in model behavior. Surprisingly, our results suggest that instruction-tuning on perturbed instructions can, in some cases, improve downstream performance. These findings highlight the importance of including perturbed instructions in instruction-tuning, which can make LLMs more resilient to noisy user inputs.</p> - https://arxiv.org/abs/2510.03528 - Fri, 03 Oct 2025 21:54:33 +0000 + RLFR: Extending Reinforcement Learning for LLMs with Flow Environment + https://arxiv.org/abs/2510.10201 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10201.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jinghao Zhang, Naishan Zheng, Ruilin Li, Dongzhou Cheng, Zheming Liang, Feng Zhao, Jiaqi Wang</p><p><b>Upvotes:</b> 35</p><p><b>Summary:</b> Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as a promising framework for improving reasoning abilities in Large Language Models (LLMs). However, policy optimized with binary verification prone to overlook potential valuable exploration in reasoning trajectory. In view of heavy annotation cost of golden Process Reward Models (PRMs), recent works attempt using auxiliary signals for reward shaping of process tokens, involving entropy and likelihood collected from logit space. In this work, we offer a novel perspective on shaping RLVR with flow rewards derived from latent space, and propose RLFR, where the flow fields of model latents are constructed from either off-policy high-quality data and on-policy rejection sampling data, and the velocity deviations of policy latents within it are quantified to serve as a reward signal. RLFR first demonstrates that a well-established flow field can be a sound environment for reward signal collection, highlighting the expressive latent space is much underexplored. Moreover, RLFR is able to compress any off-policy expert data as reference for constituting reward signals, and we show that the efficient context dependence compressed within the hidden states are utilized, rather than individual token-level denotation for context comprehending. Experiments on both language and multimodal reasoning benchmarks demonstrate the reliability of flow rewards, and suggesting a promising paradigm for reward shaping with auxiliary signals.</p> + https://arxiv.org/abs/2510.10201 + Sat, 11 Oct 2025 13:00:25 +0000 - Reactive Transformer (RxT) -- Stateful Real-Time Processing for Event-Driven Reactive Language Models - https://arxiv.org/abs/2510.03561 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03561.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Adam Filipek</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> The Transformer architecture has become the de facto standard for Large Language Models (LLMs), demonstrating remarkable capabilities in language understanding and generation. However, its application in conversational AI is fundamentally constrained by its stateless nature and the quadratic computational complexity (O(L^2)) with respect to sequence length L. Current models emulate memory by reprocessing an ever-expanding conversation history with each turn, leading to prohibitive costs and latency in long dialogues. This paper introduces the Reactive Transformer (RxT), a novel architecture designed to overcome these limitations by shifting from a data-driven to an event-driven paradigm. RxT processes each conversational turn as a discrete event in real-time, maintaining context in an integrated, fixed-size Short-Term Memory (STM) system. The architecture features a distinct operational cycle where a generator-decoder produces a response based on the current query and the previous memory state, after which a memory-encoder and a dedicated Memory Attention network asynchronously update the STM with a representation of the complete interaction. This design fundamentally alters the scaling dynamics, reducing the total user-facing cost of a conversation from quadratic (O(N^2 cdot T)) to linear (O(N cdot T)) with respect to the number of interactions N. By decoupling response generation from memory updates, RxT achieves low latency, enabling truly real-time, stateful, and economically viable long-form conversations. We validated our architecture with a series of proof-of-concept experiments on synthetic data, demonstrating superior performance and constant-time inference latency compared to a baseline stateless model of comparable size.</p> - https://arxiv.org/abs/2510.03561 - Fri, 03 Oct 2025 23:18:07 +0000 + X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model + https://arxiv.org/abs/2510.10274 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10274.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jinliang Zheng, Jianxiong Li, Zhihao Wang, Dongxiu Liu, Xirui Kang, Yuchun Feng, Yinan Zheng, Jiayin Zou, Yilun Chen, Jia Zeng, Ya-Qin Zhang, Jiangmiao Pang, Jingjing Liu, Tai Wang, Xianyuan Zhan</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> Successful generalist Vision-Language-Action (VLA) models rely on effective training across diverse robotic platforms with large-scale, cross-embodiment, heterogeneous datasets. To facilitate and leverage the heterogeneity in rich, diverse robotic data sources, we propose a novel Soft Prompt approach with minimally added parameters, by infusing prompt learning concepts into cross-embodiment robot learning and introducing separate sets of learnable embeddings for each distinct data source. These embeddings serve as embodiment-specific prompts, which in unity empower VLA models with effective exploitation of varying cross-embodiment features. Our new X-VLA, a neat flow-matching-based VLA architecture, relies exclusively on soft-prompted standard Transformer encoders, enjoying both scalability and simplicity. Evaluated across 6 simulations as well as 3 real-world robots, our 0.9B instantiation-X-VLA-0.9B simultaneously achieves SOTA performance over a sweep of benchmarks, demonstrating superior results on a wide axes of capabilities, from flexible dexterity to quick adaptation across embodiments, environments, and tasks. Website: https://thu-air-dream.github.io/X-VLA/</p> + https://arxiv.org/abs/2510.10274 + Sat, 11 Oct 2025 16:20:17 +0000 - Efficient Test-Time Scaling for Small Vision-Language Models - https://arxiv.org/abs/2510.03574 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03574.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Mehmet Onurcan Kaya, Desmond Elliott, Dim P. Papadopoulos</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Small Vision-Language Models (VLMs) provide a computationally efficient alternative to larger models, at the cost of weaker generalization abilities and downstream task performance. These shortcomings could be addressed by test-time scaling techniques, but existing methods are typically computationally demanding, contradicting the resource-efficient design goals of small models. To address these limitations, we propose two novel and efficient test-time scaling strategies that leverage the model-internal features rather than external supervision: (i) Test-Time Augmentation (TTAug), which generates multiple augmented inputs and aggregates outputs at the token level without parameter updates, and (ii) Test-Time Adaptation (TTAdapt), which adapts model parameters during inference using consensus-based pseudolabels from TTAug. Through extensive experiments across nine benchmarks, we demonstrate consistent performance improvements while maintaining computational efficiency suitable for resource-constrained environments. The generality of our approach is demonstrated both within models at different scales and across different VLMs without additional tuning.</p> - https://arxiv.org/abs/2510.03574 - Fri, 03 Oct 2025 23:49:06 +0000 + AVoCaDO: An Audiovisual Video Captioner Driven by Temporal Orchestration + https://arxiv.org/abs/2510.10395 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10395.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xinlong Chen, Yue Ding, Weihong Lin, Jingyun Hua, Linli Yao, Yang Shi, Bozhou Li, Yuanxing Zhang, Qiang Liu, Pengfei Wan, Liang Wang, Tieniu Tan</p><p><b>Upvotes:</b> 28</p><p><b>Summary:</b> Audiovisual video captioning aims to generate semantically rich descriptions with temporal alignment between visual and auditory events, thereby benefiting both video understanding and generation. In this paper, we present AVoCaDO, a powerful audiovisual video captioner driven by the temporal orchestration between audio and visual modalities. We propose a two-stage post-training pipeline: (1) AVoCaDO SFT, which fine-tunes the model on a newly curated dataset of 107K high-quality, temporally-aligned audiovisual captions; and (2) AVoCaDO GRPO, which leverages tailored reward functions to further enhance temporal coherence and dialogue accuracy while regularizing caption length and reducing collapse. Experimental results demonstrate that AVoCaDO significantly outperforms existing open-source models across four audiovisual video captioning benchmarks, and also achieves competitive performance on the VDC and DREAM-1K benchmark under visual-only settings.</p> + https://arxiv.org/abs/2510.10395 + Sun, 12 Oct 2025 01:20:22 +0000 - MITS: Enhanced Tree Search Reasoning for LLMs via Pointwise Mutual Information - https://arxiv.org/abs/2510.03632 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03632.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiaxi Li, Yucheng Shi, Jin Lu, Ninghao Liu</p><p><b>Upvotes:</b> 38</p><p><b>Summary:</b> Tree search has become as a representative framework for test-time reasoning with large language models (LLMs), exemplified by methods such as Tree-of-Thought and Monte Carlo Tree Search that explore multiple reasoning paths. However, it remains difficult to provide instant and reliable quantitative assessments of intermediate reasoning step quality, and extensive path exploration is computationally costly. To address this, we propose Mutual Information Tree Search (MITS), a novel framework that guides reasoning with information-theoretic principles. MITS introduces an effective scoring function based on pointwise mutual information (PMI), which enables step-wise evaluation of reasoning paths and search tree expansion via beam search without expensive look-ahead simulations, achieving superior reasoning performances while maintaining computational efficiency. The framework is complemented by an entropy-based dynamic sampling strategy that adaptively allocates computational resources to uncertain reasoning steps where exploration is most beneficial. For final prediction, MITS employs a weighted voting scheme that combines PMI scores with prediction consensus. Through comprehensive experiments on diverse reasoning benchmarks, MITS consistently surpasses baseline methods, establishing a principled and efficient framework for LLM reasoning.</p> - https://arxiv.org/abs/2510.03632 - Sat, 04 Oct 2025 02:30:40 +0000 + VR-Thinker: Boosting Video Reward Models through Thinking-with-Image Reasoning + https://arxiv.org/abs/2510.10518 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10518.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qunzhong Wang, Jie Liu, Jiajun Liang, Yilei Jiang, Yuanxing Zhang, Jinyuan Chen, Yaozhi Zheng, Xintao Wang, Pengfei Wan, Xiangyu Yue, Jiaheng Liu</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Recent advancements in multimodal reward models (RMs) have substantially improved post-training for visual generative models. However, current RMs face inherent limitations: (1) visual inputs consume large context budgets, forcing fewer frames and causing loss of fine-grained details; and (2) all visual information is packed into the initial prompt, exacerbating hallucination and forgetting during chain-of-thought reasoning. To overcome these issues, we introduce VideoReward Thinker (VR-Thinker), a thinking-with-image framework that equips the RM with visual reasoning operations (e.g., select frame) and a configurable visual memory window. This allows the RM to actively acquire and update visual evidence within context limits, improving reasoning fidelity and reliability. We activate visual reasoning via a reinforcement fine-tuning pipeline: (i) Cold Start with curated visual chain-of-thought data to distill basic reasoning skills and operation formatting; (ii) select samples whose per-dimension and overall judgments are all correct, then conduct Rejection sampling Fine-Tuning on these high-quality traces to further enhance reasoning; and (iii) apply Group Relative Policy Optimization (GRPO) to strengthen reasoning. Our approach delivers state-of-the-art accuracy among open-source models on video preference benchmarks, especially for longer videos: a 7B VR-Thinker achieves 80.5% on VideoGen Reward, 82.3% on GenAI-Bench, and 75.6% on MJ-Bench-Video. These results validate the effectiveness and promise of thinking-with-image multimodal reward modeling.</p> + https://arxiv.org/abs/2510.10518 + Sun, 12 Oct 2025 09:29:50 +0000 - UNIDOC-BENCH: A Unified Benchmark for Document-Centric Multimodal RAG - https://arxiv.org/abs/2510.03663 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03663.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xiangyu Peng, Cab Qin, Zeyuan Chen, Ran Xu, Caiming Xiong, Chien-Sheng Wu</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Multimodal retrieval-augmented generation (MM-RAG) is a key approach for applying large language models (LLMs) and agents to real-world knowledge bases, yet current evaluations are fragmented, focusing on either text or images in isolation or on simplified multimodal setups that fail to capture document-centric multimodal use cases. In this paper, we introduce UniDoc-Bench, the first large-scale, realistic benchmark for MM-RAG built from 70k real-world PDF pages across eight domains. Our pipeline extracts and links evidence from text, tables, and figures, then generates 1,600 multimodal QA pairs spanning factual retrieval, comparison, summarization, and logical reasoning queries. To ensure reliability, 20% of QA pairs are validated by multiple annotators and expert adjudication. UniDoc-Bench supports apples-to-apples comparison across four paradigms: (1) text-only, (2) image-only, (3) multimodal text-image fusion, and (4) multimodal joint retrieval -- under a unified protocol with standardized candidate pools, prompts, and evaluation metrics. Our experiments show that multimodal text-image fusion RAG systems consistently outperform both unimodal and jointly multimodal embedding-based retrieval, indicating that neither text nor images alone are sufficient and that current multimodal embeddings remain inadequate. Beyond benchmarking, our analysis reveals when and how visual context complements textual evidence, uncovers systematic failure modes, and offers actionable guidance for developing more robust MM-RAG pipelines.</p> - https://arxiv.org/abs/2510.03663 - Sat, 04 Oct 2025 04:30:13 +0000 + High-Fidelity Simulated Data Generation for Real-World Zero-Shot Robotic Manipulation Learning with Gaussian Splatting + https://arxiv.org/abs/2510.10637 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10637.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haoyu Zhao, Cheng Zeng, Linghao Zhuang, Yaxi Zhao, Shengke Xue, Hao Wang, Xingyue Zhao, Zhongyu Li, Kehan Li, Siteng Huang, Mingxiu Chen, Xin Li, Deli Zhao, Hua Zou</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> The scalability of robotic learning is fundamentally bottlenecked by the significant cost and labor of real-world data collection. While simulated data offers a scalable alternative, it often fails to generalize to the real world due to significant gaps in visual appearance, physical properties, and object interactions. To address this, we propose RoboSimGS, a novel Real2Sim2Real framework that converts multi-view real-world images into scalable, high-fidelity, and physically interactive simulation environments for robotic manipulation. Our approach reconstructs scenes using a hybrid representation: 3D Gaussian Splatting (3DGS) captures the photorealistic appearance of the environment, while mesh primitives for interactive objects ensure accurate physics simulation. Crucially, we pioneer the use of a Multi-modal Large Language Model (MLLM) to automate the creation of physically plausible, articulated assets. The MLLM analyzes visual data to infer not only physical properties (e.g., density, stiffness) but also complex kinematic structures (e.g., hinges, sliding rails) of objects. We demonstrate that policies trained entirely on data generated by RoboSimGS achieve successful zero-shot sim-to-real transfer across a diverse set of real-world manipulation tasks. Furthermore, data from RoboSimGS significantly enhances the performance and generalization capabilities of SOTA methods. Our results validate RoboSimGS as a powerful and scalable solution for bridging the sim-to-real gap.</p> + https://arxiv.org/abs/2510.10637 + Sun, 12 Oct 2025 14:42:07 +0000 - Optimal Scaling Needs Optimal Norm - https://arxiv.org/abs/2510.03871 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.03871.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Oleg Filatov, Jiangtao Wang, Jan Ebert, Stefan Kesselheim</p><p><b>Upvotes:</b> 27</p><p><b>Summary:</b> Despite recent progress in optimal hyperparameter transfer under model and dataset scaling, no unifying explanatory principle has been established. Using the Scion optimizer, we discover that joint optimal scaling across model and dataset sizes is governed by a single invariant: the operator norm of the output layer. Across models with up to 1.3B parameters trained on up to 138B tokens, the optimal learning rate/batch size pair (eta^{ast}, B^{ast}) consistently has the same operator norm value - a phenomenon we term norm transfer. This constant norm condition is necessary but not sufficient: while for each dataset size, multiple (eta, B) reach the optimal norm, only a unique (eta^{ast}, B^{ast}) achieves the best loss. As a sufficient condition, we provide the first measurement of (eta^{ast}, B^{ast}) scaling with dataset size for Scion, and find that the scaling rules are consistent with those of the Adam optimizer. Tuning per-layer-group learning rates also improves model performance, with the output layer being the most sensitive and hidden layers benefiting from lower learning rates. We provide practical insights on norm-guided optimal scaling and release our Distributed Scion (Disco) implementation with logs from over two thousand runs to support research on LLM training dynamics at scale.</p> - https://arxiv.org/abs/2510.03871 - Sat, 04 Oct 2025 16:48:36 +0000 + BrowserAgent: Building Web Agents with Human-Inspired Web Browsing Actions + https://arxiv.org/abs/2510.10666 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10666.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhengbo Zhang, Zhiheng Lyu, Junhao Gong, Hongzhu Yi, Xinming Wang, Yuxuan Zhou, Jiabing Yang, Ping Nie, Yan Huang, Wenhu Chen</p><p><b>Upvotes:</b> 27</p><p><b>Summary:</b> Efficiently solving real-world problems with LLMs increasingly hinges on their ability to interact with dynamic web environments and autonomously acquire external information. While recent research like Search-R1 and WebDancer demonstrates strong performance in solving web tasks, they heavily rely on additional tools to convert the interactive web environment into static text content. This is in contrast to human browsing behaviors, which involve diverse interactions with the browser, such as scrolling, clicking, and typing. In this paper, we propose BrowserAgent, a more interactive agent that solves complex tasks through human-inspired browser actions. BrowserAgent operates directly on raw web pages via Playwright through a set of predefined browser actions. We adopt a two-stage training (Supervised Fine-Tuning (SFT) and Rejection Fine-Tuning (RFT)) to improve the model's generalization abilities. Despite using significantly less training data than Search-R1, BrowserAgent achieves more competitive results across different Open-QA tasks. Additionally, we introduce an explicit memory mechanism to store key conclusions across steps, further enhancing the model's reasoning capabilities for long-horizon tasks. Notably, BrowserAgent-7B can achieve around 20\% improvement over Search-R1 on multi-hop QA tasks like HotpotQA, 2Wiki, and Bamboogle. These results indicate that BrowserAgent can serve as a more advanced framework for more interactive and scalable web agents.</p> + https://arxiv.org/abs/2510.10666 + Sun, 12 Oct 2025 15:43:37 +0000 - Scaling Code-Assisted Chain-of-Thoughts and Instructions for Model Reasoning - https://arxiv.org/abs/2510.04081 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04081.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Honglin Lin, Qizhi Pei, Xin Gao, Zhuoshi Pan, Yu Li, Juntao Li, Conghui He, Lijun Wu</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> Reasoning capability is pivotal for Large Language Models (LLMs) to solve complex tasks, yet achieving reliable and scalable reasoning remains challenging. While Chain-of-Thought (CoT) prompting has become a mainstream approach, existing methods often suffer from uncontrolled generation, insufficient quality, and limited diversity in reasoning paths. Recent efforts leverage code to enhance CoT by grounding reasoning in executable steps, but such methods are typically constrained to predefined mathematical problems, hindering scalability and generalizability. In this work, we propose Caco (Code-Assisted Chain-of-ThOught), a novel framework that automates the synthesis of high-quality, verifiable, and diverse instruction-CoT reasoning data through code-driven augmentation. Unlike prior work, Caco first fine-tunes a code-based CoT generator on existing math and programming solutions in a unified code format, then scales the data generation to a large amount of diverse reasoning traces. Crucially, we introduce automated validation via code execution and rule-based filtering to ensure logical correctness and structural diversity, followed by reverse-engineering filtered outputs into natural language instructions and language CoTs to enrich task adaptability. This closed-loop process enables fully automated, scalable synthesis of reasoning data with guaranteed executability. Experiments on our created Caco-1.3M dataset demonstrate that Caco-trained models achieve strong competitive performance on mathematical reasoning benchmarks, outperforming existing strong baselines. Further analysis reveals that Caco's code-anchored verification and instruction diversity contribute to superior generalization across unseen tasks. Our work establishes a paradigm for building self-sustaining, trustworthy reasoning systems without human intervention.</p> - https://arxiv.org/abs/2510.04081 - Sun, 05 Oct 2025 07:59:24 +0000 + AdaViewPlanner: Adapting Video Diffusion Models for Viewpoint Planning in 4D Scenes + https://arxiv.org/abs/2510.10670 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10670.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yu Li, Menghan Xia, Gongye Liu, Jianhong Bai, Xintao Wang, Conglang Zhang, Yuxuan Lin, Ruihang Chu, Pengfei Wan, Yujiu Yang</p><p><b>Upvotes:</b> 16</p><p><b>Summary:</b> Recent Text-to-Video (T2V) models have demonstrated powerful capability in visual simulation of real-world geometry and physical laws, indicating its potential as implicit world models. Inspired by this, we explore the feasibility of leveraging the video generation prior for viewpoint planning from given 4D scenes, since videos internally accompany dynamic scenes with natural viewpoints. To this end, we propose a two-stage paradigm to adapt pre-trained T2V models for viewpoint prediction, in a compatible manner. First, we inject the 4D scene representation into the pre-trained T2V model via an adaptive learning branch, where the 4D scene is viewpoint-agnostic and the conditional generated video embeds the viewpoints visually. Then, we formulate viewpoint extraction as a hybrid-condition guided camera extrinsic denoising process. Specifically, a camera extrinsic diffusion branch is further introduced onto the pre-trained T2V model, by taking the generated video and 4D scene as input. Experimental results show the superiority of our proposed method over existing competitors, and ablation studies validate the effectiveness of our key technical designs. To some extent, this work proves the potential of video generation models toward 4D interaction in real world.</p> + https://arxiv.org/abs/2510.10670 + Sun, 12 Oct 2025 15:55:44 +0000 - Drax: Speech Recognition with Discrete Flow Matching - https://arxiv.org/abs/2510.04162 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04162.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Aviv Navon, Aviv Shamsian, Neta Glazer, Yael Segal-Feldman, Gill Hetz, Joseph Keshet, Ethan Fetaya</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Diffusion and flow-based non-autoregressive (NAR) models have shown strong promise in large language modeling, however, their potential for automatic speech recognition (ASR) remains largely unexplored. We propose Drax, a discrete flow matching framework for ASR that enables efficient parallel decoding. To better align training with inference, we construct an audio-conditioned probability path that guides the model through trajectories resembling likely intermediate inference errors, rather than direct random noise to target transitions. Our theoretical analysis links the generalization gap to divergences between training and inference occupancies, controlled by cumulative velocity errors, thereby motivating our design choice. Empirical evaluation demonstrates that our approach attains recognition accuracy on par with state-of-the-art speech models while offering improved accuracy-efficiency trade-offs, highlighting discrete flow matching as a promising direction for advancing NAR ASR.</p> - https://arxiv.org/abs/2510.04162 - Sun, 05 Oct 2025 11:38:01 +0000 + OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs + https://arxiv.org/abs/2510.10689 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10689.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Caorui Li, Yu Chen, Yiyan Ji, Jin Xu, Zhenyu Cui, Shihao Li, Yuanxing Zhang, Jiafu Tang, Zhenghao Song, Dingling Zhang, Ying He, Haoxiang Liu, Yuxuan Wang, Qiufeng Wang, Zhenhe Wu, Jiehui Luo, Zhiyu Pan, Weihao Xie, Chenchen Zhang, Zhaohui Wang, Jiayi Tian, Yanghai Wang, Zhe Cao, Minxin Dai, Ke Wang, Runzhe Wen, Yinghao Ma, Yaning Pan, Sungkyun Chang, Termeh Taheri, Haiwen Xia, Christos Plachouras, Emmanouil Benetos, Yizhi Li, Ge Zhang, Jian Yang, Tianhao Peng, Zili Wang, Minghao Liu, Junran Peng, Zhaoxiang Zhang, Jiaheng Liu</p><p><b>Upvotes:</b> 44</p><p><b>Summary:</b> Recent advances in multimodal large language models (MLLMs) have demonstrated substantial potential in video understanding. However, existing benchmarks fail to comprehensively evaluate synergistic reasoning capabilities across audio and visual modalities, often neglecting either one of the modalities or integrating them in a logically inconsistent manner. To bridge this gap, we introduce OmniVideoBench, a large-scale and rigorously designed benchmark dedicated to assessing synergistic audio-visual understanding, with a strong emphasis on modality complementarity and logical consistency. Specifically, OmniVideoBench comprises 1000 high-quality question-answer(QA) pairs, each annotated with step-by-step reasoning traces, derived from 628 diverse videos ranging from several seconds to 30 minutes, and manually verified to guarantee complete correctness and uniqueness. Moreover, OmniVideoBench encompasses 13 carefully designed question types, covering temporal reasoning, spatial localization, counting, causal inference, summarization, and beyond, thereby capturing the essential challenges of video understanding. Evaluation of multiple MLLMs on OmniVideoBench reveals a pronounced gap between model performance and human reasoning, with open-source models lagging significantly behind their closed-source counterparts, underscoring the inherent difficulty of genuine audio-visual reasoning. We will release OmniVideoBench to foster the development of MLLMs with stronger and more generalizable reasoning capabilities.</p> + https://arxiv.org/abs/2510.10689 + Sun, 12 Oct 2025 16:34:00 +0000 - CALM Before the STORM: Unlocking Native Reasoning for Optimization Modeling - https://arxiv.org/abs/2510.04204 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04204.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhengyang Tang, Zihan Ye, Chenyu Huang, Xuhan Huang, Chengpeng Li, Sihang Li, Guanhua Chen, Ming Yan, Zizhuo Wang, Hongyuan Zha, Dayiheng Liu, Benyou Wang</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> Large Reasoning Models (LRMs) have demonstrated strong capabilities in complex multi-step reasoning, opening new opportunities for automating optimization modeling. However, existing domain adaptation methods, originally designed for earlier instruction-tuned models, often fail to exploit the advanced reasoning patterns of modern LRMs -- In particular, we show that direct fine-tuning on traditional non-reflective datasets leads to limited gains. To fully leverage LRMs' inherent reasoning abilities, we propose CALM (Corrective Adaptation with Lightweight Modification), a framework that progressively refines LRMs within their native reasoning modes for optimization modeling tasks. In CALM, an expert intervener identifies reasoning flaws and provides concise corrective hints, which the LRM incorporates to produce improved reasoning trajectories. These interventions modify fewer than 2.6\% of generated tokens, but generate high-quality data for soft adaptation through supervised fine-tuning. The adapted model is then further improved through reinforcement learning. Building on CALM, we develop STORM (Smart Thinking Optimization Reasoning Model), a 4B-parameter LRM that achieves a new state-of-the-art average accuracy of 68.9\% across five popular optimization modeling benchmarks, matching the performance of a 671B LRM. These results demonstrate that dynamic, hint-based data synthesis both preserves and amplifies the native reasoning patterns of modern LRMs, offering a more effective and scalable path towards expert-level performance on challenging optimization modeling tasks.</p> - https://arxiv.org/abs/2510.04204 - Sun, 05 Oct 2025 13:38:31 +0000 + FastHMR: Accelerating Human Mesh Recovery via Token and Layer Merging with Diffusion Decoding + https://arxiv.org/abs/2510.10868 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.10868.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Soroush Mehraban, Andrea Iaboni, Babak Taati</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Recent transformer-based models for 3D Human Mesh Recovery (HMR) have achieved strong performance but often suffer from high computational cost and complexity due to deep transformer architectures and redundant tokens. In this paper, we introduce two HMR-specific merging strategies: Error-Constrained Layer Merging (ECLM) and Mask-guided Token Merging (Mask-ToMe). ECLM selectively merges transformer layers that have minimal impact on the Mean Per Joint Position Error (MPJPE), while Mask-ToMe focuses on merging background tokens that contribute little to the final prediction. To further address the potential performance drop caused by merging, we propose a diffusion-based decoder that incorporates temporal context and leverages pose priors learned from large-scale motion capture datasets. Experiments across multiple benchmarks demonstrate that our method achieves up to 2.3x speed-up while slightly improving performance over the baseline.</p> + https://arxiv.org/abs/2510.10868 + Mon, 13 Oct 2025 00:23:17 +0000 - Why Low-Precision Transformer Training Fails: An Analysis on Flash Attention - https://arxiv.org/abs/2510.04212 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04212.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haiquan Qiu, Quanming Yao</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> The pursuit of computational efficiency has driven the adoption of low-precision formats for training transformer models. However, this progress is often hindered by notorious training instabilities. This paper provides the first mechanistic explanation for a long-standing and unresolved failure case where training with flash attention in low-precision settings leads to catastrophic loss explosions. Our in-depth analysis reveals that the failure is not a random artifact but caused by two intertwined phenomena: the emergence of similar low-rank representations within the attention mechanism and the compounding effect of biased rounding errors inherent in low-precision arithmetic. We demonstrate how these factors create a vicious cycle of error accumulation that corrupts weight updates, ultimately derailing the training dynamics. To validate our findings, we introduce a minimal modification to the flash attention that mitigates the bias in rounding errors. This simple change stabilizes the training process, confirming our analysis and offering a practical solution to this persistent problem.</p> - https://arxiv.org/abs/2510.04212 - Sun, 05 Oct 2025 14:01:24 +0000 + GIR-Bench: Versatile Benchmark for Generating Images with Reasoning + https://arxiv.org/abs/2510.11026 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11026.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hongxiang Li, Yaowei Li, Bin Lin, Yuwei Niu, Yuhang Yang, Xiaoshuang Huang, Jiayin Cai, Xiaolong Jiang, Yao Hu, Long Chen</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Unified multimodal models integrate the reasoning capacity of large language models with both image understanding and generation, showing great promise for advanced multimodal intelligence. However, the community still lacks a rigorous reasoning-centric benchmark to systematically evaluate the alignment between understanding and generation, and their generalization potential in complex visual tasks. To this end, we introduce GIR-Bench, a comprehensive benchmark that evaluates unified models across three complementary perspectives. Firstly, we investigate understanding-generation consistency (GIR-Bench-UGC), asking whether models can consistently leverage the same knowledge in both understanding and generation tasks. Secondly, we investigate whether models can perform reasoning-centric text-to-image generation that requires applying logical constraints and implicit knowledge to generate faithful visual content (GIR-Bench-T2I). Thirdly, we evaluate whether models can handle multi-step reasoning in editing (GIR-Bench-Edit). For each subset, we carefully design different task-specific evaluation pipelines tailored for each task. This enables fine-grained and interpretable evaluation while mitigating biases from the prevalent MLLM-as-a-Judge paradigm. Extensive ablations over various unified models and generation-only systems have shown that: Although unified models are more capable of reasoning-driven visual tasks, they still exhibit a persistent gap between understanding and generation. The data and code for GIR-Bench are available at https://hkust-longgroup.github.io/GIR-Bench{https://hkust-longgroup.github.io/GIR-Bench}.</p> + https://arxiv.org/abs/2510.11026 + Mon, 13 Oct 2025 05:50:44 +0000 - Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought - https://arxiv.org/abs/2510.04230 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04230.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Guijin Son, Donghun Yang, Hitesh Laxmichand Patel, Amit Agarwal, Hyunwoo Ko, Chanuk Lim, Srikant Panda, Minhyuk Kim, Nikunj Drolia, Dasol Choi, Kyong-Ha Lee, Youngjae Yu</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Recent frontier models employ long chain-of-thought reasoning to explore solution spaces in context and achieve stonger performance. While many works study distillation to build smaller yet capable models, most focus on English and little is known about language-specific reasoning. To bridge this gap, we first introduct **Language-Mixed CoT**, a reasoning schema that switches between English and a target language, using English as an anchor to excel in reasoning while minimizing translation artificats. As a Korean case study, we curate **Yi-Sang**: 5.79M native-Korean prompts from web Q&A, exams, STEM, and code; 3.7M long reasoning traces generated from Qwen3-32B; and a targeted 260k high-yield subset. We train ninve models (4B-35B) across six families (Qwen2.5, Llama-3.1, Gemma-3, etc). Our best model, **KO-REAson-35B**, achieves state-of-the-art performance, with the highest overall average score (64.0 \pm 25), ranking first on 5/9 benchmarks and second on the remainder. Samller and mid-sized models also benefit substantially, with an average improvement of +18.6 points across teh evaluated nine benchmarks. Ablations show **Language-Mixed CoT** is more effective than monolingual CoT, also resulting in cross-lingual and mult-modal performance gains. We release our data-curation pipeline, evaluation system, datasets, and models to advance research on language-specific reasoning. Data and model collection: https://huggingface.co/KOREAson.</p> - https://arxiv.org/abs/2510.04230 - Sun, 05 Oct 2025 14:39:41 +0000 + Vlaser: Vision-Language-Action Model with Synergistic Embodied Reasoning + https://arxiv.org/abs/2510.11027 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11027.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ganlin Yang, Tianyi Zhang, Haoran Hao, Weiyun Wang, Yibin Liu, Dehui Wang, Guanzhou Chen, Zijian Cai, Junting Chen, Weijie Su, Wengang Zhou, Yu Qiao, Jifeng Dai, Jiangmiao Pang, Gen Luo, Wenhai Wang, Yao Mu, Zhi Hou</p><p><b>Upvotes:</b> 18</p><p><b>Summary:</b> While significant research has focused on developing embodied reasoning capabilities using Vision-Language Models (VLMs) or integrating advanced VLMs into Vision-Language-Action (VLA) models for end-to-end robot control, few studies directly address the critical gap between upstream VLM-based reasoning and downstream VLA policy learning. In this work, we take an initial step toward bridging embodied reasoning with VLA policy learning by introducing Vlaser - a Vision-Language-Action Model with synergistic embodied reasoning capability, which is a foundational vision-language model designed to integrate high-level reasoning with low-level control for embodied agents. Built upon the high-quality Vlaser-6M dataset, Vlaser achieves state-of-the-art performance across a range of embodied reasoning benchmarks - including spatial reasoning, embodied grounding, embodied QA, and task planning. Furthermore, we systematically examine how different VLM initializations affect supervised VLA fine-tuning, offering novel insights into mitigating the domain shift between internet-scale pre-training data and embodied-specific policy learning data. Based on these insights, our approach achieves state-of-the-art results on the WidowX benchmark and competitive performance on the Google Robot benchmark.</p> + https://arxiv.org/abs/2510.11027 + Mon, 13 Oct 2025 05:51:22 +0000 - ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation - https://arxiv.org/abs/2510.04290 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04290.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jay Zhangjie Wu, Xuanchi Ren, Tianchang Shen, Tianshi Cao, Kai He, Yifan Lu, Ruiyuan Gao, Enze Xie, Shiyi Lan, Jose M. Alvarez, Jun Gao, Sanja Fidler, Zian Wang, Huan Ling</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Recent advances in large generative models have significantly advanced image editing and in-context image generation, yet a critical gap remains in ensuring physical consistency, where edited objects must remain coherent. This capability is especially vital for world simulation related tasks. In this paper, we present ChronoEdit, a framework that reframes image editing as a video generation problem. First, ChronoEdit treats the input and edited images as the first and last frames of a video, allowing it to leverage large pretrained video generative models that capture not only object appearance but also the implicit physics of motion and interaction through learned temporal consistency. Second, ChronoEdit introduces a temporal reasoning stage that explicitly performs editing at inference time. Under this setting, the target frame is jointly denoised with reasoning tokens to imagine a plausible editing trajectory that constrains the solution space to physically viable transformations. The reasoning tokens are then dropped after a few steps to avoid the high computational cost of rendering a full video. To validate ChronoEdit, we introduce PBench-Edit, a new benchmark of image-prompt pairs for contexts that require physical consistency, and demonstrate that ChronoEdit surpasses state-of-the-art baselines in both visual fidelity and physical plausibility. Code and models for both the 14B and 2B variants of ChronoEdit will be released on the project page: https://research.nvidia.com/labs/toronto-ai/chronoedit</p> - https://arxiv.org/abs/2510.04290 - Sun, 05 Oct 2025 17:02:01 +0000 + Latent Refinement Decoding: Enhancing Diffusion-Based Language Models by Refining Belief States + https://arxiv.org/abs/2510.11052 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11052.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qinglin Zhu, Yizhen Yao, Runcong Zhao, Yanzheng Xiang, Amrutha Saseendran, Chen Jin, Philip Alexander Teare, Bin Liang, Yulan He, Lin Gui</p><p><b>Upvotes:</b> 48</p><p><b>Summary:</b> Autoregressive (AR) models remain the standard for natural language generation but still suffer from high latency due to strictly sequential decoding. Recent diffusion-inspired approaches, such as LlaDA and Dream, mitigate this by generating in parallel, yet they suffer from two core limitations: information loss, as predictive distributions for non-finalized tokens are discarded at each step, and premature commitment, where local decisions are made without sufficient global coordination. We introduce Latent Refinement Decoding (LRD), a two-stage framework with Latent Refinement and a Predictive Feedback Loop. The first stage maintains masked positions as distributional mixtures of predicted tokens and the mask embedding, allowing the model to establish more globally consistent beliefs. The second stage progressively finalizes confident tokens while retaining uncertain ones for iterative feedback. KL-divergence dynamics provide a principled and reliable criterion for convergence and early stopping. Experiments across coding (HumanEval +6.3, MBPP +2.6) and reasoning (GSM8K +2.9, MATH500 +3.8) show that LRD improves accuracy while delivering speedups of up to 10.6x, making it a strong and versatile alternative for parallel sequence generation.</p> + https://arxiv.org/abs/2510.11052 + Mon, 13 Oct 2025 06:38:13 +0000 - GRACE: Generative Representation Learning via Contrastive Policy Optimization - https://arxiv.org/abs/2510.04506 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04506.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiashuo Sun, Shixuan Liu, Zhaochen Su, Xianrui Zhong, Pengcheng Jiang, Bowen Jin, Peiran Li, Weijia Shi, Jiawei Han</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> Prevailing methods for training Large Language Models (LLMs) as text encoders rely on contrastive losses that treat the model as a black box function, discarding its generative and reasoning capabilities in favor of static embeddings. We introduce GRACE (Generative Representation Learning via Contrastive Policy Optimization), a novel framework that reimagines contrastive signals not as losses to be minimized, but as rewards that guide a generative policy. In GRACE, the LLM acts as a policy that produces explicit, human-interpretable rationales--structured natural language explanations of its semantic understanding. These rationales are then encoded into high-quality embeddings via mean pooling. Using policy gradient optimization, we train the model with a multi-component reward function that maximizes similarity between query positive pairs and minimizes similarity with negatives. This transforms the LLM from an opaque encoder into an interpretable agent whose reasoning process is transparent and inspectable. On MTEB benchmark, GRACE yields broad cross category gains: averaged over four backbones, the supervised setting improves overall score by 11.5% over base models, and the unsupervised variant adds 6.9%, while preserving general capabilities. This work treats contrastive objectives as rewards over rationales, unifying representation learning with generation to produce stronger embeddings and transparent rationales. The model, data and code are available at https://github.com/GasolSun36/GRACE.</p> - https://arxiv.org/abs/2510.04506 - Mon, 06 Oct 2025 05:46:56 +0000 + Temporal Alignment Guidance: On-Manifold Sampling in Diffusion Models + https://arxiv.org/abs/2510.11057 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11057.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Youngrok Park, Hojung Jung, Sangmin Bae, Se-Young Yun</p><p><b>Upvotes:</b> 29</p><p><b>Summary:</b> Diffusion models have achieved remarkable success as generative models. However, even a well-trained model can accumulate errors throughout the generation process. These errors become particularly problematic when arbitrary guidance is applied to steer samples toward desired properties, which often breaks sample fidelity. In this paper, we propose a general solution to address the off-manifold phenomenon observed in diffusion models. Our approach leverages a time predictor to estimate deviations from the desired data manifold at each timestep, identifying that a larger time gap is associated with reduced generation quality. We then design a novel guidance mechanism, `Temporal Alignment Guidance' (TAG), attracting the samples back to the desired manifold at every timestep during generation. Through extensive experiments, we demonstrate that TAG consistently produces samples closely aligned with the desired manifold at each timestep, leading to significant improvements in generation quality across various downstream tasks.</p> + https://arxiv.org/abs/2510.11057 + Mon, 13 Oct 2025 06:46:57 +0000 - Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models - https://arxiv.org/abs/2510.04618 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04618.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qizheng Zhang, Changran Hu, Shubhangi Upasani, Boyuan Ma, Fenglu Hong, Vamsidhar Kamanuru, Jay Rainton, Chen Wu, Mengmeng Ji, Hanchen Li, Urmish Thakker, James Zou, Kunle Olukotun</p><p><b>Upvotes:</b> 63</p><p><b>Summary:</b> Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on context adaptation -- modifying inputs with instructions, strategies, or evidence, rather than weight updates. Prior approaches improve usability but often suffer from brevity bias, which drops domain insights for concise summaries, and from context collapse, where iterative rewriting erodes details over time. Building on the adaptive memory introduced by Dynamic Cheatsheet, we introduce ACE (Agentic Context Engineering), a framework that treats contexts as evolving playbooks that accumulate, refine, and organize strategies through a modular process of generation, reflection, and curation. ACE prevents collapse with structured, incremental updates that preserve detailed knowledge and scale with long-context models. Across agent and domain-specific benchmarks, ACE optimizes contexts both offline (e.g., system prompts) and online (e.g., agent memory), consistently outperforming strong baselines: +10.6% on agents and +8.6% on finance, while significantly reducing adaptation latency and rollout cost. Notably, ACE could adapt effectively without labeled supervision and instead by leveraging natural execution feedback. On the AppWorld leaderboard, ACE matches the top-ranked production-level agent on the overall average and surpasses it on the harder test-challenge split, despite using a smaller open-source model. These results show that comprehensive, evolving contexts enable scalable, efficient, and self-improving LLM systems with low overhead.</p> - https://arxiv.org/abs/2510.04618 - Mon, 06 Oct 2025 09:30:18 +0000 + Stronger Together: On-Policy Reinforcement Learning for Collaborative LLMs + https://arxiv.org/abs/2510.11062 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11062.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yujie Zhao, Lanxiang Hu, Yang Wang, Minmin Hou, Hao Zhang, Ke Ding, Jishen Zhao</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Multi-agent systems (MAS) and reinforcement learning (RL) are widely used to enhance the agentic capabilities of large language models (LLMs). MAS improves task performance through role-based orchestration, while RL uses environmental rewards to learn stronger policies, such as GRPO-style optimization. However, applying on-policy RL to MAS remains underexplored and presents unique challenges. Algorithmically, standard GRPO grouping assumptions break down because prompts vary by role and by turn. System-wise, the training stack must support MAS-workflow rollouts and on-policy updates for both single-policy and multi-policy models. We propose AT-GRPO, which includes (i) an agent- and turn-wise grouped RL algorithm tailored to MAS and (ii) a training system that supports both single- and multi-policy regimes. Across game, planning, coding, and math tasks, AT-GRPO delivers substantial gains. On long-horizon planning, it increases accuracy from a 14.0 to 47.0 percent single-agent RL baseline to 96.0 to 99.5 percent. It also improves reasoning performance, with average gains of 3.87 to 7.62 percent on coding tasks and 9.0 to 17.93 percent on math. Code and environments are available at: https://github.com/pettingllms-ai/PettingLLMs.</p> + https://arxiv.org/abs/2510.11062 + Mon, 13 Oct 2025 06:55:09 +0000 - Watch and Learn: Learning to Use Computers from Online Videos - https://arxiv.org/abs/2510.04673 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04673.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chan Hee Song, Yiwen Song, Palash Goyal, Yu Su, Oriana Riva, Hamid Palangi, Tomas Pfister</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> Computer use agents (CUAs) need to plan task workflows grounded in diverse, ever-changing applications and environments, but learning is hindered by the scarcity of large-scale, high-quality training data in the target application. Existing datasets are domain-specific, static, and costly to annotate, while current synthetic data generation methods often yield simplistic or misaligned task demonstrations. To address these limitations, we introduce Watch & Learn (W&L), a framework that converts human demonstration videos readily available on the Internet into executable UI trajectories at scale. Instead of directly generating trajectories or relying on ad hoc reasoning heuristics, we cast the problem as an inverse dynamics objective: predicting the user's action from consecutive screen states. This formulation reduces manual engineering, is easier to learn, and generalizes more robustly across applications. Concretely, we develop an inverse dynamics labeling pipeline with task-aware video retrieval, generate over 53k high-quality trajectories from raw web videos, and demonstrate that these trajectories improve CUAs both as in-context demonstrations and as supervised training data. On the challenging OSWorld benchmark, UI trajectories extracted with W&L consistently enhance both general-purpose and state-of-the-art frameworks in-context, and deliver stronger gains for open-source models under supervised training. These results highlight web-scale human demonstration videos as a practical and scalable foundation for advancing CUAs towards real-world deployment.</p> - https://arxiv.org/abs/2510.04673 - Mon, 06 Oct 2025 10:29:00 +0000 + InternSVG: Towards Unified SVG Tasks with Multimodal Large Language Models + https://arxiv.org/abs/2510.11341 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11341.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haomin Wang, Jinhui Yin, Qi Wei, Wenguang Zeng, Lixin Gu, Shenglong Ye, Zhangwei Gao, Yaohui Wang, Yanting Zhang, Yuanqi Li, Yanwen Guo, Wenhai Wang, Kai Chen, Yu Qiao, Hongjie Zhang</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> General SVG modeling remains challenging due to fragmented datasets, limited transferability of methods across tasks, and the difficulty of handling structural complexity. In response, we leverage the strong transfer and generalization capabilities of multimodal large language models (MLLMs) to achieve unified modeling for SVG understanding, editing, and generation. We present the InternSVG family, an integrated data-benchmark-model suite. At its core is SAgoge, the largest and most comprehensive multimodal dataset for SVG tasks, encompassing both static graphics and dynamic animations. It covers icons, long-sequence illustrations, scientific diagrams, and dynamic animations, supporting tasks of varied difficulty levels and providing deeper hierarchies with richer attributes compared to previous datasets. Based on this resource, we introduce SArena, a companion benchmark with comprehensive task definitions and standardized evaluation that aligns with the domains and difficulty spectrum covered by SAgoge. Building on these foundations, we propose InternSVG, a unified MLLM for SVG understanding, editing, and generation with SVG-specific special tokens, subword-based embedding initialization, and a two-stage training strategy that progresses from short static SVGs to long-sequence illustrations and complex animations. This unified formulation induces positive transfer and improves overall performance. Experiments on SArena and prior benchmark confirm that InternSVG achieves substantial gains and consistently outperforms leading open and proprietary counterparts.</p> + https://arxiv.org/abs/2510.11341 + Mon, 13 Oct 2025 12:38:04 +0000 - Multi-Agent Tool-Integrated Policy Optimization - https://arxiv.org/abs/2510.04678 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04678.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhanfeng Mo, Xingxuan Li, Yuntao Chen, Lidong Bing</p><p><b>Upvotes:</b> 25</p><p><b>Summary:</b> Large language models (LLMs) increasingly rely on multi-turn tool-integrated planning for knowledge-intensive and complex reasoning tasks. Existing implementations typically rely on a single agent, but they suffer from limited context length and noisy tool responses. A natural solution is to adopt a multi-agent framework with planner- and worker-agents to manage context. However, no existing methods support effective reinforcement learning post-training of tool-integrated multi-agent frameworks. To address this gap, we propose Multi-Agent Tool-Integrated Policy Optimization (MATPO), which enables distinct roles (planner and worker) to be trained within a single LLM instance using role-specific prompts via reinforcement learning. MATPO is derived from a principled credit assignment mechanism across planner and worker rollouts. This design eliminates the need to deploy multiple LLMs, which would be memory-intensive, while preserving the benefits of specialization. Experiments on GAIA-text, WebWalkerQA, and FRAMES show that MATPO consistently outperforms single-agent baselines by an average of 18.38% relative improvement in performance and exhibits greater robustness to noisy tool outputs. Our findings highlight the effectiveness of unifying multiple agent roles within a single LLM and provide practical insights for stable and efficient multi-agent RL training.</p> - https://arxiv.org/abs/2510.04678 - Mon, 06 Oct 2025 10:44:04 +0000 + DocReward: A Document Reward Model for Structuring and Stylizing + https://arxiv.org/abs/2510.11391 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11391.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Junpeng Liu, Yuzhong Zhao, Bowen Cao, Jiayu Ding, Yilin Jia, Tengchao Lv, Yupan Huang, Shaohan Huang, Nan Yang, Li Dong, Lei Cui, Tao Ge, Xun Wang, Huitian Jiao, Sun Mao, FNU Kartik, Si-Qing Chen, Wai Lam, Furu Wei</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> Recent advances in agentic workflows have enabled the automation of tasks such as professional document generation. However, they primarily focus on textual quality, neglecting visual structure and style, which are crucial for readability and engagement. This gap arises mainly from the absence of suitable reward models to guide agentic workflows toward producing documents with stronger structural and stylistic quality. To address this, we propose DocReward, a document reward model that evaluates documents based on their structure and style. We construct a multi-domain dataset DocPair of 117K paired documents, covering 32 domains and 267 document types, each including a high- and low-professionalism document with identical content but different structure and style. This enables the model to evaluate professionalism comprehensively, and in a textual-quality-agnostic way. DocReward is trained using the Bradley-Terry loss to score documents, penalizing predictions that contradict the annotated ranking. To assess the performance of reward models, we create a test dataset containing document bundles ranked by well-educated human evaluators. Notably, DocReward outperforms GPT-4o and GPT-5 in accuracy by 30.6 and 19.4 percentage points, respectively, demonstrating its superiority over baselines. In an extrinsic evaluation of document generation, DocReward achieves a significantly higher win rate of 60.8%, compared to GPT-5's 37.7% win rate, demonstrating its utility in guiding generation agents toward producing human-preferred documents.</p> + https://arxiv.org/abs/2510.11391 + Mon, 13 Oct 2025 13:36:32 +0000 - Hybrid Architectures for Language Models: Systematic Analysis and Design Insights - https://arxiv.org/abs/2510.04800 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04800.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Sangmin Bae, Bilge Acun, Haroun Habeeb, Seungyeon Kim, Chien-Yu Lin, Liang Luo, Junjie Wang, Carole-Jean Wu</p><p><b>Upvotes:</b> 32</p><p><b>Summary:</b> Recent progress in large language models demonstrates that hybrid architectures--combining self-attention mechanisms with structured state space models like Mamba--can achieve a compelling balance between modeling quality and computational efficiency, particularly for long-context tasks. While these hybrid models show promising performance, systematic comparisons of hybridization strategies and analyses on the key factors behind their effectiveness have not been clearly shared to the community. In this work, we present a holistic evaluation of hybrid architectures based on inter-layer (sequential) or intra-layer (parallel) fusion. We evaluate these designs from a variety of perspectives: language modeling performance, long-context capabilities, scaling analysis, and training and inference efficiency. By investigating the core characteristics of their computational primitive, we identify the most critical elements for each hybridization strategy and further propose optimal design recipes for both hybrid models. Our comprehensive analysis provides practical guidance and valuable insights for developing hybrid language models, facilitating the optimization of architectural configurations.</p> - https://arxiv.org/abs/2510.04800 - Mon, 06 Oct 2025 13:30:07 +0000 + ReLook: Vision-Grounded RL with a Multimodal LLM Critic for Agentic Web Coding + https://arxiv.org/abs/2510.11498 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11498.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuhang Li, Chenchen Zhang, Ruilin Lv, Ao Liu, Ken Deng, Yuanxing Zhang, Jiaheng Liu, Wiggin Zhou, Bo Zhou</p><p><b>Upvotes:</b> 10</p><p><b>Summary:</b> While Large Language Models (LLMs) excel at algorithmic code generation, they struggle with front-end development, where correctness is judged on rendered pixels and interaction. We present ReLook, an agentic, vision-grounded reinforcement learning framework that empowers an agent to close a robust generate--diagnose--refine loop by invoking a multimodal LLM (MLLM) as a tool. During training, the agent uses the MLLM-in-the-loop both as a visual critic--scoring code with screenshots--and as a source of actionable, vision-grounded feedback; a strict zero-reward rule for invalid renders anchors renderability and prevents reward hacking. To prevent behavioral collapse, we introduce Forced Optimization, a strict acceptance rule that admits only improving revisions, yielding monotonically better trajectories. At inference, we decouple the critic and run a lightweight, critic-free self-edit cycle, keeping latency comparable to base decoding while retaining most of the gains. Across three widely used benchmarks, ReLook consistently outperforms strong baselines in vision-grounded front-end code generation, highlighting the benefits of agentic perception, visual rewards, and training-inference decoupling.</p> + https://arxiv.org/abs/2510.11498 + Mon, 13 Oct 2025 15:05:50 +0000 - Less is More: Recursive Reasoning with Tiny Networks - https://arxiv.org/abs/2510.04871 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04871.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Alexia Jolicoeur-Martineau</p><p><b>Upvotes:</b> 317</p><p><b>Summary:</b> Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while trained with small models (27M parameters) on small data (around 1000 examples). HRM holds great promise for solving hard problems with small networks, but it is not yet well understood and may be suboptimal. We propose Tiny Recursive Model (TRM), a much simpler recursive reasoning approach that achieves significantly higher generalization than HRM, while using a single tiny network with only 2 layers. With only 7M parameters, TRM obtains 45% test-accuracy on ARC-AGI-1 and 8% on ARC-AGI-2, higher than most LLMs (e.g., Deepseek R1, o3-mini, Gemini 2.5 Pro) with less than 0.01% of the parameters.</p> - https://arxiv.org/abs/2510.04871 - Mon, 06 Oct 2025 14:58:08 +0000 + ACADREASON: Exploring the Limits of Reasoning Models with Academic Research Problems + https://arxiv.org/abs/2510.11652 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11652.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xin Gui, King Zhu, JinCheng Ren, Qianben Chen, Zekun Moore Wang, Yizhi LI, Xinpeng Liu, Xiaowan Li, Wenli Ren, Linyu Miao, Tianrui Qin, Ziqi Shu, He Zhu, Xiangru Tang, Dingfeng Shi, Jiaheng Liu, Yuchen Eleanor Jiang, Minghao Liu, Ge Zhang, Wangchunshu Zhou</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> In recent years, the research focus of large language models (LLMs) and agents has shifted increasingly from demonstrating novel capabilities to complex reasoning and tackling challenging tasks. However, existing evaluations focus mainly on math/code contests or general tasks, while existing multi-domain academic benchmarks lack sufficient reasoning depth, leaving the field without a rigorous benchmark for high-level reasoning. To fill this gap, we introduce the Acadreason benchmark, designed to evaluate the ability of LLMs and agents to acquire and reason over academic knowledge. It consists of 50 expert-annotated academic problems across five high-reasoning domains, including computer science, economics, law, mathematics, and philosophy. All questions are sourced from top-tier publications in recent years and undergo rigorous annotation and quality control to ensure they are both challenging and answerable. We conduct systematic evaluations of over 10 mainstream LLMs and agents. The results show that most LLMs scored below 20 points, with even the cutting-edge GPT-5 achieving only 16 points. While agents achieved higher scores, none exceeded 40 points. This demonstrates the current capability gap between LLMs and agents in super-intelligent academic research tasks and highlights the challenges of Acadreason.</p> + https://arxiv.org/abs/2510.11652 + Mon, 13 Oct 2025 17:30:36 +0000 - Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training - https://arxiv.org/abs/2510.04996 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04996.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wei Xiong, Chenlu Ye, Baohao Liao, Hanze Dong, Xinxing Xu, Christof Monz, Jiang Bian, Nan Jiang, Tong Zhang</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Reinforcement learning applied to large language models (LLMs) for reasoning tasks is often bottlenecked by unstable gradient estimates due to fixed and uniform sampling of responses across prompts. Prior work such as GVM-RAFT addresses this by dynamically allocating inference budget per prompt to minimize stochastic gradient variance under a budget constraint. Inspired by this insight, we propose Reinforce-Ada, an adaptive sampling framework for online RL post-training of LLMs that continuously reallocates sampling effort to the prompts with the greatest uncertainty or learning potential. Unlike conventional two-stage allocation methods, Reinforce-Ada interleaves estimation and sampling in an online successive elimination process, and automatically stops sampling for a prompt once sufficient signal is collected. To stabilize updates, we form fixed-size groups with enforced reward diversity and compute advantage baselines using global statistics aggregated over the adaptive sampling phase. Empirical results across multiple model architectures and reasoning benchmarks show that Reinforce-Ada accelerates convergence and improves final performance compared to GRPO, especially when using the balanced sampling variant. Our work highlights the central role of variance-aware, adaptive data curation in enabling efficient and reliable reinforcement learning for reasoning-capable LLMs. Code is available at https://github.com/RLHFlow/Reinforce-Ada.</p> - https://arxiv.org/abs/2510.04996 - Mon, 06 Oct 2025 16:34:09 +0000 + Boundary-Guided Policy Optimization for Memory-efficient RL of Diffusion Large Language Models + https://arxiv.org/abs/2510.11683 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11683.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Nianyi Lin, Jiajie Zhang, Lei Hou, Juanzi Li</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> A key challenge in applying reinforcement learning (RL) to diffusion large language models (dLLMs) lies in the intractability of their likelihood functions, which are essential for the RL objective, necessitating corresponding approximation in each training step. While existing methods approximate the log-likelihoods by their evidence lower bounds (ELBOs) via customized Monte Carlo (MC) sampling, the forward computational graphs of all MC samples need to be retained for the gradient computation of non-linear terms in the RL objective, resulting in significant memory overhead. This constraint restricts feasible sample sizes, leading to imprecise likelihood approximations and ultimately distorting the RL objective. To overcome this limitation, we propose Boundary-Guided Policy Optimization (BGPO), a memory-efficient RL algorithm that maximizes a specially constructed lower bound of the ELBO-based objective. This lower bound is carefully designed to satisfy two key properties: (1) Linearity: it is formulated in a linear sum where each term depends only on a single MC sample, thereby enabling gradient accumulation across samples and ensuring constant memory usage; (2) Equivalence: Both the value and gradient of this lower bound are equal to those of the ELBO-based objective in on-policy training, making it also an effective approximation for the original RL objective. These properties allow BGPO to adopt a large MC sample size, resulting in more accurate likelihood approximations and improved RL objective estimation, which in turn leads to enhanced performance. Experiments show that BGPO significantly outperforms previous RL algorithms for dLLMs in math problem solving, code generation, and planning tasks.</p> + https://arxiv.org/abs/2510.11683 + Mon, 13 Oct 2025 17:47:50 +0000 - Imperceptible Jailbreaking against Large Language Models - https://arxiv.org/abs/2510.05025 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05025.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kuofeng Gao, Yiming Li, Chao Du, Xin Wang, Xingjun Ma, Shu-Tao Xia, Tianyu Pang</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> Jailbreaking attacks on the vision modality typically rely on imperceptible adversarial perturbations, whereas attacks on the textual modality are generally assumed to require visible modifications (e.g., non-semantic suffixes). In this paper, we introduce imperceptible jailbreaks that exploit a class of Unicode characters called variation selectors. By appending invisible variation selectors to malicious questions, the jailbreak prompts appear visually identical to original malicious questions on screen, while their tokenization is "secretly" altered. We propose a chain-of-search pipeline to generate such adversarial suffixes to induce harmful responses. Our experiments show that our imperceptible jailbreaks achieve high attack success rates against four aligned LLMs and generalize to prompt injection attacks, all without producing any visible modifications in the written prompt. Our code is available at https://github.com/sail-sg/imperceptible-jailbreaks.</p> - https://arxiv.org/abs/2510.05025 - Mon, 06 Oct 2025 17:03:50 +0000 + Diffusion Transformers with Representation Autoencoders + https://arxiv.org/abs/2510.11690 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11690.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Boyang Zheng, Nanye Ma, Shengbang Tong, Saining Xie</p><p><b>Upvotes:</b> 144</p><p><b>Summary:</b> Latent generative modeling, where a pretrained autoencoder maps pixels into a latent space for the diffusion process, has become the standard strategy for Diffusion Transformers (DiT); however, the autoencoder component has barely evolved. Most DiTs continue to rely on the original VAE encoder, which introduces several limitations: outdated backbones that compromise architectural simplicity, low-dimensional latent spaces that restrict information capacity, and weak representations that result from purely reconstruction-based training and ultimately limit generative quality. In this work, we explore replacing the VAE with pretrained representation encoders (e.g., DINO, SigLIP, MAE) paired with trained decoders, forming what we term Representation Autoencoders (RAEs). These models provide both high-quality reconstructions and semantically rich latent spaces, while allowing for a scalable transformer-based architecture. Since these latent spaces are typically high-dimensional, a key challenge is enabling diffusion transformers to operate effectively within them. We analyze the sources of this difficulty, propose theoretically motivated solutions, and validate them empirically. Our approach achieves faster convergence without auxiliary representation alignment losses. Using a DiT variant equipped with a lightweight, wide DDT head, we achieve strong image generation results on ImageNet: 1.51 FID at 256x256 (no guidance) and 1.13 at both 256x256 and 512x512 (with guidance). RAE offers clear advantages and should be the new default for diffusion transformer training.</p> + https://arxiv.org/abs/2510.11690 + Mon, 13 Oct 2025 17:51:39 +0000 - Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models - https://arxiv.org/abs/2510.05034 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05034.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yunlong Tang, Jing Bi, Pinxin Liu, Zhenyu Pan, Zhangyun Tan, Qianxiang Shen, Jiani Liu, Hang Hua, Junjia Guo, Yunzhong Xiao, Chao Huang, Zhiyuan Wang, Susan Liang, Xinyi Liu, Yizhi Song, Yuhe Nie, Jia-Xing Zhong, Bozheng Li, Daiqing Qi, Ziyun Zeng, Ali Vosoughi, Luchuan Song, Zeliang Zhang, Daiki Shimada, Han Liu, Jiebo Luo, Chenliang Xu</p><p><b>Upvotes:</b> 42</p><p><b>Summary:</b> Video understanding represents the most challenging frontier in computer vision, requiring models to reason about complex spatiotemporal relationships, long-term dependencies, and multimodal evidence. The recent emergence of Video-Large Multimodal Models (Video-LMMs), which integrate visual encoders with powerful decoder-based language models, has demonstrated remarkable capabilities in video understanding tasks. However, the critical phase that transforms these models from basic perception systems into sophisticated reasoning engines, post-training, remains fragmented across the literature. This survey provides the first comprehensive examination of post-training methodologies for Video-LMMs, encompassing three fundamental pillars: supervised fine-tuning (SFT) with chain-of-thought, reinforcement learning (RL) from verifiable objectives, and test-time scaling (TTS) through enhanced inference computation. We present a structured taxonomy that clarifies the roles, interconnections, and video-specific adaptations of these techniques, addressing unique challenges such as temporal localization, spatiotemporal grounding, long video efficiency, and multimodal evidence integration. Through systematic analysis of representative methods, we synthesize key design principles, insights, and evaluation protocols while identifying critical open challenges in reward design, scalability, and cost-performance optimization. We further curate essential benchmarks, datasets, and metrics to facilitate rigorous assessment of post-training effectiveness. This survey aims to provide researchers and practitioners with a unified framework for advancing Video-LMM capabilities. Additional resources and updates are maintained at: https://github.com/yunlong10/Awesome-Video-LMM-Post-Training</p> - https://arxiv.org/abs/2510.05034 - Mon, 06 Oct 2025 17:10:44 +0000 + Scaling Language-Centric Omnimodal Representation Learning + https://arxiv.org/abs/2510.11693 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11693.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chenghao Xiao, Hou Pong Chan, Hao Zhang, Weiwen Xu, Mahani Aljunied, Yu Rong</p><p><b>Upvotes:</b> 94</p><p><b>Summary:</b> Recent multimodal embedding approaches leveraging multimodal large language models (MLLMs) fine-tuned with contrastive learning (CL) have shown promising results, yet the underlying reasons behind their superiority remain underexplored. This work argues that a crucial advantage of MLLM-based approaches stems from implicit cross-modal alignment achieved during generative pretraining, where the language decoder learns to exploit multimodal signals within a shared representation space for generating unimodal outputs. Through analysis of anisotropy and kernel similarity structure, we empirically confirm that latent alignment emerges within MLLM representations, allowing CL to serve as a lightweight refinement stage. Leveraging this insight, we propose a Language-Centric Omnimodal Embedding framework, termed LCO-Emb. Extensive experiments across diverse backbones and benchmarks demonstrate its effectiveness, achieving state-of-the-art performance across modalities. Furthermore, we identify a Generation-Representation Scaling Law (GRSL), showing that the representational capabilities gained through contrastive refinement scales positively with the MLLM's generative capabilities. This suggests that improving generative abilities evolves as an effective paradigm for enhancing representation quality. We provide a theoretical explanation of GRSL, which formally links the MLLM's generative quality to the upper bound on its representation performance, and validate it on a challenging, low-resource visual-document retrieval task, showing that continual generative pretraining before CL can further enhance the potential of a model's embedding capabilities. Codes, models, and resources are available at https://github.com/LCO-Embedding/LCO-Embedding.</p> + https://arxiv.org/abs/2510.11693 + Mon, 13 Oct 2025 17:53:52 +0000 - StaMo: Unsupervised Learning of Generalizable Robot Motion from Compact State Representation - https://arxiv.org/abs/2510.05057 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05057.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Mingyu Liu, Jiuhe Shu, Hui Chen, Zeju Li, Canyu Zhao, Jiange Yang, Shenyuan Gao, Hao Chen, Chunhua Shen</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> A fundamental challenge in embodied intelligence is developing expressive and compact state representations for efficient world modeling and decision making. However, existing methods often fail to achieve this balance, yielding representations that are either overly redundant or lacking in task-critical information. We propose an unsupervised approach that learns a highly compressed two-token state representation using a lightweight encoder and a pre-trained Diffusion Transformer (DiT) decoder, capitalizing on its strong generative prior. Our representation is efficient, interpretable, and integrates seamlessly into existing VLA-based models, improving performance by 14.3% on LIBERO and 30% in real-world task success with minimal inference overhead. More importantly, we find that the difference between these tokens, obtained via latent interpolation, naturally serves as a highly effective latent action, which can be further decoded into executable robot actions. This emergent capability reveals that our representation captures structured dynamics without explicit supervision. We name our method StaMo for its ability to learn generalizable robotic Motion from compact State representation, which is encoded from static images, challenging the prevalent dependence to learning latent action on complex architectures and video data. The resulting latent actions also enhance policy co-training, outperforming prior methods by 10.4% with improved interpretability. Moreover, our approach scales effectively across diverse data sources, including real-world robot data, simulation, and human egocentric video.</p> - https://arxiv.org/abs/2510.05057 - Mon, 06 Oct 2025 17:37:24 +0000 + QeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMs + https://arxiv.org/abs/2510.11696 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11696.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wei Huang, Yi Ge, Shuai Yang, Yicheng Xiao, Huizi Mao, Yujun Lin, Hanrong Ye, Sifei Liu, Ka Chun Cheung, Hongxu Yin, Yao Lu, Xiaojuan Qi, Song Han, Yukang Chen</p><p><b>Upvotes:</b> 157</p><p><b>Summary:</b> We propose QeRL, a Quantization-enhanced Reinforcement Learning framework for large language models (LLMs). While RL is essential for LLMs' reasoning capabilities, it is resource-intensive, requiring substantial GPU memory and long rollout durations. QeRL addresses these issues by combining NVFP4 quantization with Low-Rank Adaptation (LoRA), accelerating rollout phase of RL while reducing memory overhead. Beyond efficiency, our findings show that quantization noise increases policy entropy, enhancing exploration, and enabling the discovery of better strategies during RL. To further optimize exploration, QeRL introduces an Adaptive Quantization Noise (AQN) mechanism, which dynamically adjusts noise during training. Experiments demonstrate that QeRL delivers over 1.5 times speedup in the rollout phase. Moreover, this is the first framework to enable RL training of a 32B LLM on a single H100 80GB GPU, while delivering overall speedups for RL training. It also achieves faster reward growth and higher final accuracy than 16-bit LoRA and QLoRA, while matching the performance of full-parameter fine-tuning on mathematical benchmarks such as GSM8K (90.8%) and MATH 500 (77.4%) in the 7B model. These results establish QeRL as an efficient and effective framework for RL training in LLMs.</p> + https://arxiv.org/abs/2510.11696 + Mon, 13 Oct 2025 17:55:09 +0000 - SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMs - https://arxiv.org/abs/2510.05069 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05069.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Dachuan Shi, Abedelkadir Asi, Keying Li, Xiangchi Yuan, Leyan Pan, Wenke Lee, Wen Xiao</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Recent work shows that, beyond discrete reasoning through explicit chain-of-thought steps, which are limited by the boundaries of natural languages, large language models (LLMs) can also reason continuously in latent space, allowing richer information per step and thereby improving token efficiency. Despite this promise, latent reasoning still faces two challenges, especially in training-free settings: 1) purely latent reasoning broadens the search distribution by maintaining multiple implicit paths, which diffuses probability mass, introduces noise, and impedes convergence to a single high-confidence solution, thereby hurting accuracy; and 2) overthinking persists even without explicit text, wasting tokens and degrading efficiency. To address these issues, we introduce SwiReasoning, a training-free framework for LLM reasoning which features two key innovations: 1) SwiReasoning dynamically switches between explicit and latent reasoning, guided by block-wise confidence estimated from entropy trends in next-token distributions, to balance exploration and exploitation and promote timely convergence. 2) By limiting the maximum number of thinking-block switches, SwiReasoning curbs overthinking and improves token efficiency across varying problem difficulties. On widely used mathematics and STEM benchmarks, SwiReasoning consistently improves average accuracy by 1.5%-2.8% across reasoning LLMs of different model families and scales. Furthermore, under constrained budgets, SwiReasoning improves average token efficiency by 56%-79%, with larger gains as budgets tighten.</p> - https://arxiv.org/abs/2510.05069 - Mon, 06 Oct 2025 17:46:34 +0000 + GAR: Generative Adversarial Reinforcement Learning for Formal Theorem Proving + https://arxiv.org/abs/2510.11769 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11769.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ruida Wang, Jiarui Yao, Rui Pan, Shizhe Diao, Tong Zhang</p><p><b>Upvotes:</b> 25</p><p><b>Summary:</b> Solving math problems through verifiable languages such as Lean has significantly impacted both the mathematics and computer science communities. Current state-of-the-art models are often trained with expensive online Reinforcement Learning (RL) or expert iteration. However, these approaches rely on fixed problem sets, which causes inefficient training and limits the model to tackle complex problems. To overcome these limitations, we propose GAR: Generative Adversarial Reinforcement learning, a comprehensive RL training framework that jointly trains the problem composer and solver in an adversarial loop. GAR introduces an implicit curriculum learning mechanism, which aligns task difficulty with the prover's evolving capability. It thereby improves the training efficiency and enables stronger performance of proving advanced theorems. Experiments show that with GAR training, Goedel-Prover-V2-8B and DeepSeek-Prover-V2-7B achieve an average relative improvement in pass@32 of 4.20% on MiniF2F-Test benchmark, while DeepSeek-Prover-V2's pass@32 on ProofNet-Test increases from 22.58% to 25.81%. Beyond formal proving, GAR establishes a general RL paradigm for co-evolution of problem generation and solving under verifiable environments.</p> + https://arxiv.org/abs/2510.11769 + Mon, 13 Oct 2025 17:56:25 +0000 - Factuality Matters: When Image Generation and Editing Meet Structured Visuals - https://arxiv.org/abs/2510.05091 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05091.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Le Zhuo, Songhao Han, Yuandong Pu, Boxiang Qiu, Sayak Paul, Yue Liao, Yihao Liu, Jie Shao, Xi Chen, Si Liu, Hongsheng Li</p><p><b>Upvotes:</b> 15</p><p><b>Summary:</b> While modern visual generation models excel at creating aesthetically pleasing natural images, they struggle with producing or editing structured visuals like charts, diagrams, and mathematical figures, which demand composition planning, text rendering, and multimodal reasoning for factual fidelity. To address this, we present the first comprehensive, systematic investigation of this domain, encompassing data construction, model training, and an evaluation benchmark. First, we construct a large-scale dataset of 1.3 million high-quality structured image pairs derived from executable drawing programs and augmented with chain-of-thought reasoning annotations. Building on it, we train a unified model that integrates a VLM with FLUX.1 Kontext via a lightweight connector for enhanced multimodal understanding. A three-stage training curriculum enables progressive feature alignment, knowledge infusion, and reasoning-augmented generation, further boosted by an external reasoner at inference time. Finally, we introduce StructBench, a novel benchmark for generation and editing with over 1,700 challenging instances, and an accompanying evaluation metric, StructScore, which employs a multi-round Q\&A protocol to assess fine-grained factual accuracy. Evaluations of 15 models reveal that even leading closed-source systems remain far from satisfactory. Our model attains strong editing performance, and inference-time reasoning yields consistent gains across diverse architectures. By releasing the dataset, model, and benchmark, we aim to advance unified multimodal foundations for structured visuals.</p> - https://arxiv.org/abs/2510.05091 - Mon, 06 Oct 2025 17:56:55 +0000 + Demystifying Reinforcement Learning in Agentic Reasoning + https://arxiv.org/abs/2510.11701 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11701.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhaochen Yu, Ling Yang, Jiaru Zou, Shuicheng Yan, Mengdi Wang</p><p><b>Upvotes:</b> 29</p><p><b>Summary:</b> Recently, the emergence of agentic RL has showcased that RL could also effectively improve the agentic reasoning ability of LLMs, yet the key design principles and optimal practices remain unclear. In this work, we conduct a comprehensive and systematic investigation to demystify reinforcement learning in agentic reasoning from three key perspectives: data, algorithm, and reasoning mode. We highlight our key insights: (i) Replacing stitched synthetic trajectories with real end-to-end tool-use trajectories yields a far stronger SFT initialization; high-diversity, model-aware datasets sustain exploration and markedly improve RL performance. (ii) Exploration-friendly techniques are crucial for agentic RL, such as clip higher, overlong reward shaping, and maintaining adequate policy entropy could improve the training efficiency. (iii) A deliberative strategy with fewer tool calls outperforms frequent tool calls or verbose self-reasoning, improving tool efficiency and final accuracy. Together, these simple practices consistently enhance agentic reasoning and training efficiency, achieving strong results on challenging benchmarks with smaller models, and establishing a practical baseline for future agentic RL research. Beyond these empirical insights, we further contribute a high-quality, real end-to-end agentic SFT dataset along with a high-quality RL dataset, and demonstrate the effectiveness of our insights in boosting the agentic reasoning ability of LLMs across four challenging benchmarks, including AIME2024/AIME2025, GPQA-Diamond, and LiveCodeBench-v6. With our recipes, 4B-sized models could also achieve superior agentic reasoning performance compared to 32B-sized models. Code and models: https://github.com/Gen-Verse/Open-AgentRL</p> + https://arxiv.org/abs/2510.11701 + Mon, 13 Oct 2025 17:57:15 +0000 - VChain: Chain-of-Visual-Thought for Reasoning in Video Generation - https://arxiv.org/abs/2510.05094 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05094.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ziqi Huang, Ning Yu, Gordon Chen, Haonan Qiu, Paul Debevec, Ziwei Liu</p><p><b>Upvotes:</b> 34</p><p><b>Summary:</b> Recent video generation models can produce smooth and visually appealing clips, but they often struggle to synthesize complex dynamics with a coherent chain of consequences. Accurately modeling visual outcomes and state transitions over time remains a core challenge. In contrast, large language and multimodal models (e.g., GPT-4o) exhibit strong visual state reasoning and future prediction capabilities. To bridge these strengths, we introduce VChain, a novel inference-time chain-of-visual-thought framework that injects visual reasoning signals from multimodal models into video generation. Specifically, VChain contains a dedicated pipeline that leverages large multimodal models to generate a sparse set of critical keyframes as snapshots, which are then used to guide the sparse inference-time tuning of a pre-trained video generator only at these key moments. Our approach is tuning-efficient, introduces minimal overhead and avoids dense supervision. Extensive experiments on complex, multi-step scenarios show that VChain significantly enhances the quality of generated videos.</p> - https://arxiv.org/abs/2510.05094 - Mon, 06 Oct 2025 17:57:59 +0000 + DiT360: High-Fidelity Panoramic Image Generation via Hybrid Training + https://arxiv.org/abs/2510.11712 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11712.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haoran Feng, Dizhe Zhang, Xiangtai Li, Bo Du, Lu Qi</p><p><b>Upvotes:</b> 29</p><p><b>Summary:</b> In this work, we propose DiT360, a DiT-based framework that performs hybrid training on perspective and panoramic data for panoramic image generation. For the issues of maintaining geometric fidelity and photorealism in generation quality, we attribute the main reason to the lack of large-scale, high-quality, real-world panoramic data, where such a data-centric view differs from prior methods that focus on model design. Basically, DiT360 has several key modules for inter-domain transformation and intra-domain augmentation, applied at both the pre-VAE image level and the post-VAE token level. At the image level, we incorporate cross-domain knowledge through perspective image guidance and panoramic refinement, which enhance perceptual quality while regularizing diversity and photorealism. At the token level, hybrid supervision is applied across multiple modules, which include circular padding for boundary continuity, yaw loss for rotational robustness, and cube loss for distortion awareness. Extensive experiments on text-to-panorama, inpainting, and outpainting tasks demonstrate that our method achieves better boundary consistency and image fidelity across eleven quantitative metrics. Our code is available at https://github.com/Insta360-Research-Team/DiT360.</p> + https://arxiv.org/abs/2510.11712 + Mon, 13 Oct 2025 17:59:15 +0000 - Paper2Video: Automatic Video Generation from Scientific Papers - https://arxiv.org/abs/2510.05096 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05096.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zeyu Zhu, Kevin Qinghong Lin, Mike Zheng Shou</p><p><b>Upvotes:</b> 93</p><p><b>Summary:</b> Academic presentation videos have become an essential medium for research communication, yet producing them remains highly labor-intensive, often requiring hours of slide design, recording, and editing for a short 2 to 10 minutes video. Unlike natural video, presentation video generation involves distinctive challenges: inputs from research papers, dense multi-modal information (text, figures, tables), and the need to coordinate multiple aligned channels such as slides, subtitles, speech, and human talker. To address these challenges, we introduce PaperTalker, the first benchmark of 101 research papers paired with author-created presentation videos, slides, and speaker metadata. We further design four tailored evaluation metrics--Meta Similarity, PresentArena, PresentQuiz, and IP Memory--to measure how videos convey the paper's information to the audience. Building on this foundation, we propose PaperTalker, the first multi-agent framework for academic presentation video generation. It integrates slide generation with effective layout refinement by a novel effective tree search visual choice, cursor grounding, subtitling, speech synthesis, and talking-head rendering, while parallelizing slide-wise generation for efficiency. Experiments on Paper2Video demonstrate that the presentation videos produced by our approach are more faithful and informative than existing baselines, establishing a practical step toward automated and ready-to-use academic video generation. Our dataset, agent, and code are available at https://github.com/showlab/Paper2Video.</p> - https://arxiv.org/abs/2510.05096 - Mon, 06 Oct 2025 17:58:02 +0000 + CodePlot-CoT: Mathematical Visual Reasoning by Thinking with Code-Driven Images + https://arxiv.org/abs/2510.11718 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11718.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chengqi Duan, Kaiyue Sun, Rongyao Fang, Manyuan Zhang, Yan Feng, Ying Luo, Yufang Liu, Ke Wang, Peng Pei, Xunliang Cai, Hongsheng Li, Yi Ma, Xihui Liu</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> Recent advances in Large Language Models (LLMs) and Vision Language Models (VLMs) have shown significant progress in mathematical reasoning, yet they still face a critical bottleneck with problems requiring visual assistance, such as drawing auxiliary lines or plotting functions to solve the problems. Most LLMs and VLMs are constrained to text-only reasoning chains, while multimodal unified models that can generate interleaved text and images lack the necessary precision and controllability for such tasks. To address this, we propose CodePlot-CoT, a code-driven Chain-of-Thought paradigm for "thinking with images" in mathematics. Our approach leverages the VLM to generate text reasoning as well as executable plotting code, which is then rendered into images as "visual thought", to solve mathematical problems. To achieve this, we first construct Math-VR, the first large-scale, bilingual dataset and benchmark for Mathematics problems with Visual Reasoning, comprising 178K samples. Second, to create high-quality training data, we develop a state-of-the-art image-to-code converter specialized for parsing complex mathematical figures into codes. Finally, using these training data, we train the CodePlot-CoT model for solving mathematical problems. Experimental results show that our model achieves up to 21% increase over base model on our new benchmark, fully validating the efficacy of our proposed code-driven reasoning paradigm. Our work opens a new direction for multimodal mathematical reasoning and provides the community with the first large-scale dataset, comprehensive benchmark, and strong approach for such problems. To facilitate future research, we make our datasets, code, and pretrained models publicly available at https://github.com/HKU-MMLab/Math-VR-CodePlot-CoT.</p> + https://arxiv.org/abs/2510.11718 + Mon, 13 Oct 2025 17:59:55 +0000 - BIRD-INTERACT: Re-imagining Text-to-SQL Evaluation for Large Language Models via Lens of Dynamic Interactions - https://arxiv.org/abs/2510.05318 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05318.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Nan Huo, Xiaohan Xu, Jinyang Li, Per Jacobsson, Shipei Lin, Bowen Qin, Binyuan Hui, Xiaolong Li, Ge Qu, Shuzheng Si, Linheng Han, Edward Alexander, Xintong Zhu, Rui Qin, Ruihan Yu, Yiyao Jin, Feige Zhou, Weihao Zhong, Yun Chen, Hongyu Liu, Chenhao Ma, Fatma Ozcan, Yannis Papakonstantinou, Reynold Cheng</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> Large language models (LLMs) have demonstrated remarkable performance on single-turn text-to-SQL tasks, but real-world database applications predominantly require multi-turn interactions to handle ambiguous queries, execution errors, and evolving user requirements. Existing multi-turn benchmarks fall short by treating conversation histories as static context or limiting evaluation to read-only operations, failing to reflect production-grade database assistant challenges. We introduce BIRD-INTERACT, a benchmark that restores this realism through: (1) a comprehensive interaction environment coupling each database with a hierarchical knowledge base, metadata files, and a function-driven user simulator, enabling models to solicit clarifications, retrieve knowledge, and recover from errors without human supervision; (2) two evaluation settings consisting of a pre-defined conversational protocol (c-Interact) and an open-ended agentic setting (a-Interact) where models autonomously decide when to query the user simulator or explore the environment; (3) a challenging task suite covering the full CRUD spectrum for business-intelligence and operational use cases, guarded by executable test cases. Each task features ambiguous and follow-up sub-tasks requiring dynamic interaction. The suite comprises BIRD-INTERACT-FULL (600 tasks, up to 11,796 interactions) for comprehensive performance assessment, and BIRD-INTERACT-LITE (300 tasks with simplified databases) for detailed behavioral analysis and rapid method development. Our empirical results highlight BIRD-INTERACT's difficulty: GPT-5 completes only 8.67% of tasks in c-Interact and 17.00% in a-Interact. Analysis via memory grafting and Interaction Test-time Scaling validates the importance of effective interaction for complex, dynamic text-to-SQL tasks.</p> - https://arxiv.org/abs/2510.05318 - Mon, 06 Oct 2025 19:31:47 +0000 + R-WoM: Retrieval-augmented World Model For Computer-use Agents + https://arxiv.org/abs/2510.11892 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.11892.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kai Mei, Jiang Guo, Shuaichen Chang, Mingwen Dong, Dongkyu Lee, Xing Niu, Jiarong Jiang</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> Large Language Models (LLMs) can serve as world models to enhance agent decision-making in digital environments by simulating future states and predicting action outcomes, potentially eliminating costly trial-and-error exploration. However, this capability is fundamentally limited by LLMs' tendency toward hallucination and their reliance on static training knowledge, which can lead to compounding errors that inhibit long-horizon simulations. To systematically investigate whether LLMs are appropriate for world modeling, we probe two core capabilities of world models--future state prediction and reward estimation--through three tasks: next-state identification, full-procedure planning alignment, and milestone transition recognition. Our analysis shows that while LLMs effectively capture immediate next states and identify meaningful state transitions, their performance rapidly degrades in full-procedure planning. This highlights LLMs' limitations in reliably modeling environment dynamics over long horizons. To address these limitations, we propose the Retrieval-augmented World Model (R-WoM), which grounds LLM simulations by incorporating factual, up-to-date knowledge retrieved from external tutorials. Experiments show that R-WoM achieves substantial improvements of up to 25.3% (OSWorld) and 18.1% (WebArena) compared to baselines, with particular advantages in longer-horizon simulations.</p> + https://arxiv.org/abs/2510.11892 + Mon, 13 Oct 2025 19:52:04 +0000 - Scalable In-context Ranking with Generative Models - https://arxiv.org/abs/2510.05396 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05396.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Nilesh Gupta, Chong You, Srinadh Bhojanapalli, Sanjiv Kumar, Inderjit Dhillon, Felix Yu</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> In-context Ranking (ICR) is an emerging paradigm for Information Retrieval (IR), which leverages contextual understanding of LLMs by directly incorporating the task description, candidate documents, and the query into the model's input prompt and tasking the LLM to identify relevant document(s). While it is effective, efficiency is a significant challenge in this paradigm, especially as the candidate list grows due to quadratic/super-linear scaling of attention operation with context length. To this end, this paper first identifies inherent and exploitable structures in the attention of LLMs finetuned for ICR: (1) inter-document block sparsity: attention is dense within each document block but sparse across different documents in the context; and (2) query-document block relevance: the attention scores from certain query tokens to a document block in middle layers strongly correlate with that document's actual relevance. Motivated by these observations, we introduce BlockRank (Blockwise In-context Ranking), a novel method that adapts the attention operation in an LLM by (a) architecturally enforcing the observed inter-document block sparsity, reducing attention complexity from quadratic to linear without loss in performance, and (b) optimizing query-document block relevance for true relevant documents during fine-tuning using an auxiliary contrastive training objective, improving retrieval in attention. Experiments on BEIR, MSMarco and NQ with Mistral-7B demonstrate that FLARE Mistral matches or outperforms existing SOTA listwise rankers and controlled fine-tuned baseline while being significantly more efficient at inference (4.7x for 100 MSMarco documents in context) and scaling gracefully to long-context shortlists, around 500 documents in-context (approximately 100K context length) within a second, presenting a scalable and effective solution for ICR.</p> - https://arxiv.org/abs/2510.05396 - Mon, 06 Oct 2025 21:41:58 +0000 + Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model + https://arxiv.org/abs/2510.12276 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12276.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Fuhao Li, Wenxuan Song, Han Zhao, Jingbo Wang, Pengxiang Ding, Donglin Wang, Long Zeng, Haoang Li</p><p><b>Upvotes:</b> 137</p><p><b>Summary:</b> Vision-language-action (VLA) models have recently shown strong potential in enabling robots to follow language instructions and execute precise actions. However, most VLAs are built upon vision-language models pretrained solely on 2D data, which lack accurate spatial awareness and hinder their ability to operate in the 3D physical world. Existing solutions attempt to incorporate explicit 3D sensor inputs such as depth maps or point clouds, but these approaches face challenges due to sensor noise, hardware heterogeneity, and incomplete depth coverage in existing datasets. Alternative methods that estimate 3D cues from 2D images also suffer from the limited performance of depth estimators.We propose Spatial Forcing (SF), a simple yet effective alignment strategy that implicitly forces VLA models to develop spatial comprehension capabilities without relying on explicit 3D inputs or depth estimators. SF aligns intermediate visual embeddings of VLAs with geometric representations produced by pretrained 3D foundation models. By enforcing alignment at intermediate layers, SF guides VLAs to encode richer spatial representations that enhance action precision.Extensive experiments in simulation and real-world environments demonstrate that SF achieves state-of-the-art results, surpassing both 2D- and 3D-based VLAs. SF further accelerates training by up to 3.8x and improves data efficiency across diverse robotic tasks. Project page is at https://spatial-forcing.github.io/</p> + https://arxiv.org/abs/2510.12276 + Tue, 14 Oct 2025 08:27:10 +0000 - Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic Presentations - https://arxiv.org/abs/2510.05571 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05571.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chengzhi Liu, Yuzhe Yang, Kaiwen Zhou, Zhen Zhang, Yue Fan, Yannan Xie, Peng Qi, Xin Eric Wang</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> The promotion of academic papers has become an important means of enhancing research visibility. However, existing automated methods struggle limited storytelling, insufficient aesthetic quality, and constrained self-adjustment, making it difficult to achieve efficient and engaging dissemination. At the heart of those challenges is a simple principle: there is no way to improve it when you cannot evaluate it right. To address this, we introduce EvoPresent, a self-improvement agent framework that unifies coherent narratives, aesthetic-aware designs, and realistic presentation delivery via virtual characters. Central to EvoPresent is PresAesth, a multi-task reinforcement learning (RL) aesthetic model that provides reliable aesthetic scoring, defect adjustment, and comparative feedback, enabling iterative self-improvement even under limited aesthetic training data. To systematically evaluate the methods, we introduce EvoPresent Benchmark, a comprehensive benchmark comprising: Presentation Generation Quality, built on 650 top-tier AI conference papers with multimodal resources (slides, videos and scripts) to assess both content and design; and Aesthetic Awareness, consisting of 2,000 slide pairs with varying aesthetic levels, supporting joint training and evaluation on scoring, defect adjustment, and comparison. Our findings highlight that (i) High-quality feedback is essential for agent self-improvement, while initial capability alone does not guarantee effective self-correction. (ii) Automated generation pipelines exhibit a trade-off between visual design and content construction. (iii) Multi-task RL training shows stronger generalization in aesthetic awareness tasks.</p> - https://arxiv.org/abs/2510.05571 - Tue, 07 Oct 2025 04:24:26 +0000 + RAG-Anything: All-in-One RAG Framework + https://arxiv.org/abs/2510.12323 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12323.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zirui Guo, Xubin Ren, Lingrui Xu, Jiahao Zhang, Chao Huang</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> Retrieval-Augmented Generation (RAG) has emerged as a fundamental paradigm for expanding Large Language Models beyond their static training limitations. However, a critical misalignment exists between current RAG capabilities and real-world information environments. Modern knowledge repositories are inherently multimodal, containing rich combinations of textual content, visual elements, structured tables, and mathematical expressions. Yet existing RAG frameworks are limited to textual content, creating fundamental gaps when processing multimodal documents. We present RAG-Anything, a unified framework that enables comprehensive knowledge retrieval across all modalities. Our approach reconceptualizes multimodal content as interconnected knowledge entities rather than isolated data types. The framework introduces dual-graph construction to capture both cross-modal relationships and textual semantics within a unified representation. We develop cross-modal hybrid retrieval that combines structural knowledge navigation with semantic matching. This enables effective reasoning over heterogeneous content where relevant evidence spans multiple modalities. RAG-Anything demonstrates superior performance on challenging multimodal benchmarks, achieving significant improvements over state-of-the-art methods. Performance gains become particularly pronounced on long documents where traditional approaches fail. Our framework establishes a new paradigm for multimodal knowledge access, eliminating the architectural fragmentation that constrains current systems. Our framework is open-sourced at: https://github.com/HKUDS/RAG-Anything.</p> + https://arxiv.org/abs/2510.12323 + Tue, 14 Oct 2025 09:25:35 +0000 - In-the-Flow Agentic System Optimization for Effective Planning and Tool Use - https://arxiv.org/abs/2510.05592 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05592.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhuofeng Li, Haoxiang Zhang, Seungju Han, Sheng Liu, Jianwen Xie, Yu Zhang, Yejin Choi, James Zou, Pan Lu</p><p><b>Upvotes:</b> 75</p><p><b>Summary:</b> Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a single, monolithic policy that interleaves thoughts and tool calls under full context; this scales poorly with long horizons and diverse tools and generalizes weakly to new scenarios. Agentic systems offer a promising alternative by decomposing work across specialized modules, yet most remain training-free or rely on offline training decoupled from the live dynamics of multi-turn interaction. We introduce AgentFlow, a trainable, in-the-flow agentic framework that coordinates four modules (planner, executor, verifier, generator) through an evolving memory and directly optimizes its planner inside the multi-turn loop. To train on-policy in live environments, we propose Flow-based Group Refined Policy Optimization (Flow-GRPO), which tackles long-horizon, sparse-reward credit assignment by converting multi-turn optimization into a sequence of tractable single-turn policy updates. It broadcasts a single, verifiable trajectory-level outcome to every turn to align local planner decisions with global success and stabilizes learning with group-normalized advantages. Across ten benchmarks, AgentFlow with a 7B-scale backbone outperforms top-performing baselines with average accuracy gains of 14.9% on search, 14.0% on agentic, 14.5% on mathematical, and 4.1% on scientific tasks, even surpassing larger proprietary models like GPT-4o. Further analyses confirm the benefits of in-the-flow optimization, showing improved planning, enhanced tool-calling reliability, and positive scaling with model size and reasoning turns.</p> - https://arxiv.org/abs/2510.05592 - Tue, 07 Oct 2025 05:32:44 +0000 + A Survey of Vibe Coding with Large Language Models + https://arxiv.org/abs/2510.12399 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12399.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuyao Ge, Lingrui Mei, Zenghao Duan, Tianhao Li, Yujia Zheng, Yiwei Wang, Lexin Wang, Jiayu Yao, Tianyu Liu, Yujun Cai, Baolong Bi, Fangda Guo, Jiafeng Guo, Shenghua Liu, Xueqi Cheng</p><p><b>Upvotes:</b> 45</p><p><b>Summary:</b> The advancement of large language models (LLMs) has catalyzed a paradigm shift from code generation assistance to autonomous coding agents, enabling a novel development methodology termed "Vibe Coding" where developers validate AI-generated implementations through outcome observation rather than line-by-line code comprehension. Despite its transformative potential, the effectiveness of this emergent paradigm remains under-explored, with empirical evidence revealing unexpected productivity losses and fundamental challenges in human-AI collaboration. To address this gap, this survey provides the first comprehensive and systematic review of Vibe Coding with large language models, establishing both theoretical foundations and practical frameworks for this transformative development approach. Drawing from systematic analysis of over 1000 research papers, we survey the entire vibe coding ecosystem, examining critical infrastructure components including LLMs for coding, LLM-based coding agent, development environment of coding agent, and feedback mechanisms. We first introduce Vibe Coding as a formal discipline by formalizing it through a Constrained Markov Decision Process that captures the dynamic triadic relationship among human developers, software projects, and coding agents. Building upon this theoretical foundation, we then synthesize existing practices into five distinct development models: Unconstrained Automation, Iterative Conversational Collaboration, Planning-Driven, Test-Driven, and Context-Enhanced Models, thus providing the first comprehensive taxonomy in this domain. Critically, our analysis reveals that successful Vibe Coding depends not merely on agent capabilities but on systematic context engineering, well-established development environments, and human-agent collaborative development models.</p> + https://arxiv.org/abs/2510.12399 + Tue, 14 Oct 2025 11:26:56 +0000 - The African Languages Lab: A Collaborative Approach to Advancing Low-Resource African NLP - https://arxiv.org/abs/2510.05644 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05644.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Sheriff Issaka, Keyi Wang, Yinka Ajibola, Oluwatumininu Samuel-Ipaye, Zhaoyi Zhang, Nicte Aguillon Jimenez, Evans Kofi Agyei, Abraham Lin, Rohan Ramachandran, Sadick Abdul Mumin, Faith Nchifor, Mohammed Shuraim, Lieqi Liu, Erick Rosas Gonzalez, Sylvester Kpei, Jemimah Osei, Carlene Ajeneza, Persis Boateng, Prisca Adwoa Dufie Yeboah, Saadia Gabriel</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> Despite representing nearly one-third of the world's languages, African languages remain critically underserved by modern NLP technologies, with 88\% classified as severely underrepresented or completely ignored in computational linguistics. We present the African Languages Lab (All Lab), a comprehensive research initiative that addresses this technological gap through systematic data collection, model development, and capacity building. Our contributions include: (1) a quality-controlled data collection pipeline, yielding the largest validated African multi-modal speech and text dataset spanning 40 languages with 19 billion tokens of monolingual text and 12,628 hours of aligned speech data; (2) extensive experimental validation demonstrating that our dataset, combined with fine-tuning, achieves substantial improvements over baseline models, averaging +23.69 ChrF++, +0.33 COMET, and +15.34 BLEU points across 31 evaluated languages; and (3) a structured research program that has successfully mentored fifteen early-career researchers, establishing sustainable local capacity. Our comparative evaluation against Google Translate reveals competitive performance in several languages while identifying areas that require continued development.</p> - https://arxiv.org/abs/2510.05644 - Tue, 07 Oct 2025 07:42:52 +0000 + Robot Learning: A Tutorial + https://arxiv.org/abs/2510.12403 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12403.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Francesco Capuano, Caroline Pascal, Adil Zouitine, Thomas Wolf, Michel Aractingi</p><p><b>Upvotes:</b> 73</p><p><b>Summary:</b> Robot learning is at an inflection point, driven by rapid advancements in machine learning and the growing availability of large-scale robotics data. This shift from classical, model-based methods to data-driven, learning-based paradigms is unlocking unprecedented capabilities in autonomous systems. This tutorial navigates the landscape of modern robot learning, charting a course from the foundational principles of Reinforcement Learning and Behavioral Cloning to generalist, language-conditioned models capable of operating across diverse tasks and even robot embodiments. This work is intended as a guide for researchers and practitioners, and our goal is to equip the reader with the conceptual understanding and practical tools necessary to contribute to developments in robot learning, with ready-to-use examples implemented in lerobot.</p> + https://arxiv.org/abs/2510.12403 + Tue, 14 Oct 2025 11:36:46 +0000 - Revisiting Long-context Modeling from Context Denoising Perspective - https://arxiv.org/abs/2510.05862 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05862.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zecheng Tang, Baibei Ji, Juntao Li, Lijun Wu, Haijia Gui, Min Zhang</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> Long-context models (LCMs) have demonstrated great potential in processing long sequences, facilitating many real-world applications. The success of LCMs can be attributed to their ability to locate implicit critical information within the context for further prediction. However, recent research reveals that LCMs are often susceptible to contextual noise, i.e., irrelevant tokens, that can mislead model attention. In this paper, we conduct a fine-grained analysis of the context noise and propose an effective metric, the Integrated Gradient (IG) score, to detect and quantify the noise information within the context. Our findings reveal that even simple mitigation of detected context noise can substantially boost the model's attention on critical tokens and benefit subsequent predictions. Building on this insight, we propose Context Denoising Training (CDT), a straightforward yet effective training strategy that improves attention on critical tokens while reinforcing their influence on model predictions. Extensive experiments across four tasks, under both context window scaling and long-context alignment settings, demonstrate the superiority of CDT. Notably, when trained with CDT, an open-source 8B model can achieve performance (50.92) comparable to GPT-4o (51.00).</p> - https://arxiv.org/abs/2510.05862 - Tue, 07 Oct 2025 12:32:23 +0000 + Advancing End-to-End Pixel Space Generative Modeling via Self-supervised Pre-training + https://arxiv.org/abs/2510.12586 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12586.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiachen Lei, Keli Liu, Julius Berner, Haiming Yu, Hongkai Zheng, Jiahong Wu, Xiangxiang Chu</p><p><b>Upvotes:</b> 104</p><p><b>Summary:</b> Pixel-space generative models are often more difficult to train and generally underperform compared to their latent-space counterparts, leaving a persistent performance and efficiency gap. In this paper, we introduce a novel two-stage training framework that closes this gap for pixel-space diffusion and consistency models. In the first stage, we pre-train encoders to capture meaningful semantics from clean images while aligning them with points along the same deterministic sampling trajectory, which evolves points from the prior to the data distribution. In the second stage, we integrate the encoder with a randomly initialized decoder and fine-tune the complete model end-to-end for both diffusion and consistency models. Our training framework demonstrates strong empirical performance on ImageNet dataset. Specifically, our diffusion model reaches an FID of 2.04 on ImageNet-256 and 2.35 on ImageNet-512 with 75 number of function evaluations (NFE), surpassing prior pixel-space methods by a large margin in both generation quality and efficiency while rivaling leading VAE-based models at comparable training cost. Furthermore, on ImageNet-256, our consistency model achieves an impressive FID of 8.82 in a single sampling step, significantly surpassing its latent-space counterpart. To the best of our knowledge, this marks the first successful training of a consistency model directly on high-resolution images without relying on pre-trained VAEs or diffusion models.</p> + https://arxiv.org/abs/2510.12586 + Tue, 14 Oct 2025 14:41:16 +0000 - MixReasoning: Switching Modes to Think - https://arxiv.org/abs/2510.06052 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06052.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haiquan Lu, Gongfan Fang, Xinyin Ma, Qi Li, Xinchao Wang</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> Reasoning models enhance performance by tackling problems in a step-by-step manner, decomposing them into sub-problems and exploring long chains of thought before producing an answer. However, applying extended reasoning to every step introduces substantial redundancy, as sub-problems vary widely in difficulty and complexity: a small number of pivotal steps are genuinely challenging and decisive for the final answer, while many others only involve straightforward revisions or simple computations. Therefore, a natural idea is to endow reasoning models with the ability to adaptively respond to this variation, rather than treating all steps with the same level of elaboration. To this end, we propose MixReasoning, a framework that dynamically adjusts the depth of reasoning within a single response. The resulting chain of thought then becomes a mixture of detailed reasoning on difficult steps and concise inference on simpler ones. Experiments on GSM8K, MATH-500, and AIME show that MixReasoning shortens reasoning length and substantially improves efficiency without compromising accuracy.</p> - https://arxiv.org/abs/2510.06052 - Tue, 07 Oct 2025 15:46:34 +0000 + Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks + https://arxiv.org/abs/2510.12635 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12635.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuxiang Zhang, Jiangming Shu, Ye Ma, Xueyuan Lin, Shangxi Wu, Jitao Sang</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Large Language Models face challenges in long-horizon agentic tasks as their constrained memory is easily overwhelmed by distracting or irrelevant context. Existing working memory methods typically rely on external, heuristic mechanisms that are decoupled from the agent's core policy. In this work, we reframe working memory management as a learnable, intrinsic capability. We propose a novel framework, Memory-as-Action, where an agent actively manages its working memory by executing explicit editing operations as part of a unified policy. This formulation allows an agent, trained via reinforcement learning, to balance memory curation against long-term task objectives under given resource constraints. However, such memory editing actions break the standard assumption of a continuously growing prefix in LLM interactions, leading to what we call trajectory fractures. These non-prefix changes disrupt the causal continuity required by standard policy gradient methods, making those methods inapplicable. To address this, we propose a new algorithm, Dynamic Context Policy Optimization, which enables stable end-to-end reinforcement learning by segmenting trajectories at memory action points and applying trajectory-level advantages to the resulting action segments. Our results demonstrate that jointly optimizing for task reasoning and memory management in an end-to-end fashion not only reduces overall computational consumption but also improves task performance, driven by adaptive context curation strategies tailored to the model's intrinsic capabilities.</p> + https://arxiv.org/abs/2510.12635 + Tue, 14 Oct 2025 15:29:57 +0000 - ASPO: Asymmetric Importance Sampling Policy Optimization - https://arxiv.org/abs/2510.06062 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06062.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiakang Wang, Runze Liu, Lei Lin, Wenping Hu, Xiu Li, Fuzheng Zhang, Guorui Zhou, Kun Gai</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> Recent Large Language Model (LLM) post-training methods rely on token-level clipping mechanisms during Reinforcement Learning (RL). However, we identify a fundamental flaw in this Outcome-Supervised RL (OSRL) paradigm: the Importance Sampling (IS) ratios of positive-advantage tokens are mismatched, leading to unbalanced token weighting for positive and negative tokens. This mismatch suppresses the update of low-probability tokens while over-amplifying already high-probability ones. To address this, we propose Asymmetric Importance Sampling Policy Optimization (ASPO), which uses a simple yet effective strategy that flips the IS ratios of positive-advantage tokens, aligning their update direction with the learning dynamics of negative ones. AIS further incorporates a soft dual-clipping mechanism to stabilize extreme updates while maintaining gradient flow. Comprehensive experiments on coding and mathematical reasoning benchmarks demonstrate that ASPO significantly mitigates premature convergence, improves training stability, and enhances final performance over strong GRPO-based baselines. Our analysis provides new insights into the role of token-level weighting in OSRL and highlights the critical importance of correcting IS in LLM RL. The code and models of ASPO are available at https://github.com/wizard-III/Archer2.0.</p> - https://arxiv.org/abs/2510.06062 - Tue, 07 Oct 2025 15:54:24 +0000 + ERA: Transforming VLMs into Embodied Agents via Embodied Prior Learning and Online Reinforcement Learning + https://arxiv.org/abs/2510.12693 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12693.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hanyang Chen, Mark Zhao, Rui Yang, Qinwei Ma, Ke Yang, Jiarui Yao, Kangrui Wang, Hao Bai, Zhenhailong Wang, Rui Pan, Mengchao Zhang, Jose Barreiros, Aykut Onol, ChengXiang Zhai, Heng Ji, Manling Li, Huan Zhang, Tong Zhang</p><p><b>Upvotes:</b> 24</p><p><b>Summary:</b> Recent advances in embodied AI highlight the potential of vision language models (VLMs) as agents capable of perception, reasoning, and interaction in complex environments. However, top-performing systems rely on large-scale models that are costly to deploy, while smaller VLMs lack the necessary knowledge and skills to succeed. To bridge this gap, we present Embodied Reasoning Agent (ERA), a two-stage framework that integrates prior knowledge learning and online reinforcement learning (RL). The first stage, Embodied Prior Learning, distills foundational knowledge from three types of data: (1) Trajectory-Augmented Priors, which enrich existing trajectory data with structured reasoning generated by stronger models; (2) Environment-Anchored Priors, which provide in-environment knowledge and grounding supervision; and (3) External Knowledge Priors, which transfer general knowledge from out-of-environment datasets. In the second stage, we develop an online RL pipeline that builds on these priors to further enhance agent performance. To overcome the inherent challenges in agent RL, including long horizons, sparse rewards, and training instability, we introduce three key designs: self-summarization for context management, dense reward shaping, and turn-level policy optimization. Extensive experiments on both high-level planning (EB-ALFRED) and low-level control (EB-Manipulation) tasks demonstrate that ERA-3B surpasses both prompting-based large models and previous training-based baselines. Specifically, it achieves overall improvements of 8.4\% on EB-ALFRED and 19.4\% on EB-Manipulation over GPT-4o, and exhibits strong generalization to unseen tasks. Overall, ERA offers a practical path toward scalable embodied intelligence, providing methodological insights for future embodied AI systems.</p> + https://arxiv.org/abs/2510.12693 + Tue, 14 Oct 2025 16:25:46 +0000 - Discrete Diffusion Models with MLLMs for Unified Medical Multimodal Generation - https://arxiv.org/abs/2510.06131 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06131.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiawei Mao, Yuhan Wang, Lifeng Chen, Can Zhao, Yucheng Tang, Dong Yang, Liangqiong Qu, Daguang Xu, Yuyin Zhou</p><p><b>Upvotes:</b> 10</p><p><b>Summary:</b> Recent advances in generative medical models are constrained by modality-specific scenarios that hinder the integration of complementary evidence from imaging, pathology, and clinical notes. This fragmentation limits their evolution into foundation models that can learn and reason across the full spectrum of biomedical data. We propose MeDiM, the first medical discrete diffusion model that learns shared distributions across modalities without modality-specific components. MeDiM unifies multiple generative tasks: translating between images and text, and jointly producing image-report pairs across domains in response to prompts. Built on a discrete diffusion framework, MeDiM bridges vision and language representations through a shared probabilistic space. To enable unified and flexible medical generation, we employ a multimodal large language model (MLLM) as the diffusion backbone, leveraging its prior knowledge and cross-modal reasoning. Two key designs are introduced: (1) removing the causal attention mask for bidirectional context, and (2) injecting continuous timestep embeddings for diffusion awareness. Experiments demonstrate high-fidelity medical generation (FID 16.60 on MIMIC-CXR and FID 24.19 on PathGen) and accurate report generation (METEOR 0.2650 and 0.2580). Jointly generated image-report pairs further enhance downstream performance (plus6.43 percent BLEU-1, plus18.57 percent BLEU-2, plus31.58 percent BLEU-3, plus4.80 percent METEOR), showing that MeDiM supports coherent and clinically grounded multimodal outputs.</p> - https://arxiv.org/abs/2510.06131 - Tue, 07 Oct 2025 17:06:57 +0000 + FlashVSR: Towards Real-Time Diffusion-Based Streaming Video Super-Resolution + https://arxiv.org/abs/2510.12747 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12747.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Junhao Zhuang, Shi Guo, Xin Cai, Xiaohui Li, Yihao Liu, Chun Yuan, Tianfan Xue</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> Diffusion models have recently advanced video restoration, but applying them to real-world video super-resolution (VSR) remains challenging due to high latency, prohibitive computation, and poor generalization to ultra-high resolutions. Our goal in this work is to make diffusion-based VSR practical by achieving efficiency, scalability, and real-time performance. To this end, we propose FlashVSR, the first diffusion-based one-step streaming framework towards real-time VSR. FlashVSR runs at approximately 17 FPS for 768x1408 videos on a single A100 GPU by combining three complementary innovations: (i) a train-friendly three-stage distillation pipeline that enables streaming super-resolution, (ii) locality-constrained sparse attention that cuts redundant computation while bridging the train-test resolution gap, and (iii) a tiny conditional decoder that accelerates reconstruction without sacrificing quality. To support large-scale training, we also construct VSR-120K, a new dataset with 120k videos and 180k images. Extensive experiments show that FlashVSR scales reliably to ultra-high resolutions and achieves state-of-the-art performance with up to 12x speedup over prior one-step diffusion VSR models. We will release the code, pretrained models, and dataset to foster future research in efficient diffusion-based VSR.</p> + https://arxiv.org/abs/2510.12747 + Tue, 14 Oct 2025 17:25:54 +0000 - Mixing Mechanisms: How Language Models Retrieve Bound Entities In-Context - https://arxiv.org/abs/2510.06182 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06182.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yoav Gur-Arieh, Mor Geva, Atticus Geiger</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> A key component of in-context reasoning is the ability of language models (LMs) to bind entities for later retrieval. For example, an LM might represent "Ann loves pie" by binding "Ann" to "pie", allowing it to later retrieve "Ann" when asked "Who loves pie?" Prior research on short lists of bound entities found strong evidence that LMs implement such retrieval via a positional mechanism, where "Ann" is retrieved based on its position in context. In this work, we find that this mechanism generalizes poorly to more complex settings; as the number of bound entities in context increases, the positional mechanism becomes noisy and unreliable in middle positions. To compensate for this, we find that LMs supplement the positional mechanism with a lexical mechanism (retrieving "Ann" using its bound counterpart "pie") and a reflexive mechanism (retrieving "Ann" through a direct pointer). Through extensive experiments on nine models and ten binding tasks, we uncover a consistent pattern in how LMs mix these mechanisms to drive model behavior. We leverage these insights to develop a causal model combining all three mechanisms that estimates next token distributions with 95% agreement. Finally, we show that our model generalizes to substantially longer inputs of open-ended text interleaved with entity groups, further demonstrating the robustness of our findings in more natural settings. Overall, our study establishes a more complete picture of how LMs bind and retrieve entities in-context.</p> - https://arxiv.org/abs/2510.06182 - Tue, 07 Oct 2025 17:44:30 +0000 + Dr.LLM: Dynamic Layer Routing in LLMs + https://arxiv.org/abs/2510.12773 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12773.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ahmed Heakl, Martin Gubri, Salman Khan, Sangdoo Yun, Seong Joon Oh</p><p><b>Upvotes:</b> 29</p><p><b>Summary:</b> Large Language Models (LLMs) process every token through all layers of a transformer stack, causing wasted computation on simple queries and insufficient flexibility for harder ones that need deeper reasoning. Adaptive-depth methods can improve efficiency, but prior approaches rely on costly inference-time search, architectural changes, or large-scale retraining, and in practice often degrade accuracy despite efficiency gains. We introduce Dr.LLM, Dynamic routing of Layers for LLMs, a retrofittable framework that equips pretrained models with lightweight per-layer routers deciding to skip, execute, or repeat a block. Routers are trained with explicit supervision: using Monte Carlo Tree Search (MCTS), we derive high-quality layer configurations that preserve or improve accuracy under a compute budget. Our design, windowed pooling for stable routing, focal loss with class balancing, and bottleneck MLP routers, ensures robustness under class imbalance and long sequences. On ARC (logic) and DART (math), Dr.LLM improves accuracy by up to +3.4%p while saving 5 layers per example on average. Routers generalize to out-of-domain tasks (MMLU, GSM8k, AIME, TruthfulQA, SQuADv2, GPQA, PIQA, AGIEval) with only 0.85% accuracy drop while retaining efficiency, and outperform prior routing methods by up to +7.7%p. Overall, Dr.LLM shows that explicitly supervised routers retrofit frozen LLMs for budget-aware, accuracy-driven inference without altering base weights.</p> + https://arxiv.org/abs/2510.12773 + Tue, 14 Oct 2025 17:51:26 +0000 - ShapeGen4D: Towards High Quality 4D Shape Generation from Videos - https://arxiv.org/abs/2510.06208 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06208.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiraphon Yenphraphai, Ashkan Mirzaei, Jianqi Chen, Jiaxu Zou, Sergey Tulyakov, Raymond A. Yeh, Peter Wonka, Chaoyang Wang</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Video-conditioned 4D shape generation aims to recover time-varying 3D geometry and view-consistent appearance directly from an input video. In this work, we introduce a native video-to-4D shape generation framework that synthesizes a single dynamic 3D representation end-to-end from the video. Our framework introduces three key components based on large-scale pre-trained 3D models: (i) a temporal attention that conditions generation on all frames while producing a time-indexed dynamic representation; (ii) a time-aware point sampling and 4D latent anchoring that promote temporally consistent geometry and texture; and (iii) noise sharing across frames to enhance temporal stability. Our method accurately captures non-rigid motion, volume changes, and even topological transitions without per-frame optimization. Across diverse in-the-wild videos, our method improves robustness and perceptual fidelity and reduces failure modes compared with the baselines.</p> - https://arxiv.org/abs/2510.06208 - Tue, 07 Oct 2025 17:58:11 +0000 + SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models + https://arxiv.org/abs/2510.12784 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12784.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Weiyang Jin, Yuwei Niu, Jiaqi Liao, Chengqi Duan, Aoxue Li, Shenghua Gao, Xihui Liu</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> Recently, remarkable progress has been made in Unified Multimodal Models (UMMs), which integrate vision-language generation and understanding capabilities within a single framework. However, a significant gap exists where a model's strong visual understanding often fails to transfer to its visual generation. A model might correctly understand an image based on user instructions, yet be unable to generate a faithful image from text prompts. This phenomenon directly raises a compelling question: Can a model achieve self-improvement by using its understanding module to reward its generation module? To bridge this gap and achieve self-improvement, we introduce SRUM, a self-rewarding post-training framework that can be directly applied to existing UMMs of various designs. SRUM creates a feedback loop where the model's own understanding module acts as an internal ``evaluator'', providing corrective signals to improve its generation module, without requiring additional human-labeled data. To ensure this feedback is comprehensive, we designed a global-local dual reward system. To tackle the inherent structural complexity of images, this system offers multi-scale guidance: a global reward ensures the correctness of the overall visual semantics and layout, while a local reward refines fine-grained, object-level fidelity. SRUM leads to powerful capabilities and shows strong generalization, boosting performance on T2I-CompBench from 82.18 to 88.37 and on T2I-ReasonBench from 43.82 to 46.75. Overall, our work establishes a powerful new paradigm for enabling a UMMs' understanding module to guide and enhance its own generation via self-rewarding.</p> + https://arxiv.org/abs/2510.12784 + Tue, 14 Oct 2025 17:56:11 +0000 - Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding - https://arxiv.org/abs/2510.06308 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06308.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yi Xin, Qi Qin, Siqi Luo, Kaiwen Zhu, Juncheng Yan, Yan Tai, Jiayi Lei, Yuewen Cao, Keqi Wang, Yibin Wang, Jinbin Bai, Qian Yu, Dengyang Jiang, Yuandong Pu, Haoxing Chen, Le Zhuo, Junjun He, Gen Luo, Tianbin Li, Ming Hu, Jin Ye, Shenglong Ye, Bo Zhang, Chang Xu, Wenhai Wang, Hongsheng Li, Guangtao Zhai, Tianfan Xue, Bin Fu, Xiaohong Liu, Yu Qiao, Yihao Liu</p><p><b>Upvotes:</b> 47</p><p><b>Summary:</b> We introduce Lumina-DiMOO, an open-source foundational model for seamless multi-modal generation and understanding. Lumina-DiMOO sets itself apart from prior unified models by utilizing a fully discrete diffusion modeling to handle inputs and outputs across various modalities. This innovative approach allows Lumina-DiMOO to achieve higher sampling efficiency compared to previous autoregressive (AR) or hybrid AR-Diffusion paradigms and adeptly support a broad spectrum of multi-modal tasks, including text-to-image generation, image-to-image generation (e.g., image editing, subject-driven generation, and image inpainting, etc.), as well as image understanding. Lumina-DiMOO achieves state-of-the-art performance on multiple benchmarks, surpassing existing open-source unified multi-modal models. To foster further advancements in multi-modal and discrete diffusion model research, we release our code and checkpoints to the community. Project Page: https://synbol.github.io/Lumina-DiMOO.</p> - https://arxiv.org/abs/2510.06308 - Tue, 07 Oct 2025 17:59:20 +0000 + UniFusion: Vision-Language Model as Unified Encoder in Image Generation + https://arxiv.org/abs/2510.12789 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12789.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kevin Li, Manuel Brack, Sudeep Katakol, Hareesh Ravi, Ajinkya Kale</p><p><b>Upvotes:</b> 15</p><p><b>Summary:</b> Although recent advances in visual generation have been remarkable, most existing architectures still depend on distinct encoders for images and text. This separation constrains diffusion models' ability to perform cross-modal reasoning and knowledge transfer. Prior attempts to bridge this gap often use the last layer information from VLM, employ multiple visual encoders, or train large unified models jointly for text and image generation, which demands substantial computational resources and large-scale data, limiting its accessibility.We present UniFusion, a diffusion-based generative model conditioned on a frozen large vision-language model (VLM) that serves as a unified multimodal encoder. At the core of UniFusion is the Layerwise Attention Pooling (LAP) mechanism that extracts both high level semantics and low level details from text and visual tokens of a frozen VLM to condition a diffusion generative model. We demonstrate that LAP outperforms other shallow fusion architectures on text-image alignment for generation and faithful transfer of visual information from VLM to the diffusion model which is key for editing. We propose VLM-Enabled Rewriting Injection with Flexibile Inference (VERIFI), which conditions a diffusion transformer (DiT) only on the text tokens generated by the VLM during in-model prompt rewriting. VERIFI combines the alignment of the conditioning distribution with the VLM's reasoning capabilities for increased capabilities and flexibility at inference. In addition, finetuning on editing task not only improves text-image alignment for generation, indicative of cross-modality knowledge transfer, but also exhibits tremendous generalization capabilities. Our model when trained on single image editing, zero-shot generalizes to multiple image references further motivating the unified encoder design of UniFusion.</p> + https://arxiv.org/abs/2510.12789 + Tue, 14 Oct 2025 17:57:56 +0000 - TaTToo: Tool-Grounded Thinking PRM for Test-Time Scaling in Tabular Reasoning - https://arxiv.org/abs/2510.06217 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06217.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiaru Zou, Soumya Roy, Vinay Kumar Verma, Ziyi Wang, David Wipf, Pan Lu, Sumit Negi, James Zou, Jingrui He</p><p><b>Upvotes:</b> 59</p><p><b>Summary:</b> Process Reward Models (PRMs) have recently emerged as a powerful framework for enhancing the reasoning capabilities of large reasoning models (LRMs), particularly in the context of test-time scaling (TTS). However, their potential for supervising LRMs on tabular reasoning domains remains underexplored. Through detailed empirical analyses, we identify that existing PRMs, though widely adopted for supervising text-only reasoning steps, struggle with table-specific operations such as sub-table retrieval and schema interaction, leading to critical performance bottlenecks. To address this limitation, we propose TaTToo, a novel table-grounded PRM framework that (i) reasons explicitly over tabular reasoning steps and (ii) integrates tool-based verification to provide precise reward supervision. Concretely, we first design a scalable data curation pipeline that constructs over 60k high-quality step-level annotations by integrating table verification rationales with tool-based executions. Building on the collected data, we train TaTToo with a dual-stage paradigm: cold-start supervised fine-tuning to capture tool-use reasoning patterns, followed by reinforcement learning with tool-grounded reward shaping to align our model with table-based verification. We provide a comprehensive evaluation of the policy improvement induced by our newly designed PRM. Across 5 challenging tabular reasoning benchmarks covering numerical reasoning, fact-checking, and data analysis, TaTToo improves downstream policy LRMs by 30.9% at inference, surpasses strong PRM baselines such as Qwen-2.5-Math-PRM-72B with only 8B parameters, and demonstrates strong generalizability across diverse TTS strategies.</p> - https://arxiv.org/abs/2510.06217 - Tue, 07 Oct 2025 17:59:41 +0000 + Detect Anything via Next Point Prediction + https://arxiv.org/abs/2510.12798 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12798.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qing Jiang, Junan Huo, Xingyu Chen, Yuda Xiong, Zhaoyang Zeng, Yihao Chen, Tianhe Ren, Junzhi Yu, Lei Zhang</p><p><b>Upvotes:</b> 42</p><p><b>Summary:</b> Object detection has long been dominated by traditional coordinate regression-based models, such as YOLO, DETR, and Grounding DINO. Although recent efforts have attempted to leverage MLLMs to tackle this task, they face challenges like low recall rate, duplicate predictions, coordinate misalignment, etc. In this work, we bridge this gap and propose Rex-Omni, a 3B-scale MLLM that achieves state-of-the-art object perception performance. On benchmarks like COCO and LVIS, Rex-Omni attains performance comparable to or exceeding regression-based models (e.g., DINO, Grounding DINO) in a zero-shot setting. This is enabled by three key designs: 1) Task Formulation: we use special tokens to represent quantized coordinates from 0 to 999, reducing the model's learning difficulty and improving token efficiency for coordinate prediction; 2) Data Engines: we construct multiple data engines to generate high-quality grounding, referring, and pointing data, providing semantically rich supervision for training; \3) Training Pipelines: we employ a two-stage training process, combining supervised fine-tuning on 22 million data with GRPO-based reinforcement post-training. This RL post-training leverages geometry-aware rewards to effectively bridge the discrete-to-continuous coordinate prediction gap, improve box accuracy, and mitigate undesirable behaviors like duplicate predictions that stem from the teacher-guided nature of the initial SFT stage. Beyond conventional detection, Rex-Omni's inherent language understanding enables versatile capabilities such as object referring, pointing, visual prompting, GUI grounding, spatial referring, OCR and key-pointing, all systematically evaluated on dedicated benchmarks. We believe that Rex-Omni paves the way for more versatile and language-aware visual perception systems.</p> + https://arxiv.org/abs/2510.12798 + Tue, 14 Oct 2025 17:59:54 +0000 - Human3R: Everyone Everywhere All at Once - https://arxiv.org/abs/2510.06219 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06219.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yue Chen, Xingyu Chen, Yuxuan Xue, Anpei Chen, Yuliang Xiu, Gerard Pons-Moll</p><p><b>Upvotes:</b> 8</p><p><b>Summary:</b> We present Human3R, a unified, feed-forward framework for online 4D human-scene reconstruction, in the world frame, from casually captured monocular videos. Unlike previous approaches that rely on multi-stage pipelines, iterative contact-aware refinement between humans and scenes, and heavy dependencies, e.g., human detection, depth estimation, and SLAM pre-processing, Human3R jointly recovers global multi-person SMPL-X bodies ("everyone"), dense 3D scene ("everywhere"), and camera trajectories in a single forward pass ("all-at-once"). Our method builds upon the 4D online reconstruction model CUT3R, and uses parameter-efficient visual prompt tuning, to strive to preserve CUT3R's rich spatiotemporal priors, while enabling direct readout of multiple SMPL-X bodies. Human3R is a unified model that eliminates heavy dependencies and iterative refinement. After being trained on the relatively small-scale synthetic dataset BEDLAM for just one day on one GPU, it achieves superior performance with remarkable efficiency: it reconstructs multiple humans in a one-shot manner, along with 3D scenes, in one stage, at real-time speed (15 FPS) with a low memory footprint (8 GB). Extensive experiments demonstrate that Human3R delivers state-of-the-art or competitive performance across tasks, including global human motion estimation, local human mesh recovery, video depth estimation, and camera pose estimation, with a single unified model. We hope that Human3R will serve as a simple yet strong baseline, be easily extended for downstream applications.Code available in https://fanegg.github.io/Human3R</p> - https://arxiv.org/abs/2510.06219 - Tue, 07 Oct 2025 17:59:52 +0000 + DeepMMSearch-R1: Empowering Multimodal LLMs in Multimodal Web Search + https://arxiv.org/abs/2510.12801 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.12801.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kartik Narayan, Yang Xu, Tian Cao, Kavya Nerella, Vishal M. Patel, Navid Shiee, Peter Grasch, Chao Jia, Yinfei Yang, Zhe Gan</p><p><b>Upvotes:</b> 12</p><p><b>Summary:</b> Multimodal Large Language Models (MLLMs) in real-world applications require access to external knowledge sources and must remain responsive to the dynamic and ever-changing real-world information in order to address information-seeking and knowledge-intensive user queries. Existing approaches, such as retrieval augmented generation (RAG) methods, search agents, and search equipped MLLMs, often suffer from rigid pipelines, excessive search calls, and poorly constructed search queries, which result in inefficiencies and suboptimal outcomes. To address these limitations, we present DeepMMSearch-R1, the first multimodal LLM capable of performing on-demand, multi-turn web searches and dynamically crafting queries for both image and text search tools. Specifically, DeepMMSearch-R1 can initiate web searches based on relevant crops of the input image making the image search more effective, and can iteratively adapt text search queries based on retrieved information, thereby enabling self-reflection and self-correction. Our approach relies on a two-stage training pipeline: a cold start supervised finetuning phase followed by an online reinforcement learning optimization. For training, we introduce DeepMMSearchVQA, a novel multimodal VQA dataset created through an automated pipeline intermixed with real-world information from web search tools. This dataset contains diverse, multi-hop queries that integrate textual and visual information, teaching the model when to search, what to search for, which search tool to use and how to reason over the retrieved information. We conduct extensive experiments across a range of knowledge-intensive benchmarks to demonstrate the superiority of our approach. Finally, we analyze the results and provide insights that are valuable for advancing multimodal web-search.</p> + https://arxiv.org/abs/2510.12801 + Tue, 14 Oct 2025 17:59:58 +0000 - The Markovian Thinker - https://arxiv.org/abs/2510.06557 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06557.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Milad Aghajohari, Kamran Chitsaz, Amirhossein Kazemnejad, Sarath Chandar, Alessandro Sordoni, Aaron Courville, Siva Reddy</p><p><b>Upvotes:</b> 20</p><p><b>Summary:</b> Reinforcement learning (RL) has recently become a strong recipe for training reasoning LLMs that produce long chains of thought (LongCoT). Yet the standard RL "thinking environment", where the state is the prompt plus all prior reasoning tokens, makes the state unbounded and forces attention-based policies to pay quadratic compute as thoughts lengthen. We revisit the environment itself. We propose Markovian Thinking, a paradigm in which the policy advances reasoning while conditioning on a constant-size state, decoupling thinking length from context size. As an immediate consequence this yields linear compute with constant memory. We instantiate this idea with Delethink, an RL environment that structures reasoning into fixed-size chunks. Within each chunk, the model thinks as usual; at the boundary, the environment resets the context and reinitializes the prompt with a short carryover. Through RL, the policy learns to write a textual state near the end of each chunk sufficient for seamless continuation of reasoning after reset. Trained in this environment, an R1-Distill 1.5B model reasons in 8K-token chunks yet thinks up to 24K tokens, matching or surpassing LongCoT-RL trained with a 24K budget. With test-time scaling, Delethink continues to improve where LongCoT plateaus. The effect of linear compute is substantial: we empirically estimate at 96K average thinking length LongCoT-RL costs 27 H100-months vs. 7 for Delethink. Analysis at RL initialization shows off-the-shelf reasoning models (1.5B-120B) often sample Markovian traces zero-shot across diverse benchmarks, providing positive samples that make RL effective at scale. Our results show that redesigning the thinking environment is a powerful lever: it enables very long reasoning without quadratic overhead and opens a path toward efficient, scalable reasoning LLMs.</p> - https://arxiv.org/abs/2510.06557 - Wed, 08 Oct 2025 01:18:13 +0000 + LLM-guided Hierarchical Retrieval + https://arxiv.org/abs/2510.13217 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13217.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Nilesh Gupta, Wei-Cheng Chang, Ngot Bui, Cho-Jui Hsieh, Inderjit S. Dhillon</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Modern IR systems are increasingly tasked with answering complex, multi-faceted queries that require deep reasoning rather than simple keyword or semantic matching. While LLM-based IR has shown great promise, the prevailing retrieve-then-rerank paradigm inherits the limitations of embedding-based retrieval; parametric generative approaches are difficult to update with new information; and long-context methods that place the entire corpus in context are computationally infeasible for large document collections. To address these challenges, we introduce LATTICE, a hierarchical retrieval framework that enables an LLM to reason over and navigate large corpora with logarithmic search complexity by imposing a semantic tree structure on the corpus. Our approach consists of two stages: (1) an offline phase that organizes the corpus into a semantic hierarchy via either a bottom-up agglomerative strategy or a top-down divisive strategy using multi-level summaries and (2) an online traversal phase where a search LLM navigates this tree. A central challenge in such LLM-guided search is that the model's relevance judgments are noisy, context-dependent, and unaware of the hierarchy, making cross-branch and cross-level comparisons difficult. To overcome this, we propose a traversal algorithm that estimates calibrated latent relevance scores from local LLM outputs and aggregates them into a global path relevance metric. Our training-free framework achieves state-of-the-art zero-shot performance on the reasoning-intensive BRIGHT benchmark, demonstrating up to 9% improvement in Recall@100 and 5% in nDCG@10 over the next best zero-shot baseline. Furthermore, compared to the fine-tuned SOTA method DIVER-v2, LATTICE attains comparable results on BRIGHT subsets that use a static corpus for evaluation.</p> + https://arxiv.org/abs/2510.13217 + Wed, 15 Oct 2025 07:05:17 +0000 - Ming-UniVision: Joint Image Understanding and Generation with a Unified Continuous Tokenizer - https://arxiv.org/abs/2510.06590 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06590.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ziyuan Huang, DanDan Zheng, Cheng Zou, Rui Liu, Xiaolong Wang, Kaixiang Ji, Weilong Chai, Jianxin Sun, Libin Wang, Yongjie Lv, Taozhi Huang, Jiajia Liu, Qingpei Guo, Ming Yang, Jingdong Chen, Jun Zhou</p><p><b>Upvotes:</b> 66</p><p><b>Summary:</b> Visual tokenization remains a core challenge in unifying visual understanding and generation within the autoregressive paradigm. Existing methods typically employ tokenizers in discrete latent spaces to align with the tokens from large language models, where the quantization errors can limit semantic expressiveness and degrade the capability of vision-language understanding. To address this, we introduce MingTok, a new family of visual tokenizers with a continuous latent space, for unified autoregressive generation and understanding. While understanding tasks favor discriminative high-dimensional features, generation tasks prefer compact low-level codes. Thus, to reconcile these competing demands, MingTok adopts a three-stage sequential architecture involving low-level encoding, semantic expansion, and visual reconstruction. Built on top of it, Ming-UniVision eliminates the need for task-specific visual representations, and unifies diverse vision-language tasks under a single autoregrsssive prediction paradigm. By formulating both understanding and generation as next-token prediction in a shared continuous space, it seamlessly supports multi-round, in-context tasks such as iterative understanding, generation and editing. Empirically, we find that using a unified continuous visual representation reconciles the competing requirements on the tokenizers by the understanding and generation tasks, thereby leading to state-of-the-art level performance across both domains. We hope our findings will facilitate unified visual tokenization in the continuous domain. Inference code and model weights are released to benefit community.</p> - https://arxiv.org/abs/2510.06590 - Wed, 08 Oct 2025 02:50:14 +0000 + Universal Image Restoration Pre-training via Masked Degradation Classification + https://arxiv.org/abs/2510.13282 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13282.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> JiaKui Hu, Zhengjian Yao, Lujia Jin, Yinghao Chen, Yanye Lu</p><p><b>Upvotes:</b> 10</p><p><b>Summary:</b> This study introduces a Masked Degradation Classification Pre-Training method (MaskDCPT), designed to facilitate the classification of degradation types in input images, leading to comprehensive image restoration pre-training. Unlike conventional pre-training methods, MaskDCPT uses the degradation type of the image as an extremely weak supervision, while simultaneously leveraging the image reconstruction to enhance performance and robustness. MaskDCPT includes an encoder and two decoders: the encoder extracts features from the masked low-quality input image. The classification decoder uses these features to identify the degradation type, whereas the reconstruction decoder aims to reconstruct a corresponding high-quality image. This design allows the pre-training to benefit from both masked image modeling and contrastive learning, resulting in a generalized representation suited for restoration tasks. Benefit from the straightforward yet potent MaskDCPT, the pre-trained encoder can be used to address universal image restoration and achieve outstanding performance. Implementing MaskDCPT significantly improves performance for both convolution neural networks (CNNs) and Transformers, with a minimum increase in PSNR of 3.77 dB in the 5D all-in-one restoration task and a 34.8% reduction in PIQE compared to baseline in real-world degradation scenarios. It also emergences strong generalization to previously unseen degradation types and levels. In addition, we curate and release the UIR-2.5M dataset, which includes 2.5 million paired restoration samples across 19 degradation types and over 200 degradation levels, incorporating both synthetic and real-world data. The dataset, source code, and models are available at https://github.com/MILab-PKU/MaskDCPT.</p> + https://arxiv.org/abs/2510.13282 + Wed, 15 Oct 2025 08:30:15 +0000 - DreamOmni2: Multimodal Instruction-based Editing and Generation - https://arxiv.org/abs/2510.06679 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06679.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Bin Xia, Bohao Peng, Yuechen Zhang, Junjia Huang, Jiyang Liu, Jingyao Li, Haoru Tan, Sitong Wu, Chengyao Wang, Yitong Wang, Xinglong Wu, Bei Yu, Jiaya Jia</p><p><b>Upvotes:</b> 50</p><p><b>Summary:</b> Recent advancements in instruction-based image editing and subject-driven generation have garnered significant attention, yet both tasks still face limitations in meeting practical user needs. Instruction-based editing relies solely on language instructions, which often fail to capture specific editing details, making reference images necessary. Meanwhile, subject-driven generation is limited to combining concrete objects or people, overlooking broader, abstract concepts. To address these challenges, we propose two novel tasks: multimodal instruction-based editing and generation. These tasks support both text and image instructions and extend the scope to include both concrete and abstract concepts, greatly enhancing their practical applications. We introduce DreamOmni2, tackling two primary challenges: data creation and model framework design. Our data synthesis pipeline consists of three steps: (1) using a feature mixing method to create extraction data for both abstract and concrete concepts, (2) generating multimodal instruction-based editing training data using the editing and extraction models, and (3) further applying the extraction model to create training data for multimodal instruction-based editing. For the framework, to handle multi-image input, we propose an index encoding and position encoding shift scheme, which helps the model distinguish images and avoid pixel confusion. Additionally, we introduce joint training with the VLM and our generation/editing model to better process complex instructions. In addition, we have proposed comprehensive benchmarks for these two new tasks to drive their development. Experiments show that DreamOmni2 has achieved impressive results. Models and codes will be released.</p> - https://arxiv.org/abs/2510.06679 - Wed, 08 Oct 2025 06:07:14 +0000 + UniMoE-Audio: Unified Speech and Music Generation with Dynamic-Capacity MoE + https://arxiv.org/abs/2510.13344 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13344.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhenyu Liu, Yunxin Li, Xuanyu Zhang, Qixun Teng, Shenyuan Jiang, Xinyu Chen, Haoyuan Shi, Jinchao Li, Qi Wang, Haolan Chen, Fanbo Meng, Mingjun Zhao, Yu Xu, Yancheng He, Baotian Hu, Min Zhang</p><p><b>Upvotes:</b> 59</p><p><b>Summary:</b> Recent advances in unified multimodal models indicate a clear trend towards comprehensive content generation. However, the auditory domain remains a significant challenge, with music and speech often developed in isolation, hindering progress towards universal audio synthesis. This separation stems from inherent task conflicts and severe data imbalances, which impede the development of a truly unified audio generation model. To address this challenge, we propose UniMoE-Audio, a unified speech and music generation model within a novel Dynamic-Capacity Mixture-of-Experts (MoE) framework. Architecturally, UniMoE-Audio introduces a Top-P routing strategy for dynamic expert number allocation, and a hybrid expert design comprising routed experts for domain-specific knowledge, shared experts for domain-agnostic features, and null experts for adaptive computation skipping. To tackle data imbalance, we introduce a three-stage training curriculum: 1) Independent Specialist Training leverages original datasets to instill domain-specific knowledge into each "proto-expert" without interference; 2) MoE Integration and Warmup incorporates these specialists into the UniMoE-Audio architecture, warming up the gate module and shared expert using a subset of balanced dataset; and 3) Synergistic Joint Training trains the entire model end-to-end on the fully balanced dataset, fostering enhanced cross-domain synergy. Extensive experiments show that UniMoE-Audio not only achieves state-of-the-art performance on major speech and music generation benchmarks, but also demonstrates superior synergistic learning, mitigating the performance degradation typically seen in naive joint training. Our findings highlight the substantial potential of specialized MoE architecture and curated training strategies in advancing the field of universal audio generation. Homepage: https://mukioxun.github.io/Uni-MoE-site/home.html</p> + https://arxiv.org/abs/2510.13344 + Wed, 15 Oct 2025 09:30:25 +0000 - RLinf-VLA: A Unified and Efficient Framework for VLA+RL Training - https://arxiv.org/abs/2510.06710 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06710.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hongzhi Zang, Mingjie Wei, Si Xu, Yongji Wu, Zhen Guo, Yuanqing Wang, Hao Lin, Liangzhi Shi, Yuqing Xie, Zhexuan Xu, Zhihao Liu, Kang Chen, Wenhao Tang, Quanlu Zhang, Weinan Zhang, Chao Yu, Yu Wang</p><p><b>Upvotes:</b> 31</p><p><b>Summary:</b> Recent progress in vision and language foundation models has significantly advanced multimodal understanding, reasoning, and generation, inspiring a surge of interest in extending such capabilities to embodied settings through vision-language-action (VLA) models. Yet, most VLA models are still trained with supervised fine-tuning (SFT), which struggles to generalize under distribution shifts due to error accumulation. Reinforcement learning (RL) offers a promising alternative by directly optimizing task performance through interaction, but existing attempts remain fragmented and lack a unified platform for fair and systematic comparison across model architectures and algorithmic designs. To address this gap, we introduce RLinf-VLA, a unified and efficient framework for scalable RL training of VLA models. The system adopts a highly flexible resource allocation design that addresses the challenge of integrating rendering, training, and inference in RL+VLA training. In particular, for GPU-parallelized simulators, RLinf-VLA implements a novel hybrid fine-grained pipeline allocation mode, achieving a 1.61x-1.88x speedup in training. Through a unified interface, RLinf-VLA seamlessly supports diverse VLA architectures (e.g., OpenVLA, OpenVLA-OFT), multiple RL algorithms (e.g., PPO, GRPO), and various simulators (e.g., ManiSkill, LIBERO). In simulation, a unified model achieves 98.11\% across 130 LIBERO tasks and 97.66\% across 25 ManiSkill tasks. Beyond empirical performance, our study distills a set of best practices for applying RL to VLA training and sheds light on emerging patterns in this integration. Furthermore, we present preliminary deployment on a real-world Franka robot, where RL-trained policies exhibit stronger generalization than those trained with SFT. We envision RLinf-VLA as a foundation to accelerate and standardize research on embodied intelligence.</p> - https://arxiv.org/abs/2510.06710 - Wed, 08 Oct 2025 07:05:13 +0000 + UniME-V2: MLLM-as-a-Judge for Universal Multimodal Embedding Learning + https://arxiv.org/abs/2510.13515 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13515.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Tiancheng Gu, Kaicheng Yang, Kaichen Zhang, Xiang An, Ziyong Feng, Yueyi Zhang, Weidong Cai, Jiankang Deng, Lidong Bing</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Universal multimodal embedding models are foundational to various tasks. Existing approaches typically employ in-batch negative mining by measuring the similarity of query-candidate pairs. However, these methods often struggle to capture subtle semantic differences among candidates and lack diversity in negative samples. Moreover, the embeddings exhibit limited discriminative ability in distinguishing false and hard negatives. In this paper, we leverage the advanced understanding capabilities of MLLMs to enhance representation learning and present a novel Universal Multimodal Embedding (UniME-V2) model. Our approach first constructs a potential hard negative set through global retrieval. We then introduce the MLLM-as-a-Judge mechanism, which utilizes MLLMs to assess the semantic alignment of query-candidate pairs and generate soft semantic matching scores. These scores serve as a foundation for hard negative mining, mitigating the impact of false negatives and enabling the identification of diverse, high-quality hard negatives. Furthermore, the semantic matching scores are used as soft labels to mitigate the rigid one-to-one mapping constraint. By aligning the similarity matrix with the soft semantic matching score matrix, the model learns semantic distinctions among candidates, significantly enhancing its discriminative capacity. To further improve performance, we propose UniME-V2-Reranker, a reranking model trained on our mined hard negatives through a joint pairwise and listwise optimization approach. We conduct comprehensive experiments on the MMEB benchmark and multiple retrieval tasks, demonstrating that our method achieves state-of-the-art performance on average across all tasks.</p> + https://arxiv.org/abs/2510.13515 + Wed, 15 Oct 2025 13:07:00 +0000 - OBS-Diff: Accurate Pruning For Diffusion Models in One-Shot - https://arxiv.org/abs/2510.06751 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06751.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Junhan Zhu, Hesong Wang, Mingluo Su, Zefang Wang, Huan Wang</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Large-scale text-to-image diffusion models, while powerful, suffer from prohibitive computational cost. Existing one-shot network pruning methods can hardly be directly applied to them due to the iterative denoising nature of diffusion models. To bridge the gap, this paper presents OBS-Diff, a novel one-shot pruning framework that enables accurate and training-free compression of large-scale text-to-image diffusion models. Specifically, (i) OBS-Diff revitalizes the classic Optimal Brain Surgeon (OBS), adapting it to the complex architectures of modern diffusion models and supporting diverse pruning granularity, including unstructured, N:M semi-structured, and structured (MHA heads and FFN neurons) sparsity; (ii) To align the pruning criteria with the iterative dynamics of the diffusion process, by examining the problem from an error-accumulation perspective, we propose a novel timestep-aware Hessian construction that incorporates a logarithmic-decrease weighting scheme, assigning greater importance to earlier timesteps to mitigate potential error accumulation; (iii) Furthermore, a computationally efficient group-wise sequential pruning strategy is proposed to amortize the expensive calibration process. Extensive experiments show that OBS-Diff achieves state-of-the-art one-shot pruning for diffusion models, delivering inference acceleration with minimal degradation in visual quality.</p> - https://arxiv.org/abs/2510.06751 - Wed, 08 Oct 2025 08:19:15 +0000 + LLMs Can Get "Brain Rot"! + https://arxiv.org/abs/2510.13928 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13928.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shuo Xing, Junyuan Hong, Yifan Wang, Runjin Chen, Zhenyu Zhang, Ananth Grama, Zhengzhong Tu, Zhangyang Wang</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> We propose and test the LLM Brain Rot Hypothesis: continual exposure to junk web text induces lasting cognitive decline in large language models (LLMs). To causally isolate data quality, we run controlled experiments on real Twitter/X corpora, constructing junk and reversely controlled datasets via two orthogonal operationalizations: M1 (engagement degree) and M2 (semantic quality), with matched token scale and training operations across conditions. Contrary to the control group, continual pre-training of 4 LLMs on the junk dataset causes non-trivial declines (Hedges' g>0.3) on reasoning, long-context understanding, safety, and inflating "dark traits" (e.g., psychopathy, narcissism). The gradual mixtures of junk and control datasets also yield dose-response cognition decay: for example, under M1, ARC-Challenge with Chain Of Thoughts drops 74.9 rightarrow 57.2 and RULER-CWE 84.4 rightarrow 52.3 as junk ratio rises from 0% to 100%. Error forensics reveal several key insights. First, we identify thought-skipping as the primary lesion: models increasingly truncate or skip reasoning chains, explaining most of the error growth. Second, partial but incomplete healing is observed: scaling instruction tuning and clean data pre-training improve the declined cognition yet cannot restore baseline capability, suggesting persistent representational drift rather than format mismatch. Finally, we discover that the popularity, a non-semantic metric, of a tweet is a better indicator of the Brain Rot effect than the length in M1. Together, the results provide significant, multi-perspective evidence that data quality is a causal driver of LLM capability decay, reframing curation for continual pretraining as a training-time safety problem and motivating routine "cognitive health checks" for deployed LLMs.</p> + https://arxiv.org/abs/2510.13928 + Wed, 15 Oct 2025 13:28:49 +0000 - TTRV: Test-Time Reinforcement Learning for Vision Language Models - https://arxiv.org/abs/2510.06783 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06783.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Akshit Singh, Shyam Marjit, Wei Lin, Paul Gavrikov, Serena Yeung-Levy, Hilde Kuehne, Rogerio Feris, Sivan Doveh, James Glass, M. Jehanzeb Mirza</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Existing methods for extracting reward signals in Reinforcement Learning typically rely on labeled data and dedicated training splits, a setup that contrasts with how humans learn directly from their environment. In this work, we propose TTRV to enhance vision language understanding by adapting the model on the fly at inference time, without the need for any labeled data. Concretely, we enhance the Group Relative Policy Optimization (GRPO) framework by designing rewards based on the frequency of the base model's output, while inferring on each test sample multiple times. Further, we also propose to control the diversity of the model's output by simultaneously rewarding the model for obtaining low entropy of the output empirical distribution. Our approach delivers consistent gains across both object recognition and visual question answering (VQA), with improvements of up to 52.4% and 29.8%, respectively, and average boosts of 24.6% and 10.0% across 16 datasets.Remarkably, on image recognition, TTRV applied to InternVL 8B surpasses GPT-4o by an average of 2.3% over 8 benchmarks, while remaining highly competitive on VQA, demonstrating that test-time reinforcement learning can match or exceed the strongest proprietary models. Finally, we find many interesting properties of test-time RL for VLMs: for example, even in extremely data-constrained scenarios, where adaptation is performed on a single randomly chosen unlabeled test example, TTRV still yields non-trivial improvements of up to 5.5% in recognition tasks.</p> - https://arxiv.org/abs/2510.06783 - Wed, 08 Oct 2025 09:10:31 +0000 + Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization + https://arxiv.org/abs/2510.13554 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13554.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yang Li, Zhichen Dong, Yuhan Sun, Weixun Wang, Shaopan Xiong, Yijia Luo, Jiashun Liu, Han Lu, Jiamang Wang, Wenbo Su, Bo Zheng, Junchi Yan</p><p><b>Upvotes:</b> 52</p><p><b>Summary:</b> The reasoning pattern of Large language models (LLMs) remains opaque, and Reinforcement learning (RL) typically applies uniform credit across an entire generation, blurring the distinction between pivotal and routine steps. This work positions attention as a privileged substrate that renders the internal logic of LLMs legible, not merely as a byproduct of computation, but as a mechanistic blueprint of reasoning itself. We first distinguish attention heads between locally and globally focused information processing and reveal that locally focused heads produce a sawtooth pattern near the diagonal indicating phrasal chunks, while globally focused heads expose tokens that exert broad downstream influence over future tokens. We formalize these with two metrics: 1) Windowed Average Attention Distance, which measures the extent of backward attention within a clipped window; 2) Future Attention Influence, which quantifies a token's global importance as the average attention it receives from subsequent tokens. Taken together, these signals reveal a recurring preplan-and-anchor mechanism, where the model first performs a long-range contextual reference to generate an introductory token, which is immediately followed by or coincides with a semantic anchor token that organizes subsequent reasoning. Leveraging these insights, we introduce three novel RL strategies that dynamically perform targeted credit assignment to critical nodes (preplan tokens, anchor tokens, and their temporal coupling) and show consistent performance gains across various reasoning tasks. By aligning optimization with the model's intrinsic reasoning rhythm, we aim to transform opaque optimization into an actionable structure-aware process, hoping to offer a potential step toward more transparent and effective optimization of LLM reasoning.</p> + https://arxiv.org/abs/2510.13554 + Wed, 15 Oct 2025 13:49:51 +0000 - LongRM: Revealing and Unlocking the Context Boundary of Reward Modeling - https://arxiv.org/abs/2510.06915 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06915.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zecheng Tang, Baibei Ji, Quantong Qiu, Haitian Wang, Xiaobo Liang, Juntao Li, Min Zhang</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> Reward model (RM) plays a pivotal role in aligning large language model (LLM) with human preferences. As real-world applications increasingly involve long history trajectories, e.g., LLM agent, it becomes indispensable to evaluate whether a model's responses are not only high-quality but also grounded in and consistent with the provided context. Yet, current RMs remain confined to short-context settings and primarily focus on response-level attributes (e.g., safety or helpfulness), while largely neglecting the critical dimension of long context-response consistency. In this work, we introduce Long-RewardBench, a benchmark specifically designed for long-context RM evaluation, featuring both Pairwise Comparison and Best-of-N tasks. Our preliminary study reveals that even state-of-the-art generative RMs exhibit significant fragility in long-context scenarios, failing to maintain context-aware preference judgments. Motivated by the analysis of failure patterns observed in model outputs, we propose a general multi-stage training strategy that effectively scales arbitrary models into robust Long-context RMs (LongRMs). Experiments show that our approach not only substantially improves performance on long-context evaluation but also preserves strong short-context capability. Notably, our 8B LongRM outperforms much larger 70B-scale baselines and matches the performance of the proprietary Gemini 2.5 Pro model.</p> - https://arxiv.org/abs/2510.06915 - Wed, 08 Oct 2025 11:48:16 +0000 + The Role of Computing Resources in Publishing Foundation Model Research + https://arxiv.org/abs/2510.13621 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13621.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuexing Hao, Yue Huang, Haoran Zhang, Chenyang Zhao, Zhenwen Liang, Paul Pu Liang, Yue Zhao, Lichao Sun, Saleh Kalantari, Xiangliang Zhang, Marzyeh Ghassemi</p><p><b>Upvotes:</b> 13</p><p><b>Summary:</b> Cutting-edge research in Artificial Intelligence (AI) requires considerable resources, including Graphics Processing Units (GPUs), data, and human resources. In this paper, we evaluate of the relationship between these resources and the scientific advancement of foundation models (FM). We reviewed 6517 FM papers published between 2022 to 2024, and surveyed 229 first-authors to the impact of computing resources on scientific output. We find that increased computing is correlated with national funding allocations and citations, but our findings don't observe the strong correlations with research environment (academic or industrial), domain, or study methodology. We advise that individuals and institutions focus on creating shared and affordable computing opportunities to lower the entry barrier for under-resourced researchers. These steps can help expand participation in FM research, foster diversity of ideas and contributors, and sustain innovation and progress in AI. The data will be available at: https://mit-calc.csail.mit.edu/</p> + https://arxiv.org/abs/2510.13621 + Wed, 15 Oct 2025 14:50:45 +0000 - SHANKS: Simultaneous Hearing and Thinking for Spoken Language Models - https://arxiv.org/abs/2510.06917 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06917.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Cheng-Han Chiang, Xiaofei Wang, Linjie Li, Chung-Ching Lin, Kevin Lin, Shujie Liu, Zhendong Wang, Zhengyuan Yang, Hung-yi Lee, Lijuan Wang</p><p><b>Upvotes:</b> 34</p><p><b>Summary:</b> Current large language models (LLMs) and spoken language models (SLMs) begin thinking and taking actions only after the user has finished their turn. This prevents the model from interacting during the user's turn and can lead to high response latency while it waits to think. Consequently, thinking after receiving the full input is not suitable for speech-to-speech interaction, where real-time, low-latency exchange is important. We address this by noting that humans naturally "think while listening." In this paper, we propose SHANKS, a general inference framework that enables SLMs to generate unspoken chain-of-thought reasoning while listening to the user input. SHANKS streams the input speech in fixed-duration chunks and, as soon as a chunk is received, generates unspoken reasoning based on all previous speech and reasoning, while the user continues speaking. SHANKS uses this unspoken reasoning to decide whether to interrupt the user and to make tool calls to complete the task. We demonstrate that SHANKS enhances real-time user-SLM interaction in two scenarios: (1) when the user is presenting a step-by-step solution to a math problem, SHANKS can listen, reason, and interrupt when the user makes a mistake, achieving 37.1% higher interruption accuracy than a baseline that interrupts without thinking; and (2) in a tool-augmented dialogue, SHANKS can complete 56.9% of the tool calls before the user finishes their turn. Overall, SHANKS moves toward models that keep thinking throughout the conversation, not only after a turn ends. Animated illustrations of Shanks can be found at https://d223302.github.io/SHANKS/</p> - https://arxiv.org/abs/2510.06917 - Wed, 08 Oct 2025 11:48:59 +0000 + LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models + https://arxiv.org/abs/2510.13626 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13626.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Senyu Fei, Siyin Wang, Junhao Shi, Zihao Dai, Jikun Cai, Pengfang Qian, Li Ji, Xinzhe He, Shiduo Zhang, Zhaoye Fei, Jinlan Fu, Jingjing Gong, Xipeng Qiu</p><p><b>Upvotes:</b> 41</p><p><b>Summary:</b> Visual-Language-Action (VLA) models report impressive success rates on robotic manipulation benchmarks, yet these results may mask fundamental weaknesses in robustness. We perform a systematic vulnerability analysis by introducing controlled perturbations across seven dimensions: objects layout, camera viewpoints, robot initial states, language instructions, light conditions, background textures and sensor noise. We comprehensively analyzed multiple state-of-the-art models and revealed consistent brittleness beneath apparent competence. Our analysis exposes critical weaknesses: models exhibit extreme sensitivity to perturbation factors, including camera viewpoints and robot initial states, with performance dropping from 95% to below 30% under modest perturbations. Surprisingly, models are largely insensitive to language variations, with further experiments revealing that models tend to ignore language instructions completely. Our findings challenge the assumption that high benchmark scores equate to true competency and highlight the need for evaluation practices that assess reliability under realistic variation.</p> + https://arxiv.org/abs/2510.13626 + Wed, 15 Oct 2025 14:51:36 +0000 - Native Hybrid Attention for Efficient Sequence Modeling - https://arxiv.org/abs/2510.07019 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07019.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jusen Du, Jiaxi Hu, Tao Zhang, Weigao Sun, Yu Cheng</p><p><b>Upvotes:</b> 16</p><p><b>Summary:</b> Transformers excel at sequence modeling but face quadratic complexity, while linear attention offers improved efficiency but often compromises recall accuracy over long contexts. In this work, we introduce Native Hybrid Attention (NHA), a novel hybrid architecture of linear and full attention that integrates both intra \& inter-layer hybridization into a unified layer design. NHA maintains long-term context in key-value slots updated by a linear RNN, and augments them with short-term tokens from a sliding window. A single softmax attention operation is then applied over all keys and values, enabling per-token and per-head context-dependent weighting without requiring additional fusion parameters. The inter-layer behavior is controlled through a single hyperparameter, the sliding window size, which allows smooth adjustment between purely linear and full attention while keeping all layers structurally uniform. Experimental results show that NHA surpasses Transformers and other hybrid baselines on recall-intensive and commonsense reasoning tasks. Furthermore, pretrained LLMs can be structurally hybridized with NHA, achieving competitive accuracy while delivering significant efficiency gains. Code is available at https://github.com/JusenD/NHA.</p> - https://arxiv.org/abs/2510.07019 - Wed, 08 Oct 2025 13:44:57 +0000 + FlashWorld: High-quality 3D Scene Generation within Seconds + https://arxiv.org/abs/2510.13678 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13678.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xinyang Li, Tengfei Wang, Zixiao Gu, Shengchuan Zhang, Chunchao Guo, Liujuan Cao</p><p><b>Upvotes:</b> 63</p><p><b>Summary:</b> We propose FlashWorld, a generative model that produces 3D scenes from a single image or text prompt in seconds, 10~100times faster than previous works while possessing superior rendering quality. Our approach shifts from the conventional multi-view-oriented (MV-oriented) paradigm, which generates multi-view images for subsequent 3D reconstruction, to a 3D-oriented approach where the model directly produces 3D Gaussian representations during multi-view generation. While ensuring 3D consistency, 3D-oriented method typically suffers poor visual quality. FlashWorld includes a dual-mode pre-training phase followed by a cross-mode post-training phase, effectively integrating the strengths of both paradigms. Specifically, leveraging the prior from a video diffusion model, we first pre-train a dual-mode multi-view diffusion model, which jointly supports MV-oriented and 3D-oriented generation modes. To bridge the quality gap in 3D-oriented generation, we further propose a cross-mode post-training distillation by matching distribution from consistent 3D-oriented mode to high-quality MV-oriented mode. This not only enhances visual quality while maintaining 3D consistency, but also reduces the required denoising steps for inference. Also, we propose a strategy to leverage massive single-view images and text prompts during this process to enhance the model's generalization to out-of-distribution inputs. Extensive experiments demonstrate the superiority and efficiency of our method.</p> + https://arxiv.org/abs/2510.13678 + Wed, 15 Oct 2025 15:35:48 +0000 - Are We Using the Right Benchmark: An Evaluation Framework for Visual Token Compression Methods - https://arxiv.org/abs/2510.07143 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07143.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chenfei Liao, Wensong Wang, Zichen Wen, Xu Zheng, Yiyu Wang, Haocong He, Yuanhuiyi Lyu, Lutao Jiang, Xin Zou, Yuqian Fu, Bin Ren, Linfeng Zhang, Xuming Hu</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Recent endeavors to accelerate inference in Multimodal Large Language Models (MLLMs) have primarily focused on visual token compression. The effectiveness of these methods is typically assessed by measuring the accuracy drop on established benchmarks, comparing model performance before and after compression. However, these benchmarks are originally designed to assess the perception and reasoning capabilities of MLLMs, rather than to evaluate compression techniques. As a result, directly applying them to visual token compression introduces a task mismatch. Strikingly, our investigation reveals that simple image downsampling consistently outperforms many advanced compression methods across multiple widely used benchmarks. Through extensive experiments, we make the following observations: (i) Current benchmarks are noisy for the visual token compression task. (ii) Down-sampling is able to serve as a data filter to evaluate the difficulty of samples in the visual token compression task. Motivated by these findings, we introduce VTC-Bench, an evaluation framework that incorporates a data filtering mechanism to denoise existing benchmarks, thereby enabling fairer and more accurate assessment of visual token compression methods. All data and code are available at https://github.com/Chenfei-Liao/VTC-Bench.</p> - https://arxiv.org/abs/2510.07143 - Wed, 08 Oct 2025 15:44:28 +0000 + InteractiveOmni: A Unified Omni-modal Model for Audio-Visual Multi-turn Dialogue + https://arxiv.org/abs/2510.13747 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13747.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wenwen Tong, Hewei Guo, Dongchuan Ran, Jiangnan Chen, Jiefan Lu, Kaibin Wang, Keqiang Li, Xiaoxu Zhu, Jiakui Li, Kehan Li, Xueheng Li, Lumin Li, Chenxu Guo, Jiasheng Zhou, Jiandong Chen, Xianye Wu, Jiahao Wang, Silei Wu, Lei Chen, Hanming Deng, Yuxuan Song, Dinghao Zhou, Guiping Zhong, Ken Zheng, Shiyin Kang, Lewei Lu</p><p><b>Upvotes:</b> 28</p><p><b>Summary:</b> We introduce InteractiveOmni, a unified and open-source omni-modal large language model for audio-visual multi-turn interaction, ranging from 4B to 8B parameters, designed to lead the field of lightweight models by offering comprehensive omni-modal understanding and speech generation capabilities. To achieve this, we integrate the vision encoder, audio encoder, large language model, and speech decoder into a unified model for understanding and generation tasks. We design a multi-stage training strategy to ensure robust cross-modal capabilities, including pre-training for omni-modal understanding, followed by post-training with speech conversation and audio-visual interaction. To enable human-like long-term conversational ability, we meticulously curate a multi-turn training dataset that enhances the model's ability to handle complex and multi-turn interactions. To effectively evaluate the multi-turn memory and speech interaction capabilities, we construct the multi-modal multi-turn memory benchmark and the multi-turn speech interaction benchmark. Experiments demonstrate that InteractiveOmni significantly outperforms leading open-source models and provides a more intelligent multi-turn audio-visual experience, particularly in its long-term memory capabilities. Notably, InteractiveOmni-4B is comparable to the much larger model like Qwen2.5-Omni-7B on general benchmarks, and it can retain 97% of the performance of the InteractiveOmni-8B while utilizing only 50% of the model size. Achieving state-of-the-art results against similarly sized models across image, audio, video understanding, and speech generation tasks, InteractiveOmni is an accessible, open-source foundation for next-generation intelligent interactive systems.</p> + https://arxiv.org/abs/2510.13747 + Wed, 15 Oct 2025 16:52:48 +0000 - NewtonBench: Benchmarking Generalizable Scientific Law Discovery in LLM Agents - https://arxiv.org/abs/2510.07172 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07172.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Tianshi Zheng, Kelvin Kiu-Wai Tam, Newt Hue-Nam K. Nguyen, Baixuan Xu, Zhaowei Wang, Jiayang Cheng, Hong Ting Tsang, Weiqi Wang, Jiaxin Bai, Tianqing Fang, Yangqiu Song, Ginny Y. Wong, Simon See</p><p><b>Upvotes:</b> 27</p><p><b>Summary:</b> Large language models are emerging as powerful tools for scientific law discovery, a foundational challenge in AI-driven science. However, existing benchmarks for this task suffer from a fundamental methodological trilemma, forcing a trade-off between scientific relevance, scalability, and resistance to memorization. Furthermore, they oversimplify discovery as static function fitting, failing to capture the authentic scientific process of uncovering embedded laws through the interactive exploration of complex model systems. To address these critical gaps, we introduce NewtonBench, a benchmark comprising 324 scientific law discovery tasks across 12 physics domains. Our design mitigates the evaluation trilemma by using metaphysical shifts - systematic alterations of canonical laws - to generate a vast suite of problems that are scalable, scientifically relevant, and memorization-resistant. Moreover, we elevate the evaluation from static function fitting to interactive model discovery, requiring agents to experimentally probe simulated complex systems to uncover hidden principles. Our extensive experiment reveals a clear but fragile capability for discovery in frontier LLMs: this ability degrades precipitously with increasing system complexity and exhibits extreme sensitivity to observational noise. Notably, we uncover a paradoxical effect of tool assistance: providing a code interpreter can hinder more capable models by inducing a premature shift from exploration to exploitation, causing them to satisfice on suboptimal solutions. These results demonstrate that robust, generalizable discovery in complex, interactive environments remains the core challenge. By providing a scalable, robust, and scientifically authentic testbed, NewtonBench offers a crucial tool for measuring true progress and guiding the development of next-generation AI agents capable of genuine scientific discovery.</p> - https://arxiv.org/abs/2510.07172 - Wed, 08 Oct 2025 16:12:11 +0000 + Uni-MMMU: A Massive Multi-discipline Multimodal Unified Benchmark + https://arxiv.org/abs/2510.13759 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13759.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kai Zou, Ziqi Huang, Yuhao Dong, Shulin Tian, Dian Zheng, Hongbo Liu, Jingwen He, Bin Liu, Yu Qiao, Ziwei Liu</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Unified multimodal models aim to jointly enable visual understanding and generation, yet current benchmarks rarely examine their true integration. Existing evaluations either treat the two abilities in isolation or overlook tasks that inherently couple them. To address this gap, we present Uni-MMMU, a comprehensive and discipline-aware benchmark that systematically unfolds the bidirectional synergy between generation and understanding across eight reasoning-centric domains, including science, coding, mathematics, and puzzles. Each task is bidirectionally coupled, demanding models to (i) leverage conceptual understanding to guide precise visual synthesis, or (ii) utilize generation as a cognitive scaffold for analytical reasoning. Uni-MMMU incorporates verifiable intermediate reasoning steps, unique ground truths, and a reproducible scoring protocol for both textual and visual outputs. Through extensive evaluation of state-of-the-art unified, generation-only, and understanding-only models, we reveal substantial performance disparities and cross-modal dependencies, offering new insights into when and how these abilities reinforce one another, and establishing a reliable foundation for advancing unified models.</p> + https://arxiv.org/abs/2510.13759 + Wed, 15 Oct 2025 17:10:35 +0000 - When Benchmarks Age: Temporal Misalignment through Large Language Model Factuality Evaluation - https://arxiv.org/abs/2510.07238 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07238.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xunyi Jiang, Dingyi Chang, Julian McAuley, Xin Xu</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> The rapid evolution of large language models (LLMs) and the real world has outpaced the static nature of widely used evaluation benchmarks, raising concerns about their reliability for evaluating LLM factuality. While substantial works continue to rely on the popular but old benchmarks, their temporal misalignment with real-world facts and modern LLMs, and their effects on LLM factuality evaluation remain underexplored. Therefore, in this work, we present a systematic investigation of this issue by examining five popular factuality benchmarks and eight LLMs released across different years. An up-to-date fact retrieval pipeline and three metrics are tailored to quantify benchmark aging and its impact on LLM factuality evaluation. Experimental results and analysis illustrate that a considerable portion of samples in the widely used factuality benchmarks are outdated, leading to unreliable assessments of LLM factuality. We hope our work can provide a testbed to assess the reliability of a benchmark for LLM factuality evaluation and inspire more research on the benchmark aging issue. Codes are available in https://github.com/JiangXunyi/BenchAge.</p> - https://arxiv.org/abs/2510.07238 - Wed, 08 Oct 2025 17:06:07 +0000 + InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy + https://arxiv.org/abs/2510.13778 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13778.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xinyi Chen, Yilun Chen, Yanwei Fu, Ning Gao, Jiaya Jia, Weiyang Jin, Hao Li, Yao Mu, Jiangmiao Pang, Yu Qiao, Yang Tian, Bin Wang, Bolun Wang, Fangjing Wang, Hanqing Wang, Tai Wang, Ziqin Wang, Xueyuan Wei, Chao Wu, Shuai Yang, Jinhui Ye, Junqiu Yu, Jia Zeng, Jingjing Zhang, Jinyu Zhang, Shi Zhang, Feng Zheng, Bowen Zhou, Yangkun Zhu</p><p><b>Upvotes:</b> 15</p><p><b>Summary:</b> We introduce InternVLA-M1, a unified framework for spatial grounding and robot control that advances instruction-following robots toward scalable, general-purpose intelligence. Its core idea is spatially guided vision-language-action training, where spatial grounding serves as the critical link between instructions and robot actions. InternVLA-M1 employs a two-stage pipeline: (i) spatial grounding pre-training on over 2.3M spatial reasoning data to determine ``where to act'' by aligning instructions with visual, embodiment-agnostic positions, and (ii) spatially guided action post-training to decide ``how to act'' by generating embodiment-aware actions through plug-and-play spatial prompting. This spatially guided training recipe yields consistent gains: InternVLA-M1 outperforms its variant without spatial guidance by +14.6% on SimplerEnv Google Robot, +17% on WidowX, and +4.3% on LIBERO Franka, while demonstrating stronger spatial reasoning capability in box, point, and trace prediction. To further scale instruction following, we built a simulation engine to collect 244K generalizable pick-and-place episodes, enabling a 6.2% average improvement across 200 tasks and 3K+ objects. In real-world clustered pick-and-place, InternVLA-M1 improved by 7.3%, and with synthetic co-training, achieved +20.6% on unseen objects and novel configurations. Moreover, in long-horizon reasoning-intensive scenarios, it surpassed existing works by over 10%. These results highlight spatially guided training as a unifying principle for scalable and resilient generalist robots. Code and models are available at https://github.com/InternRobotics/InternVLA-M1.</p> + https://arxiv.org/abs/2510.13778 + Wed, 15 Oct 2025 17:30:05 +0000 - Hybrid Reinforcement: When Reward Is Sparse, It's Better to Be Dense - https://arxiv.org/abs/2510.07242 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07242.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Leitian Tao, Ilia Kulikov, Swarnadeep Saha, Tianlu Wang, Jing Xu, Yixuan Li, Jason E Weston, Ping Yu</p><p><b>Upvotes:</b> 27</p><p><b>Summary:</b> Post-training for reasoning of large language models (LLMs) increasingly relies on verifiable rewards: deterministic checkers that provide 0-1 correctness signals. While reliable, such binary feedback is brittle--many tasks admit partially correct or alternative answers that verifiers under-credit, and the resulting all-or-nothing supervision limits learning. Reward models offer richer, continuous feedback, which can serve as a complementary supervisory signal to verifiers. We introduce HERO (Hybrid Ensemble Reward Optimization), a reinforcement learning framework that integrates verifier signals with reward-model scores in a structured way. HERO employs stratified normalization to bound reward-model scores within verifier-defined groups, preserving correctness while refining quality distinctions, and variance-aware weighting to emphasize challenging prompts where dense signals matter most. Across diverse mathematical reasoning benchmarks, HERO consistently outperforms RM-only and verifier-only baselines, with strong gains on both verifiable and hard-to-verify tasks. Our results show that hybrid reward design retains the stability of verifiers while leveraging the nuance of reward models to advance reasoning.</p> - https://arxiv.org/abs/2510.07242 - Wed, 08 Oct 2025 17:09:41 +0000 + The Art of Scaling Reinforcement Learning Compute for LLMs + https://arxiv.org/abs/2510.13786 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13786.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Devvrit Khatri, Lovish Madaan, Rishabh Tiwari, Rachit Bansal, Sai Surya Duvvuri, Manzil Zaheer, Inderjit S. Dhillon, David Brandfonbrener, Rishabh Agarwal</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Reinforcement learning (RL) has become central to training large language models (LLMs), yet the field lacks predictive scaling methodologies comparable to those established for pre-training. Despite rapidly rising compute budgets, there is no principled understanding of how to evaluate algorithmic improvements for scaling RL compute. We present the first large-scale systematic study, amounting to more than 400,000 GPU-hours, that defines a principled framework for analyzing and predicting RL scaling in LLMs. We fit sigmoidal compute-performance curves for RL training and ablate a wide range of common design choices to analyze their effects on asymptotic performance and compute efficiency. We observe: (1) Not all recipes yield similar asymptotic performance, (2) Details such as loss aggregation, normalization, curriculum, and off-policy algorithm primarily modulate compute efficiency without materially shifting the asymptote, and (3) Stable, scalable recipes follow predictable scaling trajectories, enabling extrapolation from smaller-scale runs. Combining these insights, we propose a best-practice recipe, ScaleRL, and demonstrate its effectiveness by successfully scaling and predicting validation performance on a single RL run scaled up to 100,000 GPU-hours. Our work provides both a scientific framework for analyzing scaling in RL and a practical recipe that brings RL training closer to the predictability long achieved in pre-training.</p> + https://arxiv.org/abs/2510.13786 + Wed, 15 Oct 2025 17:43:03 +0000 - MATRIX: Mask Track Alignment for Interaction-aware Video Generation - https://arxiv.org/abs/2510.07310 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07310.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Siyoon Jin, Seongchan Kim, Dahyun Chung, Jaeho Lee, Hyunwook Choi, Jisu Nam, Jiyoung Kim, Seungryong Kim</p><p><b>Upvotes:</b> 32</p><p><b>Summary:</b> Video DiTs have advanced video generation, yet they still struggle to model multi-instance or subject-object interactions. This raises a key question: How do these models internally represent interactions? To answer this, we curate MATRIX-11K, a video dataset with interaction-aware captions and multi-instance mask tracks. Using this dataset, we conduct a systematic analysis that formalizes two perspectives of video DiTs: semantic grounding, via video-to-text attention, which evaluates whether noun and verb tokens capture instances and their relations; and semantic propagation, via video-to-video attention, which assesses whether instance bindings persist across frames. We find both effects concentrate in a small subset of interaction-dominant layers. Motivated by this, we introduce MATRIX, a simple and effective regularization that aligns attention in specific layers of video DiTs with multi-instance mask tracks from the MATRIX-11K dataset, enhancing both grounding and propagation. We further propose InterGenEval, an evaluation protocol for interaction-aware video generation. In experiments, MATRIX improves both interaction fidelity and semantic alignment while reducing drift and hallucination. Extensive ablations validate our design choices. Codes and weights will be released.</p> - https://arxiv.org/abs/2510.07310 - Wed, 08 Oct 2025 17:57:38 +0000 + Bee: A High-Quality Corpus and Full-Stack Suite to Unlock Advanced Fully Open MLLMs + https://arxiv.org/abs/2510.13795 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13795.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yi Zhang, Bolin Ni, Xin-Sheng Chen, Heng-Rui Zhang, Yongming Rao, Houwen Peng, Qinglin Lu, Han Hu, Meng-Hao Guo, Shi-Min Hu</p><p><b>Upvotes:</b> 44</p><p><b>Summary:</b> Fully open multimodal large language models (MLLMs) currently lag behind proprietary counterparts, primarily due to a significant gap in data quality for supervised fine-tuning (SFT). Existing open-source datasets are often plagued by widespread noise and a critical deficit in complex reasoning data, such as Chain-of-Thought (CoT), which hinders the development of advanced model capabilities. Addressing these challenges, our work makes three primary contributions. First, we introduce Honey-Data-15M, a new SFT dataset comprising approximately 15 million QA pairs, processed through multiple cleaning techniques and enhanced with a novel dual-level (short and long) CoT enrichment strategy. Second, we introduce HoneyPipe, the data curation pipeline, and its underlying framework DataStudio, providing the community with a transparent and adaptable methodology for data curation that moves beyond static dataset releases. Finally, to validate our dataset and pipeline, we train Bee-8B, an 8B model on Honey-Data-15M. Experiments show that Bee-8B establishes a new state-of-the-art (SOTA) for fully open MLLMs, achieving performance that is competitive with, and in some cases surpasses, recent semi-open models such as InternVL3.5-8B. Our work delivers to the community a suite of foundational resources, including: the Honey-Data-15M corpus; the full-stack suite comprising HoneyPipe and DataStudio; training recipes; an evaluation harness; and the model weights. This effort demonstrates that a principled focus on data quality is a key pathway to developing fully open MLLMs that are highly competitive with their semi-open counterparts.</p> + https://arxiv.org/abs/2510.13795 + Wed, 15 Oct 2025 17:52:59 +0000 - Vibe Checker: Aligning Code Evaluation with Human Preference - https://arxiv.org/abs/2510.07315 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07315.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ming Zhong, Xiang Zhou, Ting-Yun Chang, Qingze Wang, Nan Xu, Xiance Si, Dan Garrette, Shyam Upadhyay, Jeremiah Liu, Jiawei Han, Benoit Schillings, Jiao Sun</p><p><b>Upvotes:</b> 30</p><p><b>Summary:</b> Large Language Models (LLMs) have catalyzed vibe coding, where users leverage LLMs to generate and iteratively refine code through natural language interactions until it passes their vibe check. Vibe check is tied to real-world human preference and goes beyond functionality: the solution should feel right, read cleanly, preserve intent, and remain correct. However, current code evaluation remains anchored to pass@k and captures only functional correctness, overlooking the non-functional instructions that users routinely apply. In this paper, we hypothesize that instruction following is the missing piece underlying vibe check that represents human preference in coding besides functional correctness. To quantify models' code instruction following capabilities with measurable signals, we present VeriCode, a taxonomy of 30 verifiable code instructions together with corresponding deterministic verifiers. We use the taxonomy to augment established evaluation suites, resulting in Vibe Checker, a testbed to assess both code instruction following and functional correctness. Upon evaluating 31 leading LLMs, we show that even the strongest models struggle to comply with multiple instructions and exhibit clear functional regression. Most importantly, a composite score of functional correctness and instruction following correlates the best with human preference, with the latter emerging as the primary differentiator on real-world programming tasks. Our work identifies core factors of the vibe check, providing a concrete path for benchmarking and developing models that better align with user preferences in coding.</p> - https://arxiv.org/abs/2510.07315 - Wed, 08 Oct 2025 17:59:19 +0000 + Reasoning in Space via Grounding in the World + https://arxiv.org/abs/2510.13800 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13800.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yiming Chen, Zekun Qi, Wenyao Zhang, Xin Jin, Li Zhang, Peidong Liu</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> In this paper, we claim that 3D visual grounding is the cornerstone of spatial reasoning and introduce the Grounded-Spatial Reasoner (GS-Reasoner) to explore the effective spatial representations that bridge the gap between them. Existing 3D LLMs suffer from the absence of a unified 3D representation capable of jointly capturing semantic and geometric information. This deficiency is manifested either in poor performance on grounding or in an excessive reliance on external modules, ultimately hindering the seamless integration of grounding and spatial reasoning. To address this, we propose a simple yet effective dual-path pooling mechanism that tightly aligns geometric features with both semantic and positional cues, constructing a unified image patch-based 3D representation that encapsulates all essential information without increasing the number of input tokens. Leveraging this holistic representation, GS-Reasoner is the first 3D LLM that achieves autoregressive grounding entirely without external modules while delivering performance comparable to state-of-the-art models, establishing a unified and self-contained framework for 3D spatial reasoning. To further bridge grounding and spatial reasoning, we introduce the Grounded Chain-of-Thought (GCoT) dataset. This dataset is meticulously curated to include both 3D bounding box annotations for objects referenced in reasoning questions and step-by-step reasoning paths that integrate grounding as a core component of the problem-solving process. Extensive experiments demonstrate that GS-Reasoner achieves impressive results on 3D visual grounding, which in turn significantly enhances its spatial reasoning capabilities, leading to state-of-the-art performance.</p> + https://arxiv.org/abs/2510.13800 + Wed, 15 Oct 2025 17:58:08 +0000 - Artificial Hippocampus Networks for Efficient Long-Context Modeling - https://arxiv.org/abs/2510.07318 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07318.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yunhao Fang, Weihao Yu, Shu Zhong, Qinghao Ye, Xuehan Xiong, Lai Wei</p><p><b>Upvotes:</b> 22</p><p><b>Summary:</b> Long-sequence modeling faces a fundamental trade-off between the efficiency of compressive fixed-size memory in RNN-like models and the fidelity of lossless growing memory in attention-based Transformers. Inspired by the Multi-Store Model in cognitive science, we introduce a memory framework of artificial neural networks. Our method maintains a sliding window of the Transformer's KV cache as lossless short-term memory, while a learnable module termed Artificial Hippocampus Network (AHN) recurrently compresses out-of-window information into a fixed-size compact long-term memory. To validate this framework, we instantiate AHNs using modern RNN-like architectures, including Mamba2, DeltaNet, and Gated DeltaNet. Extensive experiments on long-context benchmarks LV-Eval and InfiniteBench demonstrate that AHN-augmented models consistently outperform sliding window baselines and achieve performance comparable or even superior to full-attention models, while substantially reducing computational and memory requirements. For instance, augmenting the Qwen2.5-3B-Instruct with AHNs reduces inference FLOPs by 40.5% and memory cache by 74.0%, while improving its average score on LV-Eval (128k sequence length) from 4.41 to 5.88. Code is available at: https://github.com/ByteDance-Seed/AHN.</p> - https://arxiv.org/abs/2510.07318 - Wed, 08 Oct 2025 17:59:55 +0000 + Trace Anything: Representing Any Video in 4D via Trajectory Fields + https://arxiv.org/abs/2510.13802 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13802.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xinhang Liu, Yuxi Xiao, Donny Y. Chen, Jiashi Feng, Yu-Wing Tai, Chi-Keung Tang, Bingyi Kang</p><p><b>Upvotes:</b> 29</p><p><b>Summary:</b> Effective spatio-temporal representation is fundamental to modeling, understanding, and predicting dynamics in videos. The atomic unit of a video, the pixel, traces a continuous 3D trajectory over time, serving as the primitive element of dynamics. Based on this principle, we propose representing any video as a Trajectory Field: a dense mapping that assigns a continuous 3D trajectory function of time to each pixel in every frame. With this representation, we introduce Trace Anything, a neural network that predicts the entire trajectory field in a single feed-forward pass. Specifically, for each pixel in each frame, our model predicts a set of control points that parameterizes a trajectory (i.e., a B-spline), yielding its 3D position at arbitrary query time instants. We trained the Trace Anything model on large-scale 4D data, including data from our new platform, and our experiments demonstrate that: (i) Trace Anything achieves state-of-the-art performance on our new benchmark for trajectory field estimation and performs competitively on established point-tracking benchmarks; (ii) it offers significant efficiency gains thanks to its one-pass paradigm, without requiring iterative optimization or auxiliary estimators; and (iii) it exhibits emergent abilities, including goal-conditioned manipulation, motion forecasting, and spatio-temporal fusion. Project page: https://trace-anything.github.io/.</p> + https://arxiv.org/abs/2510.13802 + Wed, 15 Oct 2025 17:59:04 +0000 - When Thoughts Meet Facts: Reusable Reasoning for Long-Context LMs - https://arxiv.org/abs/2510.07499 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07499.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Soyeong Jeong, Taehee Jung, Sung Ju Hwang, Joo-Kyung Kim, Dongyeop Kang</p><p><b>Upvotes:</b> 40</p><p><b>Summary:</b> Recent Long-Context Language Models (LCLMs) can process hundreds of thousands of tokens in a single prompt, enabling new opportunities for knowledge-intensive multi-hop reasoning by integrating large sets of retrieved documents or, in some cases, directly all necessary information. However, simply feeding more documents into the context window fails to capture how evidence should be connected. We address this gap with thought templates, which recast reasoning as reusable thought caches, derived from prior problem solving traces, structuring how evidence is combined and guiding multi-hop inference with factual documents. To keep these templates effective, we propose an update strategy that iteratively refines templates derived from training data through natural-language feedback. Across diverse benchmarks and LCLM families, our approach delivers consistent gains over strong baselines in both retrieval-based and retrieval-free settings. Furthermore, we show that optimized templates can be distilled into smaller open-source models, demonstrating its broad applicability and transparent reasoning reuse. We refer to our framework as Thought Template Augmented LCLMs (ToTAL).</p> - https://arxiv.org/abs/2510.07499 - Wed, 08 Oct 2025 19:52:35 +0000 + Generative Universal Verifier as Multimodal Meta-Reasoner + https://arxiv.org/abs/2510.13804 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13804.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xinchen Zhang, Xiaoying Zhang, Youbin Wu, Yanbin Cao, Renrui Zhang, Ruihang Chu, Ling Yang, Yujiu Yang</p><p><b>Upvotes:</b> 24</p><p><b>Summary:</b> We introduce Generative Universal Verifier, a novel concept and plugin designed for next-generation multimodal reasoning in vision-language models and unified multimodal models, providing the fundamental capability of reflection and refinement on visual outcomes during the reasoning and generation process. This work makes three main contributions: (1) We build ViVerBench, a comprehensive benchmark spanning 16 categories of critical tasks for evaluating visual outcomes in multimodal reasoning. Results show that existing VLMs consistently underperform across these tasks, underscoring a substantial gap from human-level capability in reliable visual verification. (2) We design two automated pipelines to construct large-scale visual verification data and train OmniVerifier-7B, the first omni-capable generative verifier trained for universal visual verification and achieves notable gains on ViVerBench(+8.3). Through training, we identify three atomic capabilities in visual verification and demonstrate how they generalize and interact synergistically. (3) We propose OmniVerifier-TTS, a sequential test-time scaling paradigm that leverages the universal verifier to bridge image generation and editing within unified models, enhancing the upper bound of generative ability through iterative fine-grained optimization. Beyond generation, we extend universal verifier to broader world-modeling interleaved reasoning scenarios. Empirically, OmniVerifier-TTS achieves improvements on T2I-ReasonBench(+3.7), and GenEval++(+4.3), outperforming existing parallel test-time scaling methods, such as Best-of-N. By endowing multimodal reasoning with reliable visual verification, OmniVerifier advances both reliable reflection during generation and scalable test-time refinement, marking a step toward more trustworthy and controllable next-generation reasoning systems.</p> + https://arxiv.org/abs/2510.13804 + Wed, 15 Oct 2025 17:59:24 +0000 - PickStyle: Video-to-Video Style Transfer with Context-Style Adapters - https://arxiv.org/abs/2510.07546 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07546.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Soroush Mehraban, Vida Adeli, Jacob Rommann, Babak Taati, Kyryl Truskovskyi</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> We address the task of video style transfer with diffusion models, where the goal is to preserve the context of an input video while rendering it in a target style specified by a text prompt. A major challenge is the lack of paired video data for supervision. We propose PickStyle, a video-to-video style transfer framework that augments pretrained video diffusion backbones with style adapters and benefits from paired still image data with source-style correspondences for training. PickStyle inserts low-rank adapters into the self-attention layers of conditioning modules, enabling efficient specialization for motion-style transfer while maintaining strong alignment between video content and style. To bridge the gap between static image supervision and dynamic video, we construct synthetic training clips from paired images by applying shared augmentations that simulate camera motion, ensuring temporal priors are preserved. In addition, we introduce Context-Style Classifier-Free Guidance (CS-CFG), a novel factorization of classifier-free guidance into independent text (style) and video (context) directions. CS-CFG ensures that context is preserved in generated video while the style is effectively transferred. Experiments across benchmarks show that our approach achieves temporally coherent, style-faithful, and content-preserving video translations, outperforming existing baselines both qualitatively and quantitatively.</p> - https://arxiv.org/abs/2510.07546 - Wed, 08 Oct 2025 21:02:55 +0000 + PhysMaster: Mastering Physical Representation for Video Generation via Reinforcement Learning + https://arxiv.org/abs/2510.13809 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13809.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Sihui Ji, Xi Chen, Xin Tao, Pengfei Wan, Hengshuang Zhao</p><p><b>Upvotes:</b> 35</p><p><b>Summary:</b> Video generation models nowadays are capable of generating visually realistic videos, but often fail to adhere to physical laws, limiting their ability to generate physically plausible videos and serve as ''world models''. To address this issue, we propose PhysMaster, which captures physical knowledge as a representation for guiding video generation models to enhance their physics-awareness. Specifically, PhysMaster is based on the image-to-video task where the model is expected to predict physically plausible dynamics from the input image. Since the input image provides physical priors like relative positions and potential interactions of objects in the scenario, we devise PhysEncoder to encode physical information from it as an extra condition to inject physical knowledge into the video generation process. The lack of proper supervision on the model's physical performance beyond mere appearance motivates PhysEncoder to apply reinforcement learning with human feedback to physical representation learning, which leverages feedback from generation models to optimize physical representations with Direct Preference Optimization (DPO) in an end-to-end manner. PhysMaster provides a feasible solution for improving physics-awareness of PhysEncoder and thus of video generation, proving its ability on a simple proxy task and generalizability to wide-ranging physical scenarios. This implies that our PhysMaster, which unifies solutions for various physical processes via representation learning in the reinforcement learning paradigm, can act as a generic and plug-in solution for physics-aware video generation and broader applications.</p> + https://arxiv.org/abs/2510.13809 + Wed, 15 Oct 2025 17:59:59 +0000 - Learning on the Job: An Experience-Driven Self-Evolving Agent for Long-Horizon Tasks - https://arxiv.org/abs/2510.08002 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08002.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Cheng Yang, Xuemeng Yang, Licheng Wen, Daocheng Fu, Jianbiao Mei, Rong Wu, Pinlong Cai, Yufan Shen, Nianchen Deng, Botian Shi, Yu Qiao, Haifeng Li</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Large Language Models have demonstrated remarkable capabilities across diverse domains, yet significant challenges persist when deploying them as AI agents for real-world long-horizon tasks. Existing LLM agents suffer from a critical limitation: they are test-time static and cannot learn from experience, lacking the ability to accumulate knowledge and continuously improve on the job. To address this challenge, we propose MUSE, a novel agent framework that introduces an experience-driven, self-evolving system centered around a hierarchical Memory Module. MUSE organizes diverse levels of experience and leverages them to plan and execute long-horizon tasks across multiple applications. After each sub-task execution, the agent autonomously reflects on its trajectory, converting the raw trajectory into structured experience and integrating it back into the Memory Module. This mechanism enables the agent to evolve beyond its static pretrained parameters, fostering continuous learning and self-evolution. We evaluate MUSE on the long-horizon productivity benchmark TAC. It achieves new SOTA performance by a significant margin using only a lightweight Gemini-2.5 Flash model. Sufficient Experiments demonstrate that as the agent autonomously accumulates experience, it exhibits increasingly superior task completion capabilities, as well as robust continuous learning and self-evolution capabilities. Moreover, the accumulated experience from MUSE exhibits strong generalization properties, enabling zero-shot improvement on new tasks. MUSE establishes a new paradigm for AI agents capable of real-world productivity task automation.</p> - https://arxiv.org/abs/2510.08002 - Thu, 09 Oct 2025 09:40:34 +0000 + BitNet Distillation + https://arxiv.org/abs/2510.13998 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.13998.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xun Wu, Shaohan Huang, Wenhui Wang, Ting Song, Li Dong, Yan Xia, Furu Wei</p><p><b>Upvotes:</b> 42</p><p><b>Summary:</b> In this paper, we present BitNet Distillation (BitDistill), a lightweight pipeline that fine-tunes off-the-shelf full-precision LLMs (e.g., Qwen) into 1.58-bit precision (i.e., ternary weights {-1, 0, 1}) for specific downstream tasks, achieving strong task-specific performance with minimal computational cost. Specifically, BitDistill incorporates three key techniques: the SubLN module, as introduced in BitNet; multi-head attention distillation, based on MiniLM; and continual pre-training, which serves as a crucial warm-up step to mitigate the scalability issue of the performance gap between finetuned full-precision and 1.58-bit LLMs on specific tasks. Experimental results show that BitDistill achieves performance comparable to the full-precision counterpart models across model size, while enabling up to 10x memory savings and 2.65x faster inference on CPUs. Code is available at https://github.com/microsoft/BitNet.</p> + https://arxiv.org/abs/2510.13998 + Wed, 15 Oct 2025 18:28:12 +0000 - UniMMVSR: A Unified Multi-Modal Framework for Cascaded Video Super-Resolution - https://arxiv.org/abs/2510.08143 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08143.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shian Du, Menghan Xia, Chang Liu, Quande Liu, Xintao Wang, Pengfei Wan, Xiangyang Ji</p><p><b>Upvotes:</b> 19</p><p><b>Summary:</b> Cascaded video super-resolution has emerged as a promising technique for decoupling the computational burden associated with generating high-resolution videos using large foundation models. Existing studies, however, are largely confined to text-to-video tasks and fail to leverage additional generative conditions beyond text, which are crucial for ensuring fidelity in multi-modal video generation. We address this limitation by presenting UniMMVSR, the first unified generative video super-resolution framework to incorporate hybrid-modal conditions, including text, images, and videos. We conduct a comprehensive exploration of condition injection strategies, training schemes, and data mixture techniques within a latent video diffusion model. A key challenge was designing distinct data construction and condition utilization methods to enable the model to precisely utilize all condition types, given their varied correlations with the target video. Our experiments demonstrate that UniMMVSR significantly outperforms existing methods, producing videos with superior detail and a higher degree of conformity to multi-modal conditions. We also validate the feasibility of combining UniMMVSR with a base model to achieve multi-modal guided generation of 4K video, a feat previously unattainable with existing techniques.</p> - https://arxiv.org/abs/2510.08143 - Thu, 09 Oct 2025 12:25:16 +0000 + Qwen3Guard Technical Report + https://arxiv.org/abs/2510.14276 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14276.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haiquan Zhao, Chenhan Yuan, Fei Huang, Xiaomeng Hu, Yichang Zhang, An Yang, Bowen Yu, Dayiheng Liu, Jingren Zhou, Junyang Lin, Baosong Yang, Chen Cheng, Jialong Tang, Jiandong Jiang, Jianwei Zhang, Jijie Xu, Ming Yan, Minmin Sun, Pei Zhang, Pengjun Xie, Qiaoyu Tang, Qin Zhu, Rong Zhang, Shibin Wu, Shuo Zhang, Tao He, Tianyi Tang, Tingyu Xia, Wei Liao, Weizhou Shen, Wenbiao Yin, Wenmeng Zhou, Wenyuan Yu, Xiaobin Wang, Xiaodong Deng, Xiaodong Xu, Xinyu Zhang, Yang Liu, Yeqiu Li, Yi Zhang, Yong Jiang, Yu Wan, Yuxin Zhou</p><p><b>Upvotes:</b> 10</p><p><b>Summary:</b> As large language models (LLMs) become more capable and widely used, ensuring the safety of their outputs is increasingly critical. Existing guardrail models, though useful in static evaluation settings, face two major limitations in real-world applications: (1) they typically output only binary "safe/unsafe" labels, which can be interpreted inconsistently across diverse safety policies, rendering them incapable of accommodating varying safety tolerances across domains; and (2) they require complete model outputs before performing safety checks, making them fundamentally incompatible with streaming LLM inference, thereby preventing timely intervention during generation and increasing exposure to harmful partial outputs. To address these challenges, we present Qwen3Guard, a series of multilingual safety guardrail models with two specialized variants: Generative Qwen3Guard, which casts safety classification as an instruction-following task to enable fine-grained tri-class judgments (safe, controversial, unsafe); and Stream Qwen3Guard, which introduces a token-level classification head for real-time safety monitoring during incremental text generation. Both variants are available in three sizes (0.6B, 4B, and 8B parameters) and support up to 119 languages and dialects, providing comprehensive, scalable, and low-latency safety moderation for global LLM deployments. Evaluated across English, Chinese, and multilingual benchmarks, Qwen3Guard achieves state-of-the-art performance in both prompt and response safety classification. All models are released under the Apache 2.0 license for public use.</p> + https://arxiv.org/abs/2510.14276 + Thu, 16 Oct 2025 04:00:18 +0000 - Training-Free Group Relative Policy Optimization - https://arxiv.org/abs/2510.08191 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08191.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuzheng Cai, Siqi Cai, Yuchen Shi, Zihan Xu, Lichao Chen, Yulei Qin, Xiaoyu Tan, Gang Li, Zongyi Li, Haojia Lin, Yong Mao, Ke Li, Xing Sun</p><p><b>Upvotes:</b> 32</p><p><b>Summary:</b> Recent advances in Large Language Model (LLM) agents have demonstrated their promising general capabilities. However, their performance in specialized real-world domains often degrades due to challenges in effectively integrating external tools and specific prompting strategies. While methods like agentic reinforcement learning have been proposed to address this, they typically rely on costly parameter updates, for example, through a process that uses Supervised Fine-Tuning (SFT) followed by a Reinforcement Learning (RL) phase with Group Relative Policy Optimization (GRPO) to alter the output distribution. However, we argue that LLMs can achieve a similar effect on the output distribution by learning experiential knowledge as a token prior, which is a far more lightweight approach that not only addresses practical data scarcity but also avoids the common issue of overfitting. To this end, we propose Training-Free Group Relative Policy Optimization (Training-Free GRPO), a cost-effective solution that enhances LLM agent performance without any parameter updates. Our method leverages the group relative semantic advantage instead of numerical ones within each group of rollouts, iteratively distilling high-quality experiential knowledge during multi-epoch learning on a minimal ground-truth data. Such knowledge serves as the learned token prior, which is seamlessly integrated during LLM API calls to guide model behavior. Experiments on mathematical reasoning and web searching tasks demonstrate that Training-Free GRPO, when applied to DeepSeek-V3.1-Terminus, significantly improves out-of-domain performance. With just a few dozen training samples, Training-Free GRPO outperforms fine-tuned small LLMs with marginal training data and cost.</p> - https://arxiv.org/abs/2510.08191 - Thu, 09 Oct 2025 13:18:17 +0000 + AI for Service: Proactive Assistance with AI Glasses + https://arxiv.org/abs/2510.14359 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14359.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zichen Wen, Yiyu Wang, Chenfei Liao, Boxue Yang, Junxian Li, Weifeng Liu, Haocong He, Bolong Feng, Xuyang Liu, Yuanhuiyi Lyu, Xu Zheng, Xuming Hu, Linfeng Zhang</p><p><b>Upvotes:</b> 65</p><p><b>Summary:</b> In an era where AI is evolving from a passive tool into an active and adaptive companion, we introduce AI for Service (AI4Service), a new paradigm that enables proactive and real-time assistance in daily life. Existing AI services remain largely reactive, responding only to explicit user commands. We argue that a truly intelligent and helpful assistant should be capable of anticipating user needs and taking actions proactively when appropriate. To realize this vision, we propose Alpha-Service, a unified framework that addresses two fundamental challenges: Know When to intervene by detecting service opportunities from egocentric video streams, and Know How to provide both generalized and personalized services. Inspired by the von Neumann computer architecture and based on AI glasses, Alpha-Service consists of five key components: an Input Unit for perception, a Central Processing Unit for task scheduling, an Arithmetic Logic Unit for tool utilization, a Memory Unit for long-term personalization, and an Output Unit for natural human interaction. As an initial exploration, we implement Alpha-Service through a multi-agent system deployed on AI glasses. Case studies, including a real-time Blackjack advisor, a museum tour guide, and a shopping fit assistant, demonstrate its ability to seamlessly perceive the environment, infer user intent, and provide timely and useful assistance without explicit prompts.</p> + https://arxiv.org/abs/2510.14359 + Thu, 16 Oct 2025 06:55:28 +0000 - LLMs Learn to Deceive Unintentionally: Emergent Misalignment in Dishonesty from Misaligned Samples to Biased Human-AI Interactions - https://arxiv.org/abs/2510.08211 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08211.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> XuHao Hu, Peng Wang, Xiaoya Lu, Dongrui Liu, Xuanjing Huang, Jing Shao</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> Previous research has shown that LLMs finetuned on malicious or incorrect completions within narrow domains (e.g., insecure code or incorrect medical advice) can become broadly misaligned to exhibit harmful behaviors, which is called emergent misalignment. In this work, we investigate whether this phenomenon can extend beyond safety behaviors to a broader spectrum of dishonesty and deception under high-stakes scenarios (e.g., lying under pressure and deceptive behavior). To explore this, we finetune open-sourced LLMs on misaligned completions across diverse domains. Experimental results demonstrate that LLMs show broadly misaligned behavior in dishonesty. Additionally, we further explore this phenomenon in a downstream combined finetuning setting, and find that introducing as little as 1% of misalignment data into a standard downstream task is sufficient to decrease honest behavior over 20%. Furthermore, we consider a more practical human-AI interaction environment where we simulate both benign and biased users to interact with the assistant LLM. Notably, we find that the assistant can be misaligned unintentionally to exacerbate its dishonesty with only 10% biased user population. In summary, we extend the study of emergent misalignment to the domain of dishonesty and deception under high-stakes scenarios, and demonstrate that this risk arises not only through direct finetuning, but also in downstream mixture tasks and practical human-AI interactions.</p> - https://arxiv.org/abs/2510.08211 - Thu, 09 Oct 2025 13:35:19 +0000 + PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model + https://arxiv.org/abs/2510.14528 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14528.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Cheng Cui, Ting Sun, Suyin Liang, Tingquan Gao, Zelun Zhang, Jiaxuan Liu, Xueqing Wang, Changda Zhou, Hongen Liu, Manhui Lin, Yue Zhang, Yubo Zhang, Handong Zheng, Jing Zhang, Jun Zhang, Yi Liu, Dianhai Yu, Yanjun Ma</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> In this report, we propose PaddleOCR-VL, a SOTA and resource-efficient model tailored for document parsing. Its core component is PaddleOCR-VL-0.9B, a compact yet powerful vision-language model (VLM) that integrates a NaViT-style dynamic resolution visual encoder with the ERNIE-4.5-0.3B language model to enable accurate element recognition. This innovative model efficiently supports 109 languages and excels in recognizing complex elements (e.g., text, tables, formulas, and charts), while maintaining minimal resource consumption. Through comprehensive evaluations on widely used public benchmarks and in-house benchmarks, PaddleOCR-VL achieves SOTA performance in both page-level document parsing and element-level recognition. It significantly outperforms existing solutions, exhibits strong competitiveness against top-tier VLMs, and delivers fast inference speeds. These strengths make it highly suitable for practical deployment in real-world scenarios.</p> + https://arxiv.org/abs/2510.14528 + Thu, 16 Oct 2025 10:18:48 +0000 - The Alignment Waltz: Jointly Training Agents to Collaborate for Safety - https://arxiv.org/abs/2510.08240 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08240.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jingyu Zhang, Haozhu Wang, Eric Michael Smith, Sid Wang, Amr Sharaf, Mahesh Pasupuleti, Benjamin Van Durme, Daniel Khashabi, Jason Weston, Hongyuan Zhan</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> Harnessing the power of LLMs requires a delicate dance between being helpful and harmless. This creates a fundamental tension between two competing challenges: vulnerability to adversarial attacks that elicit unsafe content, and a tendency for overrefusal on benign but sensitive prompts. Current approaches often navigate this dance with safeguard models that completely reject any content that contains unsafe portions. This approach cuts the music entirely-it may exacerbate overrefusals and fails to provide nuanced guidance for queries it refuses. To teach models a more coordinated choreography, we propose WaltzRL, a novel multi-agent reinforcement learning framework that formulates safety alignment as a collaborative, positive-sum game. WaltzRL jointly trains a conversation agent and a feedback agent, where the latter is incentivized to provide useful suggestions that improve the safety and helpfulness of the conversation agent's responses. At the core of WaltzRL is a Dynamic Improvement Reward (DIR) that evolves over time based on how well the conversation agent incorporates the feedback. At inference time, unsafe or overrefusing responses from the conversation agent are improved rather than discarded. The feedback agent is deployed together with the conversation agent and only engages adaptively when needed, preserving helpfulness and low latency on safe queries. Our experiments, conducted across five diverse datasets, demonstrate that WaltzRL significantly reduces both unsafe responses (e.g., from 39.0% to 4.6% on WildJailbreak) and overrefusals (from 45.3% to 9.9% on OR-Bench) compared to various baselines. By enabling the conversation and feedback agents to co-evolve and adaptively apply feedback, WaltzRL enhances LLM safety without degrading general capabilities, thereby advancing the Pareto front between helpfulness and harmlessness.</p> - https://arxiv.org/abs/2510.08240 - Thu, 09 Oct 2025 14:03:05 +0000 + Agentic Entropy-Balanced Policy Optimization + https://arxiv.org/abs/2510.14545 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14545.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Guanting Dong, Licheng Bao, Zhongyuan Wang, Kangzhi Zhao, Xiaoxi Li, Jiajie Jin, Jinghan Yang, Hangyu Mao, Fuzheng Zhang, Kun Gai, Guorui Zhou, Yutao Zhu, Ji-Rong Wen, Zhicheng Dou</p><p><b>Upvotes:</b> 90</p><p><b>Summary:</b> Recently, Agentic Reinforcement Learning (Agentic RL) has made significant progress in incentivizing the multi-turn, long-horizon tool-use capabilities of web agents. While mainstream agentic RL algorithms autonomously explore high-uncertainty tool-call steps under the guidance of entropy, excessive reliance on entropy signals can impose further constraints, leading to the training collapse. In this paper, we delve into the challenges caused by entropy and propose the Agentic Entropy-Balanced Policy Optimization (AEPO), an agentic RL algorithm designed to balance entropy in both the rollout and policy update phases. AEPO comprises two core components: (1) a dynamic entropy-balanced rollout mechanism that adaptively allocate global and branch sampling budget through entropy pre-monitoring, while imposing a branch penalty on consecutive high-entropy tool-call steps to prevent over-branching issues; and (2) Entropy-Balanced Policy Optimization that inserts a stop-gradient operation into the high-entropy clipping term to preserve and properly rescale gradients on high-entropy tokens, while incorporating entropy-aware advantage estimation to prioritize learning on high-uncertainty tokens. Results across 14 challenging datasets show that AEPO consistently outperforms 7 mainstream RL algorithms. With just 1K RL samples, Qwen3-14B with AEPO achieves impressive results: 47.6% on GAIA, 11.2% on Humanity's Last Exam, and 43.0% on WebWalker for Pass@1; 65.0% on GAIA, 26.0% on Humanity's Last Exam, and 70.0% on WebWalker for Pass@5. Further analysis reveals that AEPO improves rollout sampling diversity while maintaining stable policy entropy, facilitating scalable web agent training.</p> + https://arxiv.org/abs/2510.14545 + Thu, 16 Oct 2025 10:40:52 +0000 - Beyond Turn Limits: Training Deep Search Agents with Dynamic Context Window - https://arxiv.org/abs/2510.08276 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08276.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Qiaoyu Tang, Hao Xiang, Le Yu, Bowen Yu, Yaojie Lu, Xianpei Han, Le Sun, WenJuan Zhang, Pengbo Wang, Shixuan Liu, Zhenru Zhang, Jianhong Tu, Hongyu Lin, Junyang Lin</p><p><b>Upvotes:</b> 7</p><p><b>Summary:</b> While recent advances in reasoning models have demonstrated cognitive behaviors through reinforcement learning, existing approaches struggle to invoke deep reasoning capabilities in multi-turn agents with long-horizon interactions. We propose DeepMiner, a novel framework that elicits such abilities by introducing high-difficulty training tasks and dynamic context window. DeepMiner presents a reverse construction method to generate complex but verifiable question-answer pairs from authentic web sources, which ensures the challenge and reliability of training data while injecting cognitive capabilities into multi-turn reasoning scenarios. We further design an elegant yet effective dynamic context management strategy for both training and inference, utilizing sliding window mechanisms while eliminating the dependency on external summarization models, thereby efficiently empowering the model to handle continuously expanding long-horizon contexts. Through reinforcement learning on Qwen3-32B, we develop DeepMiner-32B, which achieves substantial performance improvements across multiple search agent benchmarks. DeepMiner attains 33.5% accuracy on BrowseComp-en, surpassing the previous best open-source agent by almost 20 percentage points, and demonstrates consistent improvements on BrowseComp-zh, XBench-DeepSearch, and GAIA. Notably, our dynamic context management enables sustained interactions of nearly 100 turns within standard 32k context length, effectively addressing the context limitations that constrain existing multi-turn interaction systems.</p> - https://arxiv.org/abs/2510.08276 - Thu, 09 Oct 2025 14:31:39 +0000 + Beyond Correctness: Evaluating Subjective Writing Preferences Across Cultures + https://arxiv.org/abs/2510.14616 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14616.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shuangshuang Ying, Yunwen Li, Xingwei Qu, Xin Li, Sheng Jin, Minghao Liu, Zhoufutu Wen, Xeron Du, Tianyu Zheng, Yichi Zhang, Letian Ni, Yuyang Cheng, Qiguang Chen, Jingzhe Ding, Shengda Long, Wangchunshu Zhou, Jiazhan Feng, Wanjun Zhong, Libo Qin, Ge Zhang, Wenhao Huang, Wanxiang Che, Chenghua Lin</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> Current preference learning methods achieve high accuracy on standard benchmarks but exhibit significant performance degradation when objective quality signals are removed. We introduce WritingPreferenceBench, a dataset of 1,800 human-annotated preference pairs (1,200 English, 600 Chinese) across 8 creative writing genres, where responses are matched for objective correctness, factual accuracy, and length. On this benchmark, sequence-based reward models--the standard architecture for RLHF--achieve only 52.7% mean accuracy, while zero-shot language model judges perform at 53.9%. In contrast, generative reward models that produce explicit reasoning chains achieve 81.8% accuracy. We observe high within-model variance across genres: individual models range from 18.2% to 81.8% accuracy across different writing categories, with standard deviations averaging 10.1%. This variance persists regardless of model scale, with 27B parameter models showing no consistent improvement over 8B variants. Our results suggest that current RLHF methods primarily learn to detect objective errors rather than capture subjective quality preferences (e.g., creativity, stylistic flair, and emotional resonance), and that successful preference modeling may require intermediate reasoning representations rather than direct classification.</p> + https://arxiv.org/abs/2510.14616 + Thu, 16 Oct 2025 12:23:13 +0000 - First Try Matters: Revisiting the Role of Reflection in Reasoning Models - https://arxiv.org/abs/2510.08308 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08308.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Liwei Kang, Yue Deng, Yao Xiao, Zhanfeng Mo, Wee Sun Lee, Lidong Bing</p><p><b>Upvotes:</b> 22</p><p><b>Summary:</b> Large language models have recently demonstrated significant gains in reasoning ability, often attributed to their capacity to generate longer chains of thought and engage in reflective reasoning. However, the contribution of reflections to performance improvement remains unclear. In this paper, we systematically analyze the rollouts of eight reasoning models on five mathematical datasets. We focus on reflective behaviours where the model has already produced an answer but continues reflecting before finalizing its output. Our analysis reveals that reflections are predominantly confirmatory and rarely alter the model's initial answer, a pattern consistent across models and datasets. To understand the role of reflections in training, we construct supervised fine-tuning (SFT) datasets with varying amounts of reflection steps. We observe that training models on rollouts with more reflection steps primarily enhances first-answer correctness rather than the ability to correct initially wrong answers through reflections. This motivates us to propose a question-aware early-stopping method that enhances inference-time token efficiency by stopping the reasoning process once a few plausible candidate answers are generated, thereby reducing unnecessary reflection steps. Motivated by this, we further propose to dynamically truncate the reflections after a candidate answer has appeared during generation, which reduces reasoning tokens by 24.5% across five mathematical datasets, within a 2.9% drop in accuracy.</p> - https://arxiv.org/abs/2510.08308 - Thu, 09 Oct 2025 14:57:10 +0000 + COIG-Writer: A High-Quality Dataset for Chinese Creative Writing with Thought Processes + https://arxiv.org/abs/2510.14763 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14763.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yunwen Li, Shuangshuang Ying, Xingwei Qu, Xin Li, Sheng Jin, Minghao Liu, Zhoufutu Wen, Tianyu Zheng, Xeron Du, Qiguang Chen, Jiajun Shi, Wangchunshu Zhou, Jiazhan Feng, Wanjun Zhong, Libo Qin, Stephen Huang, Wanxiang Che, Chenghua Lin, Eli Zhang</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Large language models exhibit systematic deficiencies in creative writing, particularly in non-English contexts where training data is scarce and lacks process-level supervision. We present COIG-Writer, a novel Chinese creative writing dataset that captures both diverse outputs and their underlying thought processes through systematic reverse-engineering of high-quality texts. Unlike existing datasets that provide only input-output pairs, COIG-Writer comprises 1,665 meticulously curated triplets spanning 51 genres, each containing: (1) a reverse-engineered prompt, (2) detailed creative reasoning documenting decision-making processes, and (3) the final text. Through comprehensive experiments, we identify a two-component model of creative writing: narrative logic (provided by process supervision) and linguistic expression (maintained by general-purpose data). Our findings reveal three critical insights: (1) Process supervision is highly effective but requires stabilization with general data. A ratio of at least one creative sample to twelve general samples is needed to achieve optimal performance; below this threshold, the win rate progressively degrades (from 62.75% down to 35.78%)., (2) creative capabilities are culturally-bound with no cross-lingual transfer (89.26pp gap between Chinese and English performance), and (3) lexical diversity inversely correlates with creative quality (TTR paradox), suggesting high diversity signals compensatory behavior for logical deficiencies. These findings establish that creative excellence emerges from the interaction between logical scaffolding and linguistic grounding, analogous to how mathematical reasoning enhances but cannot replace linguistic competence in foundation models.</p> + https://arxiv.org/abs/2510.14763 + Thu, 16 Oct 2025 15:01:19 +0000 - UniVideo: Unified Understanding, Generation, and Editing for Videos - https://arxiv.org/abs/2510.08377 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08377.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Cong Wei, Quande Liu, Zixuan Ye, Qiulin Wang, Xintao Wang, Pengfei Wan, Kun Gai, Wenhu Chen</p><p><b>Upvotes:</b> 56</p><p><b>Summary:</b> Unified multimodal models have shown promising results in multimodal content generation and editing but remain largely limited to the image domain. In this work, we present UniVideo, a versatile framework that extends unified modeling to the video domain. UniVideo adopts a dual-stream design, combining a Multimodal Large Language Model (MLLM) for instruction understanding with a Multimodal DiT (MMDiT) for video generation. This design enables accurate interpretation of complex multimodal instructions while preserving visual consistency. Built on this architecture, UniVideo unifies diverse video generation and editing tasks under a single multimodal instruction paradigm and is jointly trained across them. Extensive experiments demonstrate that UniVideo matches or surpasses state-of-the-art task-specific baselines in text/image-to-video generation, in-context video generation and in-context video editing. Notably, the unified design of UniVideo enables two forms of generalization. First, UniVideo supports task composition, such as combining editing with style transfer, by integrating multiple capabilities within a single instruction. Second, even without explicit training on free-form video editing, UniVideo transfers its editing capability from large-scale image editing data to this setting, handling unseen instructions such as green-screening characters or changing materials within a video. Beyond these core capabilities, UniVideo also supports visual-prompt-based video generation, where the MLLM interprets visual prompts and guides the MMDiT during synthesis. To foster future research, we will release our model and code.</p> - https://arxiv.org/abs/2510.08377 - Thu, 09 Oct 2025 16:01:30 +0000 + ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints + https://arxiv.org/abs/2510.14847 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14847.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Meiqi Wu, Jiashu Zhu, Xiaokun Feng, Chubin Chen, Chen Zhu, Bingze Song, Fangyuan Mao, Jiahong Wu, Xiangxiang Chu, Kaiqi Huang</p><p><b>Upvotes:</b> 48</p><p><b>Summary:</b> Video generation models have achieved remarkable progress, particularly excelling in realistic scenarios; however, their performance degrades notably in imaginative scenarios. These prompts often involve rarely co-occurring concepts with long-distance semantic relationships, falling outside training distributions. Existing methods typically apply test-time scaling for improving video quality, but their fixed search spaces and static reward designs limit adaptability to imaginative scenarios. To fill this gap, we propose ImagerySearch, a prompt-guided adaptive test-time search strategy that dynamically adjusts both the inference search space and reward function according to semantic relationships in the prompt. This enables more coherent and visually plausible videos in challenging imaginative settings. To evaluate progress in this direction, we introduce LDT-Bench, the first dedicated benchmark for long-distance semantic prompts, consisting of 2,839 diverse concept pairs and an automated protocol for assessing creative generation capabilities. Extensive experiments show that ImagerySearch consistently outperforms strong video generation baselines and existing test-time scaling approaches on LDT-Bench, and achieves competitive improvements on VBench, demonstrating its effectiveness across diverse prompt types. We will release LDT-Bench and code to facilitate future research on imaginative video generation.</p> + https://arxiv.org/abs/2510.14847 + Thu, 16 Oct 2025 16:19:13 +0000 - Reinforcing Diffusion Models by Direct Group Preference Optimization - https://arxiv.org/abs/2510.08425 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08425.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yihong Luo, Tianyang Hu, Jing Tang</p><p><b>Upvotes:</b> 9</p><p><b>Summary:</b> While reinforcement learning methods such as Group Relative Preference Optimization (GRPO) have significantly enhanced Large Language Models, adapting them to diffusion models remains challenging. In particular, GRPO demands a stochastic policy, yet the most cost-effective diffusion samplers are based on deterministic ODEs. Recent work addresses this issue by using inefficient SDE-based samplers to induce stochasticity, but this reliance on model-agnostic Gaussian noise leads to slow convergence. To resolve this conflict, we propose Direct Group Preference Optimization (DGPO), a new online RL algorithm that dispenses with the policy-gradient framework entirely. DGPO learns directly from group-level preferences, which utilize relative information of samples within groups. This design eliminates the need for inefficient stochastic policies, unlocking the use of efficient deterministic ODE samplers and faster training. Extensive results show that DGPO trains around 20 times faster than existing state-of-the-art methods and achieves superior performance on both in-domain and out-of-domain reward metrics. Code is available at https://github.com/Luo-Yihong/DGPO.</p> - https://arxiv.org/abs/2510.08425 - Thu, 09 Oct 2025 16:40:43 +0000 + Fantastic (small) Retrievers and How to Train Them: mxbai-edge-colbert-v0 Tech Report + https://arxiv.org/abs/2510.14880 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14880.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Rikiya Takehi, Benjamin Clavié, Sean Lee, Aamir Shakir</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> In this work, we introduce mxbai-edge-colbert-v0 models, at two different parameter counts: 17M and 32M. As part of our research, we conduct numerous experiments to improve retrieval and late-interaction models, which we intend to distill into smaller models as proof-of-concepts. Our ultimate aim is to support retrieval at all scales, from large-scale retrieval which lives in the cloud to models that can run locally, on any device. mxbai-edge-colbert-v0 is a model that we hope will serve as a solid foundation backbone for all future experiments, representing the first version of a long series of small proof-of-concepts. As part of the development of mxbai-edge-colbert-v0, we conducted multiple ablation studies, of which we report the results. In terms of downstream performance, mxbai-edge-colbert-v0 is a particularly capable small model, outperforming ColBERTv2 on common short-text benchmarks (BEIR) and representing a large step forward in long-context tasks, with unprecedented efficiency.</p> + https://arxiv.org/abs/2510.14880 + Thu, 16 Oct 2025 17:00:35 +0000 - DeepPrune: Parallel Scaling without Inter-trace Redundancy - https://arxiv.org/abs/2510.08483 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08483.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shangqing Tu, Yaxuan Li, Yushi Bai, Lei Hou, Juanzi Li</p><p><b>Upvotes:</b> 23</p><p><b>Summary:</b> Parallel scaling has emerged as a powerful paradigm to enhance reasoning capabilities in large language models (LLMs) by generating multiple Chain-of-Thought (CoT) traces simultaneously. However, this approach introduces significant computational inefficiency due to inter-trace redundancy -- our analysis reveals that over 80% of parallel reasoning traces yield identical final answers, representing substantial wasted computation. To address this critical efficiency bottleneck, we propose DeepPrune, a novel framework that enables efficient parallel scaling through dynamic pruning. Our method features a specialized judge model trained with focal loss and oversampling techniques to accurately predict answer equivalence from partial reasoning traces which realizes 0.87 AUROC on equivalence prediction, combined with an online greedy clustering algorithm that dynamically prunes redundant paths while preserving answer diversity. Comprehensive evaluations across three challenging benchmarks (AIME 2024, AIME 2025, and GPQA) and multiple reasoning models demonstrate that DeepPrune achieves remarkable token reduction by over 80% compared to conventional consensus sampling on most cases, while maintaining competitive accuracy within 3 percentage points. Our work establishes a new standard for efficient parallel reasoning, making high-performance reasoning more efficient. Our code and data are here: https://deepprune.github.io/</p> - https://arxiv.org/abs/2510.08483 - Thu, 09 Oct 2025 17:24:54 +0000 + VLA^2: Empowering Vision-Language-Action Models with an Agentic Framework for Unseen Concept Manipulation + https://arxiv.org/abs/2510.14902 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14902.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Han Zhao, Jiaxuan Zhang, Wenxuan Song, Pengxiang Ding, Donglin Wang</p><p><b>Upvotes:</b> 11</p><p><b>Summary:</b> Current vision-language-action (VLA) models, pre-trained on large-scale robotic data, exhibit strong multi-task capabilities and generalize well to variations in visual and language instructions for manipulation. However, their success rate drops significantly when faced with object concepts outside the training data, such as unseen object descriptions and textures in the dataset. To address this, we propose a novel agentic framework, VLA^2, which leverages OpenVLA as the execution backbone and effectively leverages external modules such as web retrieval and object detection to provide visual and textual knowledge about target objects to the VLA. This approach mitigates generalization failure when handling out-of-distribution objects. Based on the LIBERO simulation environment, we introduced novel objects and object descriptions to construct a new evaluation benchmark with three difficulty levels to test the effectiveness of our method. Our framework successfully outperformed the current state-of-the-art models on our designed hard-level generalization benchmark. Compared to the standalone OpenVLA baseline, VLA^2 achieves a 44.2% improvement in the success rate in the hard-level benchmark and an average improvement of 20.2% in all customized environments without any performance degradation on in-domain tasks. Project website: https://vla-2.github.io.</p> + https://arxiv.org/abs/2510.14902 + Thu, 16 Oct 2025 17:18:34 +0000 - InstructX: Towards Unified Visual Editing with MLLM Guidance - https://arxiv.org/abs/2510.08485 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08485.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Chong Mou, Qichao Sun, Yanze Wu, Pengze Zhang, Xinghui Li, Fulong Ye, Songtao Zhao, Qian He</p><p><b>Upvotes:</b> 14</p><p><b>Summary:</b> With recent advances in Multimodal Large Language Models (MLLMs) showing strong visual understanding and reasoning, interest is growing in using them to improve the editing performance of diffusion models. Despite rapid progress, most studies lack an in-depth analysis of MLLM design choices. Moreover, the integration of MLLMs and diffusion models remains an open challenge in some difficult tasks, such as video editing. In this paper, we present InstructX, a unified framework for image and video editing. Specifically, we conduct a comprehensive study on integrating MLLMs and diffusion models for instruction-driven editing across diverse tasks. Building on this study, we analyze the cooperation and distinction between images and videos in unified modeling. (1) We show that training on image data can lead to emergent video editing capabilities without explicit supervision, thereby alleviating the constraints imposed by scarce video training data. (2) By incorporating modality-specific MLLM features, our approach effectively unifies image and video editing tasks within a single model. Extensive experiments demonstrate that our method can handle a broad range of image and video editing tasks and achieves state-of-the-art performance.</p> - https://arxiv.org/abs/2510.08485 - Thu, 09 Oct 2025 17:26:09 +0000 + LaSeR: Reinforcement Learning with Last-Token Self-Rewarding + https://arxiv.org/abs/2510.14943 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14943.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wenkai Yang, Weijie Liu, Ruobing Xie, Yiju Guo, Lulu Wu, Saiyong Yang, Yankai Lin</p><p><b>Upvotes:</b> 37</p><p><b>Summary:</b> Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as a core paradigm for enhancing the reasoning capabilities of Large Language Models (LLMs). To address the lack of verification signals at test time, prior studies incorporate the training of model's self-verification capability into the standard RLVR process, thereby unifying reasoning and verification capabilities within a single LLM. However, previous practice requires the LLM to sequentially generate solutions and self-verifications using two separate prompt templates, which significantly reduces efficiency. In this work, we theoretically reveal that the closed-form solution to the RL objective of self-verification can be reduced to a remarkably simple form: the true reasoning reward of a solution is equal to its last-token self-rewarding score, which is computed as the difference between the policy model's next-token log-probability assigned to any pre-specified token at the solution's last token and a pre-calculated constant, scaled by the KL coefficient. Based on this insight, we propose LaSeR (Reinforcement Learning with Last-Token Self-Rewarding), an algorithm that simply augments the original RLVR loss with a MSE loss that aligns the last-token self-rewarding scores with verifier-based reasoning rewards, jointly optimizing the reasoning and self-rewarding capabilities of LLMs. The optimized self-rewarding scores can be utilized in both training and testing to enhance model performance. Notably, our algorithm derives these scores from the predicted next-token probability distribution of the last token immediately after generation, incurring only the minimal extra cost of one additional token inference. Experiments show that our method not only improves the model's reasoning performance but also equips it with remarkable self-rewarding capability, thereby boosting its inference-time scaling performance.</p> + https://arxiv.org/abs/2510.14943 + Thu, 16 Oct 2025 17:55:11 +0000 - CoMAS: Co-Evolving Multi-Agent Systems via Interaction Rewards - https://arxiv.org/abs/2510.08529 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08529.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xiangyuan Xue, Yifan Zhou, Guibin Zhang, Zaibin Zhang, Yijiang Li, Chen Zhang, Zhenfei Yin, Philip Torr, Wanli Ouyang, Lei Bai</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Self-evolution is a central research topic in enabling large language model (LLM)-based agents to continually improve their capabilities after pretraining. Recent research has witnessed a transition from reinforcement learning (RL)-free to RL-based methods. Current RL-based methods either rely on dense external reward signals or extract intrinsic reward signals from LLMs themselves. However, these approaches diverge from the self-evolution mechanisms observed in human intelligence, where individuals learn and improve through mutual discussion and collaboration. In this work, we introduce Co-Evolving Multi-Agent Systems (CoMAS), a novel framework that enables agents to improve autonomously by learning from inter-agent interactions without external supervision. CoMAS generates intrinsic rewards from rich discussion dynamics, employs an LLM-as-a-judge mechanism to formulate these rewards, and optimizes each agent's policy through RL, thereby enabling decentralized and scalable co-evolution. Experimental results demonstrate that CoMAS consistently outperforms untrained agents and achieves state-of-the-art performance across most evaluation settings. Ablation studies confirm the necessity of interaction-based reward signals and reveal promising scalability as the number and diversity of agents increase. These findings establish CoMAS as a novel and effective paradigm for self-evolution in LLM-based agents.</p> - https://arxiv.org/abs/2510.08529 - Thu, 09 Oct 2025 17:50:26 +0000 + MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning + https://arxiv.org/abs/2510.14958 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14958.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Weikang Shi, Aldrich Yu, Rongyao Fang, Houxing Ren, Ke Wang, Aojun Zhou, Changyao Tian, Xinyu Fu, Yuxuan Hu, Zimu Lu, Linjiang Huang, Si Liu, Rui Liu, Hongsheng Li</p><p><b>Upvotes:</b> 21</p><p><b>Summary:</b> While Large Language Models (LLMs) have excelled in textual reasoning, they struggle with mathematical domains like geometry that intrinsically rely on visual aids. Existing approaches to Visual Chain-of-Thought (VCoT) are often limited by rigid external tools or fail to generate the high-fidelity, strategically-timed diagrams necessary for complex problem-solving. To bridge this gap, we introduce MathCanvas, a comprehensive framework designed to endow unified Large Multimodal Models (LMMs) with intrinsic VCoT capabilities for mathematics. Our approach consists of two phases. First, a Visual Manipulation stage pre-trains the model on a novel 15.2M-pair corpus, comprising 10M caption-to-diagram pairs (MathCanvas-Imagen) and 5.2M step-by-step editing trajectories (MathCanvas-Edit), to master diagram generation and editing. Second, a Strategic Visual-Aided Reasoning stage fine-tunes the model on MathCanvas-Instruct, a new 219K-example dataset of interleaved visual-textual reasoning paths, teaching it when and how to leverage visual aids. To facilitate rigorous evaluation, we introduce MathCanvas-Bench, a challenging benchmark with 3K problems that require models to produce interleaved visual-textual solutions. Our model, BAGEL-Canvas, trained under this framework, achieves an 86% relative improvement over strong LMM baselines on MathCanvas-Bench, demonstrating excellent generalization to other public math benchmarks. Our work provides a complete toolkit-framework, datasets, and benchmark-to unlock complex, human-like visual-aided reasoning in LMMs. Project Page: https://mathcanvas.github.io/</p> + https://arxiv.org/abs/2510.14958 + Thu, 16 Oct 2025 17:58:58 +0000 - MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization - https://arxiv.org/abs/2510.08540 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08540.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xiangyu Zhao, Junming Lin, Tianhao Liang, Yifan Zhou, Wenhao Chai, Yuzhe Gu, Weiyun Wang, Kai Chen, Gen Luo, Wenwei Zhang, Junchi Yan, Hua Yang, Haodong Duan, Xue Yang</p><p><b>Upvotes:</b> 98</p><p><b>Summary:</b> While current Multimodal Large Language Models (MLLMs) have demonstrated proficiency in reasoning tasks such as mathematics and logic, their capacity for long-chain reflective reasoning, a prerequisite for solving complex real-world problems, remains largely underexplored. In this work, we first conduct an extensive empirical investigation to evaluate this capability. Leveraging a carefully designed data synthesis engine, we construct MM-HELIX, a multimodal benchmark consisting 1,260 samples of 42 challenging synthetic tasks that require iterative thinking and backtracking. Empirical results on this benchmark reveal that existing MLLMs exhibit significant performance deficits in long-chain reflective reasoning. To address this limitation, we generate post-training data and further explore learning paradigms for exploiting such data. We first develop the Step-Elicited Response Generation pipeline to create MM-HELIX-100K, a large-scale dataset of 100k high-quality, reflective reasoning traces for instruction-tuning stage. Given that standard Reinforcement Learning fails on complex tasks due to sparse reward signals and catastrophic forgetting after Supervised Fine-Tuning, we propose Adaptive Hybrid Policy Optimization (AHPO), a novel training strategy that dynamically unifies offline supervision and online optimization into a single stage. This strategy enables the model to learn from expert data when rewards are sparse and conduct independent exploration once proficient. When applied to the Qwen2.5-VL-7B baseline, our method achieves a +18.6\% accuracy improvement on MM-HELIX benchmark and demonstrates strong generalization with a +5.7\% average performance gain on general mathematic and logic tasks. Our work demonstrate that reflective reasoning in MLLMs can be effectively learned and generalized, paving the way for developing more capable MLLMs.</p> - https://arxiv.org/abs/2510.08540 - Thu, 09 Oct 2025 17:53:58 +0000 + Information Gain-based Policy Optimization: A Simple and Effective Approach for Multi-Turn LLM Agents + https://arxiv.org/abs/2510.14967 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14967.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Guoqing Wang, Sunhao Dai, Guangze Ye, Zeyu Gan, Wei Yao, Yong Deng, Xiaofeng Wu, Zhenzhe Ying</p><p><b>Upvotes:</b> 30</p><p><b>Summary:</b> Large language model (LLM)-based agents are increasingly trained with reinforcement learning (RL) to enhance their ability to interact with external environments through tool use, particularly in search-based settings that require multi-turn reasoning and knowledge acquisition. However, existing approaches typically rely on outcome-based rewards that are only provided at the final answer. This reward sparsity becomes particularly problematic in multi-turn settings, where long trajectories exacerbate two critical issues: (i) advantage collapse, where all rollouts receive identical rewards and provide no useful learning signals, and (ii) lack of fine-grained credit assignment, where dependencies between turns are obscured, especially in long-horizon tasks. In this paper, we propose Information Gain-based Policy Optimization (IGPO), a simple yet effective RL framework that provides dense and intrinsic supervision for multi-turn agent training. IGPO models each interaction turn as an incremental process of acquiring information about the ground truth, and defines turn-level rewards as the marginal increase in the policy's probability of producing the correct answer. Unlike prior process-level reward approaches that depend on external reward models or costly Monte Carlo estimation, IGPO derives intrinsic rewards directly from the model's own belief updates. These intrinsic turn-level rewards are combined with outcome-level supervision to form dense reward trajectories. Extensive experiments on both in-domain and out-of-domain benchmarks demonstrate that IGPO consistently outperforms strong baselines in multi-turn scenarios, achieving higher accuracy and improved sample efficiency.</p> + https://arxiv.org/abs/2510.14967 + Thu, 16 Oct 2025 17:59:32 +0000 - ARTDECO: Towards Efficient and High-Fidelity On-the-Fly 3D Reconstruction with Structured Scene Representation - https://arxiv.org/abs/2510.08551 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08551.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Guanghao Li, Kerui Ren, Linning Xu, Zhewen Zheng, Changjian Jiang, Xin Gao, Bo Dai, Jian Pu, Mulin Yu, Jiangmiao Pang</p><p><b>Upvotes:</b> 26</p><p><b>Summary:</b> On-the-fly 3D reconstruction from monocular image sequences is a long-standing challenge in computer vision, critical for applications such as real-to-sim, AR/VR, and robotics. Existing methods face a major tradeoff: per-scene optimization yields high fidelity but is computationally expensive, whereas feed-forward foundation models enable real-time inference but struggle with accuracy and robustness. In this work, we propose ARTDECO, a unified framework that combines the efficiency of feed-forward models with the reliability of SLAM-based pipelines. ARTDECO uses 3D foundation models for pose estimation and point prediction, coupled with a Gaussian decoder that transforms multi-scale features into structured 3D Gaussians. To sustain both fidelity and efficiency at scale, we design a hierarchical Gaussian representation with a LoD-aware rendering strategy, which improves rendering fidelity while reducing redundancy. Experiments on eight diverse indoor and outdoor benchmarks show that ARTDECO delivers interactive performance comparable to SLAM, robustness similar to feed-forward systems, and reconstruction quality close to per-scene optimization, providing a practical path toward on-the-fly digitization of real-world environments with both accurate geometry and high visual fidelity. Explore more demos on our project page: https://city-super.github.io/artdeco/.</p> - https://arxiv.org/abs/2510.08551 - Thu, 09 Oct 2025 17:57:38 +0000 + TokDrift: When LLM Speaks in Subwords but Code Speaks in Grammar + https://arxiv.org/abs/2510.14972 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14972.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yinxi Li, Yuntian Deng, Pengyu Nie</p><p><b>Upvotes:</b> 28</p><p><b>Summary:</b> Large language models (LLMs) for code rely on subword tokenizers, such as byte-pair encoding (BPE), learned from mixed natural language text and programming language code but driven by statistics rather than grammar. As a result, semantically identical code snippets can be tokenized differently depending on superficial factors such as whitespace or identifier naming. To measure the impact of this misalignment, we introduce TokDrift, a framework that applies semantic-preserving rewrite rules to create code variants differing only in tokenization. Across nine code LLMs, including large ones with over 30B parameters, even minor formatting changes can cause substantial shifts in model behavior. Layer-wise analysis shows that the issue originates in early embeddings, where subword segmentation fails to capture grammar token boundaries. Our findings identify misaligned tokenization as a hidden obstacle to reliable code understanding and generation, highlighting the need for grammar-aware tokenization for future code LLMs.</p> + https://arxiv.org/abs/2510.14972 + Thu, 16 Oct 2025 17:59:45 +0000 - VideoCanvas: Unified Video Completion from Arbitrary Spatiotemporal Patches via In-Context Conditioning - https://arxiv.org/abs/2510.08555 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08555.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Minghong Cai, Qiulin Wang, Zongli Ye, Wenze Liu, Quande Liu, Weicai Ye, Xintao Wang, Pengfei Wan, Kun Gai, Xiangyu Yue</p><p><b>Upvotes:</b> 56</p><p><b>Summary:</b> We introduce the task of arbitrary spatio-temporal video completion, where a video is generated from arbitrary, user-specified patches placed at any spatial location and timestamp, akin to painting on a video canvas. This flexible formulation naturally unifies many existing controllable video generation tasks--including first-frame image-to-video, inpainting, extension, and interpolation--under a single, cohesive paradigm. Realizing this vision, however, faces a fundamental obstacle in modern latent video diffusion models: the temporal ambiguity introduced by causal VAEs, where multiple pixel frames are compressed into a single latent representation, making precise frame-level conditioning structurally difficult. We address this challenge with VideoCanvas, a novel framework that adapts the In-Context Conditioning (ICC) paradigm to this fine-grained control task with zero new parameters. We propose a hybrid conditioning strategy that decouples spatial and temporal control: spatial placement is handled via zero-padding, while temporal alignment is achieved through Temporal RoPE Interpolation, which assigns each condition a continuous fractional position within the latent sequence. This resolves the VAE's temporal ambiguity and enables pixel-frame-aware control on a frozen backbone. To evaluate this new capability, we develop VideoCanvasBench, the first benchmark for arbitrary spatio-temporal video completion, covering both intra-scene fidelity and inter-scene creativity. Experiments demonstrate that VideoCanvas significantly outperforms existing conditioning paradigms, establishing a new state of the art in flexible and unified video generation.</p> - https://arxiv.org/abs/2510.08555 - Thu, 09 Oct 2025 17:58:59 +0000 + Attention Is All You Need for KV Cache in Diffusion LLMs + https://arxiv.org/abs/2510.14973 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14973.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Quan Nguyen-Tri, Mukul Ranjan, Zhiqiang Shen</p><p><b>Upvotes:</b> 33</p><p><b>Summary:</b> This work studies how to adaptively recompute key-value (KV) caches for diffusion large language models (DLMs) to maximize prediction accuracy while minimizing decoding latency. Prior methods' decoders recompute QKV for all tokens at every denoising step and layer, despite KV states changing little across most steps, especially in shallow layers, leading to substantial redundancy. We make three observations: (1) distant {bf MASK} tokens primarily act as a length-bias and can be cached block-wise beyond the active prediction window; (2) KV dynamics increase with depth, suggesting that selective refresh starting from deeper layers is sufficient; and (3) the most-attended token exhibits the smallest KV drift, providing a conservative lower bound on cache change for other tokens. Building on these, we propose {bf Elastic-Cache}, a training-free, architecture-agnostic strategy that jointly decides {when} to refresh (via an attention-aware drift test on the most-attended token) and {where} to refresh (via a depth-aware schedule that recomputes from a chosen layer onward while reusing shallow-layer caches and off-window MASK caches). Unlike fixed-period schemes, Elastic-Cache performs adaptive, layer-aware cache updates for diffusion LLMs, reducing redundant computation and accelerating decoding with negligible loss in generation quality. Experiments on LLaDA-Instruct, LLaDA-1.5, and LLaDA-V across mathematical reasoning and code generation tasks demonstrate consistent speedups: 8.7times on GSM8K (256 tokens), 45.1times on longer sequences, and 4.8times on HumanEval, while consistently maintaining higher accuracy than the baseline. Our method achieves significantly higher throughput (6.8times on GSM8K) than existing confidence-based approaches while preserving generation quality, enabling practical deployment of diffusion LLMs.</p> + https://arxiv.org/abs/2510.14973 + Thu, 16 Oct 2025 17:59:48 +0000 - Agent Learning via Early Experience - https://arxiv.org/abs/2510.08558 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08558.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kai Zhang, Xiangchao Chen, Bo Liu, Tianci Xue, Zeyi Liao, Zhihan Liu, Xiyao Wang, Yuting Ning, Zhaorun Chen, Xiaohan Fu, Jian Xie, Yuxuan Sun, Boyu Gou, Qi Qi, Zihang Meng, Jianwei Yang, Ning Zhang, Xian Li, Ashish Shah, Dat Huynh, Hengduo Li, Zi Yang, Sara Cao, Lawrence Jang, Shuyan Zhou, Jiacheng Zhu, Huan Sun, Jason Weston, Yu Su, Yifan Wu</p><p><b>Upvotes:</b> 175</p><p><b>Summary:</b> A long-term goal of language agents is to learn and improve through their own experience, ultimately outperforming humans in complex, real-world tasks. However, training agents from experience data with reinforcement learning remains difficult in many environments, which either lack verifiable rewards (e.g., websites) or require inefficient long-horizon rollouts (e.g., multi-turn tool use). As a result, most current agents rely on supervised fine-tuning on expert data, which is challenging to scale and generalizes poorly. This limitation stems from the nature of expert demonstrations: they capture only a narrow range of scenarios and expose the agent to limited environment diversity. We address this limitation with a middle-ground paradigm we call early experience: interaction data generated by the agent's own actions, where the resulting future states serve as supervision without reward signals. Within this paradigm we study two strategies of using such data: (1) Implicit world modeling, which uses collected states to ground the policy in environment dynamics; and (2) Self-reflection, where the agent learns from its suboptimal actions to improve reasoning and decision-making. We evaluate across eight diverse environments and multiple model families. Our approaches consistently improve effectiveness and out-of-domain generalization, highlighting the value of early experience. Moreover, in environments with verifiable rewards, our results provide promising signals that early experience offers a strong foundation for subsequent reinforcement learning, positioning it as a practical bridge between imitation learning and fully experience-driven agents.</p> - https://arxiv.org/abs/2510.08558 - Thu, 09 Oct 2025 17:59:17 +0000 + WithAnyone: Towards Controllable and ID Consistent Image Generation + https://arxiv.org/abs/2510.14975 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14975.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hengyuan Xu, Wei Cheng, Peng Xing, Yixiao Fang, Shuhan Wu, Rui Wang, Xianfang Zeng, Daxin Jiang, Gang Yu, Xingjun Ma, Yu-Gang Jiang</p><p><b>Upvotes:</b> 72</p><p><b>Summary:</b> Identity-consistent generation has become an important focus in text-to-image research, with recent models achieving notable success in producing images aligned with a reference identity. Yet, the scarcity of large-scale paired datasets containing multiple images of the same individual forces most approaches to adopt reconstruction-based training. This reliance often leads to a failure mode we term copy-paste, where the model directly replicates the reference face rather than preserving identity across natural variations in pose, expression, or lighting. Such over-similarity undermines controllability and limits the expressive power of generation. To address these limitations, we (1) construct a large-scale paired dataset MultiID-2M, tailored for multi-person scenarios, providing diverse references for each identity; (2) introduce a benchmark that quantifies both copy-paste artifacts and the trade-off between identity fidelity and variation; and (3) propose a novel training paradigm with a contrastive identity loss that leverages paired data to balance fidelity with diversity. These contributions culminate in WithAnyone, a diffusion-based model that effectively mitigates copy-paste while preserving high identity similarity. Extensive qualitative and quantitative experiments demonstrate that WithAnyone significantly reduces copy-paste artifacts, improves controllability over pose and expression, and maintains strong perceptual quality. User studies further validate that our method achieves high identity fidelity while enabling expressive controllable generation.</p> + https://arxiv.org/abs/2510.14975 + Thu, 16 Oct 2025 17:59:54 +0000 - NaViL: Rethinking Scaling Properties of Native Multimodal Large Language Models under Data Constraints - https://arxiv.org/abs/2510.08565 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08565.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Changyao Tian, Hao Li, Gen Luo, Xizhou Zhu, Weijie Su, Hanming Deng, Jinguo Zhu, Jie Shao, Ziran Zhu, Yunpeng Liu, Lewei Lu, Wenhai Wang, Hongsheng Li, Jifeng Dai</p><p><b>Upvotes:</b> 17</p><p><b>Summary:</b> Compositional training has been the de-facto paradigm in existing Multimodal Large Language Models (MLLMs), where pre-trained vision encoders are connected with pre-trained LLMs through continuous multimodal pre-training. However, the multimodal scaling property of this paradigm remains difficult to explore due to the separated training. In this paper, we focus on the native training of MLLMs in an end-to-end manner and systematically study its design space and scaling property under a practical setting, i.e., data constraint. Through careful study of various choices in MLLM, we obtain the optimal meta-architecture that best balances performance and training cost. After that, we further explore the scaling properties of the native MLLM and indicate the positively correlated scaling relationship between visual encoders and LLMs. Based on these findings, we propose a native MLLM called NaViL, combined with a simple and cost-effective recipe. Experimental results on 14 multimodal benchmarks confirm the competitive performance of NaViL against existing MLLMs. Besides that, our findings and results provide in-depth insights for the future study of native MLLMs.</p> - https://arxiv.org/abs/2510.08565 - Thu, 09 Oct 2025 17:59:37 +0000 + From Pixels to Words -- Towards Native Vision-Language Primitives at Scale + https://arxiv.org/abs/2510.14979 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.14979.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Haiwen Diao, Mingxuan Li, Silei Wu, Linjun Dai, Xiaohua Wang, Hanming Deng, Lewei Lu, Dahua Lin, Ziwei Liu</p><p><b>Upvotes:</b> 59</p><p><b>Summary:</b> The edifice of native Vision-Language Models (VLMs) has emerged as a rising contender to typical modular VLMs, shaped by evolving model architectures and training paradigms. Yet, two lingering clouds cast shadows over its widespread exploration and promotion: (-) What fundamental constraints set native VLMs apart from modular ones, and to what extent can these barriers be overcome? (-) How to make research in native VLMs more accessible and democratized, thereby accelerating progress in the field. In this paper, we clarify these challenges and outline guiding principles for constructing native VLMs. Specifically, one native VLM primitive should: (i) effectively align pixel and word representations within a shared semantic space; (ii) seamlessly integrate the strengths of formerly separate vision and language modules; (iii) inherently embody various cross-modal properties that support unified vision-language encoding, aligning, and reasoning. Hence, we launch NEO, a novel family of native VLMs built from first principles, capable of rivaling top-tier modular counterparts across diverse real-world scenarios. With only 390M image-text examples, NEO efficiently develops visual perception from scratch while mitigating vision-language conflicts inside a dense and monolithic model crafted from our elaborate primitives. We position NEO as a cornerstone for scalable and powerful native VLMs, paired with a rich set of reusable components that foster a cost-effective and extensible ecosystem. Our code and models are publicly available at: https://github.com/EvolvingLMMs-Lab/NEO.</p> + https://arxiv.org/abs/2510.14979 + Thu, 16 Oct 2025 17:59:58 +0000