From 832ab9c7025665f99fc122a39920e4c871c161c0 Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" Date: Sat, 11 Oct 2025 00:02:28 +0000 Subject: [PATCH] bot: update RSS feed --- feed.xml | 478 +++++++++++++++++++++++++++++++------------------------ 1 file changed, 267 insertions(+), 211 deletions(-) diff --git a/feed.xml b/feed.xml index 168d983..dd0dc1d 100644 --- a/feed.xml +++ b/feed.xml @@ -7,300 +7,356 @@ http://www.rssboard.org/rss-specification python-feedgen en - Fri, 10 Oct 2025 00:02:54 +0000 + Sat, 11 Oct 2025 00:02:28 +0000 - DeepTravel: An End-to-End Agentic Reinforcement Learning Framework for Autonomous Travel Planning Agents - https://arxiv.org/abs/2509.21842 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.21842.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yansong Ning, Rui Liu, Jun Wang, Kai Chen, Wei Li, Jun Fang, Kan Zheng, Naiqiang Tan, Hao Liu</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Travel planning (TP) agent has recently worked as an emerging building block to interact with external tools and resources for travel itinerary generation, ensuring enjoyable user experience. Despite its benefits, existing studies rely on hand craft prompt and fixed agent workflow, hindering more flexible and autonomous TP agent. This paper proposes DeepTravel, an end to end agentic reinforcement learning framework for building autonomous travel planning agent, capable of autonomously planning, executing tools, and reflecting on tool responses to explore, verify, and refine intermediate actions in multi step reasoning. To achieve this, we first construct a robust sandbox environment by caching transportation, accommodation and POI data, facilitating TP agent training without being constrained by real world APIs limitations (e.g., inconsistent outputs). Moreover, we develop a hierarchical reward modeling system, where a trajectory level verifier first checks spatiotemporal feasibility and filters unsatisfied travel itinerary, and then the turn level verifier further validate itinerary detail consistency with tool responses, enabling efficient and precise reward service. Finally, we propose the reply augmented reinforcement learning method that enables TP agent to periodically replay from a failures experience buffer, emerging notable agentic capacity. We deploy trained TP agent on DiDi Enterprise Solutions App and conduct comprehensive online and offline evaluations, demonstrating that DeepTravel enables small size LLMs (e.g., Qwen3 32B) to significantly outperform existing frontier LLMs such as OpenAI o1, o3 and DeepSeek R1 in travel planning tasks.</p> - https://arxiv.org/abs/2509.21842 - Fri, 26 Sep 2025 04:03:52 +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> 38</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 - 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> 4</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 + Beyond Outliers: A Study of Optimizers Under Quantization + https://arxiv.org/abs/2509.23500 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.23500.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Georgios Vlassis, Saleh Ashkboos, Alexandra Volkova, Torsten Hoefler, Dan Alistarh</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> As new optimizers gain traction and model quantization becomes standard for efficient deployment, a key question arises: how does the choice of optimizer affect model performance in the presence of quantization? Despite progress in both areas, systematic evidence on optimizer-quantization interactions remains limited. To fill this gap, we study the impact of optimizer choice on model robustness under quantization, considering both post-training quantization (PTQ), and quantization-aware training (QAT). We first train full-precision models, ranging from 50M to 1.5B parameters, with six optimizers, to explore the hyperparameter landscape, and establish well-tuned baselines. We then apply PTQ to evaluate how model performance degrades when trained with different optimizers. We find that outlier-related metrics, such as the max-to-mean ratio (MMR) and Kurtosis, fail to predict the PTQ performance across different optimizers. We show analytically that this is due to the MMR capturing only isolated layer errors, while ignoring how quantization errors accumulate and propagate through the network. To study the QAT degradation, we train quantized models from scratch and compare them to our original-precision baselines. We find that optimizers performing well in the original pretraining setup may not remain optimal under QAT, and that models trained with Shampoo show the lowest accuracy degradation. Finally, we derive scaling laws for quantization-aware training under different optimizers, showing that Shampoo achieves the highest parameter efficiency of all tested optimizers.</p> + https://arxiv.org/abs/2509.23500 + Sat, 27 Sep 2025 21:15:22 +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 + 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> 42</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 - G^2RPO: Granular GRPO for Precise Reward in Flow Models - https://arxiv.org/abs/2510.01982 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.01982.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yujie Zhou, Pengyang Ling, Jiazi Bu, Yibin Wang, Yuhang Zang, Jiaqi Wang, Li Niu, Guangtao Zhai</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> The integration of online reinforcement learning (RL) into diffusion and flow models has recently emerged as a promising approach for aligning generative models with human preferences. Stochastic sampling via Stochastic Differential Equations (SDE) is employed during the denoising process to generate diverse denoising directions for RL exploration. While existing methods effectively explore potential high-value samples, they suffer from sub-optimal preference alignment due to sparse and narrow reward signals. To address these challenges, we propose a novel Granular-GRPO (G^2RPO ) framework that achieves precise and comprehensive reward assessments of sampling directions in reinforcement learning of flow models. Specifically, a Singular Stochastic Sampling strategy is introduced to support step-wise stochastic exploration while enforcing a high correlation between the reward and the injected noise, thereby facilitating a faithful reward for each SDE perturbation. Concurrently, to eliminate the bias inherent in fixed-granularity denoising, we introduce a Multi-Granularity Advantage Integration module that aggregates advantages computed at multiple diffusion scales, producing a more comprehensive and robust evaluation of the sampling directions. Experiments conducted on various reward models, including both in-domain and out-of-domain evaluations, demonstrate that our G^2RPO significantly outperforms existing flow-based GRPO baselines,highlighting its effectiveness and robustness.</p> - https://arxiv.org/abs/2510.01982 - Thu, 02 Oct 2025 12:57:12 +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> 55</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 - A Single Character can Make or Break Your LLM Evals - https://arxiv.org/abs/2510.05152 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05152.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jingtong Su, Jianyu Zhang, Karen Ullrich, Léon Bottou, Mark Ibrahim</p><p><b>Upvotes:</b> 0</p><p><b>Summary:</b> Common Large Language model (LLM) evaluations rely on demonstration examples to steer models' responses to the desired style. While the number of examples used has been studied and standardized, the choice of how to format examples is less investigated. In evaluation protocols and real world usage, users face the choice how to separate in-context examples: use a comma? new line? semi-colon? hashtag? etc.? Surprisingly, we find this seemingly minor choice can dramatically alter model response quality. Across leading model families (Llama, Qwen, Gemma), performance on MMLU for example can vary by pm 23% depending on the choice of delimiter. In fact, one can manipulate model rankings to put any model in the lead by only modifying the single character separating examples. We find LLMs' brittleness pervades topics, model families, and doesn't improve with scale. By probing attention head scores, we find that good-performing delimiters steer attention towards key tokens in the input. Finally, we explore methods to improve LLMs' robustness to the choice of delimiter. We find specifying the selected delimiter in the prompt boosts robustness and offer practical recommendations for the best-performing delimiters to select.</p> - https://arxiv.org/abs/2510.05152 - Thu, 02 Oct 2025 13:27:28 +0000 + Fidelity-Aware Data Composition for Robust Robot Generalization + https://arxiv.org/abs/2509.24797 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.24797.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zizhao Tong, Di Chen, Sicheng Hu, Hongwei Fan, Liliang Chen, Guanghui Ren, Hao Tang, Hao Dong, Ling Shao</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> Generalist robot policies trained on large-scale, visually homogeneous datasets can be susceptible to shortcut learning, which impairs their out-of-distribution (OOD) generalization. While generative data augmentation is a common approach to introduce diversity, it presents a subtle challenge: data composition. Naively mixing real and synthetic data can corrupt the learning signal, as this process often prioritizes visual diversity at the expense of information fidelity. This paper suggests that robust generalization depends on principled, fidelity-aware data composition. We introduce Coherent Information Fidelity Tuning (CIFT), a framework that treats data composition as an optimization problem. CIFT uses a practical proxy for Information Fidelity based on the feature-space geometry of a dataset. This enables the identification of a phase transition, termed the Decoherence Point, where training stability degrades. The framework includes a generative engine, Multi-View Video Augmentation (MVAug), to synthesize a causally disentangled data spectrum for this tuning process. Applying CIFT to policy architectures such as pi_0 and Diffusion Policy improves OOD success rates by over 54\%. These results indicate that fidelity-aware composition, beyond data synthesis alone, is an important component for developing robust, general-purpose robots.</p> + https://arxiv.org/abs/2509.24797 + Mon, 29 Sep 2025 13:48:36 +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> 63</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 + UP2You: Fast Reconstruction of Yourself from Unconstrained Photo Collections + https://arxiv.org/abs/2509.24817 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.24817.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zeyu Cai, Ziyang Li, Xiaoben Li, Boqian Li, Zeyu Wang, Zhenyu Zhang, Yuliang Xiu</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> We present UP2You, the first tuning-free solution for reconstructing high-fidelity 3D clothed portraits from extremely unconstrained in-the-wild 2D photos. Unlike previous approaches that require "clean" inputs (e.g., full-body images with minimal occlusions, or well-calibrated cross-view captures), UP2You directly processes raw, unstructured photographs, which may vary significantly in pose, viewpoint, cropping, and occlusion. Instead of compressing data into tokens for slow online text-to-3D optimization, we introduce a data rectifier paradigm that efficiently converts unconstrained inputs into clean, orthogonal multi-view images in a single forward pass within seconds, simplifying the 3D reconstruction. Central to UP2You is a pose-correlated feature aggregation module (PCFA), that selectively fuses information from multiple reference images w.r.t. target poses, enabling better identity preservation and nearly constant memory footprint, with more observations. We also introduce a perceiver-based multi-reference shape predictor, removing the need for pre-captured body templates. Extensive experiments on 4D-Dress, PuzzleIOI, and in-the-wild captures demonstrate that UP2You consistently surpasses previous methods in both geometric accuracy (Chamfer-15%, P2S-18% on PuzzleIOI) and texture fidelity (PSNR-21%, LPIPS-46% on 4D-Dress). UP2You is efficient (1.5 minutes per person), and versatile (supports arbitrary pose control, and training-free multi-garment 3D virtual try-on), making it practical for real-world scenarios where humans are casually captured. Both models and code will be released to facilitate future research on this underexplored task. Project Page: https://zcai0612.github.io/UP2You</p> + https://arxiv.org/abs/2509.24817 + Mon, 29 Sep 2025 14:06:00 +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> 17</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 + OmniRetarget: Interaction-Preserving Data Generation for Humanoid Whole-Body Loco-Manipulation and Scene Interaction + https://arxiv.org/abs/2509.26633 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2509.26633.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Lujie Yang, Xiaoyu Huang, Zhen Wu, Angjoo Kanazawa, Pieter Abbeel, Carmelo Sferrazza, C. Karen Liu, Rocky Duan, Guanya Shi</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> A dominant paradigm for teaching humanoid robots complex skills is to retarget human motions as kinematic references to train reinforcement learning (RL) policies. However, existing retargeting pipelines often struggle with the significant embodiment gap between humans and robots, producing physically implausible artifacts like foot-skating and penetration. More importantly, common retargeting methods neglect the rich human-object and human-environment interactions essential for expressive locomotion and loco-manipulation. To address this, we introduce OmniRetarget, an interaction-preserving data generation engine based on an interaction mesh that explicitly models and preserves the crucial spatial and contact relationships between an agent, the terrain, and manipulated objects. By minimizing the Laplacian deformation between the human and robot meshes while enforcing kinematic constraints, OmniRetarget generates kinematically feasible trajectories. Moreover, preserving task-relevant interactions enables efficient data augmentation, from a single demonstration to different robot embodiments, terrains, and object configurations. We comprehensively evaluate OmniRetarget by retargeting motions from OMOMO, LAFAN1, and our in-house MoCap datasets, generating over 8-hour trajectories that achieve better kinematic constraint satisfaction and contact preservation than widely used baselines. Such high-quality data enables proprioceptive RL policies to successfully execute long-horizon (up to 30 seconds) parkour and loco-manipulation skills on a Unitree G1 humanoid, trained with only 5 reward terms and simple domain randomization shared by all tasks, without any learning curriculum.</p> + https://arxiv.org/abs/2509.26633 + Tue, 30 Sep 2025 17:59:02 +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> 17</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 + Use the Online Network If You Can: Towards Fast and Stable Reinforcement Learning + https://arxiv.org/abs/2510.02590 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.02590.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ahmed Hendawy, Henrik Metternich, Théo Vincent, Mahdi Kallel, Jan Peters, Carlo D'Eramo</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> The use of target networks is a popular approach for estimating value functions in deep Reinforcement Learning (RL). While effective, the target network remains a compromise solution that preserves stability at the cost of slowly moving targets, thus delaying learning. Conversely, using the online network as a bootstrapped target is intuitively appealing, albeit well-known to lead to unstable learning. In this work, we aim to obtain the best out of both worlds by introducing a novel update rule that computes the target using the MINimum estimate between the Target and Online network, giving rise to our method, MINTO. Through this simple, yet effective modification, we show that MINTO enables faster and stable value function learning, by mitigating the potential overestimation bias of using the online network for bootstrapping. Notably, MINTO can be seamlessly integrated into a wide range of value-based and actor-critic algorithms with a negligible cost. We evaluate MINTO extensively across diverse benchmarks, spanning online and offline RL, as well as discrete and continuous action spaces. Across all benchmarks, MINTO consistently improves performance, demonstrating its broad applicability and effectiveness.</p> + https://arxiv.org/abs/2510.02590 + Thu, 02 Oct 2025 21:48:01 +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> 15</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 + Towards Scalable and Consistent 3D Editing + https://arxiv.org/abs/2510.02994 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.02994.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ruihao Xia, Yang Tang, Pan Zhou</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> 3D editing - the task of locally modifying the geometry or appearance of a 3D asset - has wide applications in immersive content creation, digital entertainment, and AR/VR. However, unlike 2D editing, it remains challenging due to the need for cross-view consistency, structural fidelity, and fine-grained controllability. Existing approaches are often slow, prone to geometric distortions, or dependent on manual and accurate 3D masks that are error-prone and impractical. To address these challenges, we advance both the data and model fronts. On the data side, we introduce 3DEditVerse, the largest paired 3D editing benchmark to date, comprising 116,309 high-quality training pairs and 1,500 curated test pairs. Built through complementary pipelines of pose-driven geometric edits and foundation model-guided appearance edits, 3DEditVerse ensures edit locality, multi-view consistency, and semantic alignment. On the model side, we propose 3DEditFormer, a 3D-structure-preserving conditional transformer. By enhancing image-to-3D generation with dual-guidance attention and time-adaptive gating, 3DEditFormer disentangles editable regions from preserved structure, enabling precise and consistent edits without requiring auxiliary 3D masks. Extensive experiments demonstrate that our framework outperforms state-of-the-art baselines both quantitatively and qualitatively, establishing a new standard for practical and scalable 3D editing. Dataset and code will be released. Project: https://www.lv-lab.org/3DEditFormer/</p> + https://arxiv.org/abs/2510.02994 + Fri, 03 Oct 2025 13:34:55 +0000 - AlphaApollo: Orchestrating Foundation Models and Professional Tools into a Self-Evolving System for Deep Agentic Reasoning - https://arxiv.org/abs/2510.06261 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06261.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zhanke Zhou, Chentao Cao, Xiao Feng, Xuan Li, Zongze Li, Xiangyu Lu, Jiangchao Yao, Weikai Huang, Linrui Xu, Tian Cheng, Guanyu Jiang, Yiming Zheng, Brando Miranda, Tongliang Liu, Sanmi Koyejo, Masashi Sugiyama, Bo Han</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> We present AlphaApollo, a self-evolving agentic reasoning system that aims to address two bottlenecks in foundation model (FM) reasoning-limited model-intrinsic capacity and unreliable test-time iteration. AlphaApollo orchestrates multiple models with professional tools to enable deliberate, verifiable reasoning. It couples (i) a computation tool (Python with numerical and symbolic libraries) and (ii) a retrieval tool (task-relevant external information) to execute exact calculations and ground decisions. The system further supports multi-round, multi-model solution evolution via a shared state map that records candidates, executable checks, and feedback for iterative refinement. In evaluations on AIME 2024/2025 across multiple models, AlphaApollo delivers consistent gains: +5.15% Average@32 and +23.34% Pass@32 for Qwen2.5-14B-Instruct, and +8.91% Average@32 with +26.67% Pass@32 for Llama-3.3-70B-Instruct. Tool-use analysis shows that more than 80% of tool calls are successfully executed, with consistent outperformance of non-tool baselines, thereby lifting the capability ceiling of FMs. More empirical results and implementation details will be updated at https://github.com/tmlr-group/AlphaApollo.</p> - https://arxiv.org/abs/2510.06261 - Sun, 05 Oct 2025 15:42:24 +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 - 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> 18</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 + 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> 14</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 - Glocal Information Bottleneck for Time Series Imputation - https://arxiv.org/abs/2510.04910 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04910.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jie Yang, Kexin Zhang, Guibin Zhang, Philip S. Yu, Kaize Ding</p><p><b>Upvotes:</b> 0</p><p><b>Summary:</b> Time Series Imputation (TSI), which aims to recover missing values in temporal data, remains a fundamental challenge due to the complex and often high-rate missingness in real-world scenarios. Existing models typically optimize the point-wise reconstruction loss, focusing on recovering numerical values (local information). However, we observe that under high missing rates, these models still perform well in the training phase yet produce poor imputations and distorted latent representation distributions (global information) in the inference phase. This reveals a critical optimization dilemma: current objectives lack global guidance, leading models to overfit local noise and fail to capture global information of the data. To address this issue, we propose a new training paradigm, Glocal Information Bottleneck (Glocal-IB). Glocal-IB is model-agnostic and extends the standard IB framework by introducing a Global Alignment loss, derived from a tractable mutual information approximation. This loss aligns the latent representations of masked inputs with those of their originally observed counterparts. It helps the model retain global structure and local details while suppressing noise caused by missing values, giving rise to better generalization under high missingness. Extensive experiments on nine datasets confirm that Glocal-IB leads to consistently improved performance and aligned latent representations under missingness. Our code implementation is available in https://github.com/Muyiiiii/NeurIPS-25-Glocal-IB.</p> - https://arxiv.org/abs/2510.04910 - Mon, 06 Oct 2025 15:24:44 +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> 13</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 - Bridging Text and Video Generation: A Survey - https://arxiv.org/abs/2510.04999 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.04999.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Nilay Kumar, Priyansh Bhandari, G. Maragatham</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Text-to-video (T2V) generation technology holds potential to transform multiple domains such as education, marketing, entertainment, and assistive technologies for individuals with visual or reading comprehension challenges, by creating coherent visual content from natural language prompts. From its inception, the field has advanced from adversarial models to diffusion-based models, yielding higher-fidelity, temporally consistent outputs. Yet challenges persist, such as alignment, long-range coherence, and computational efficiency. Addressing this evolving landscape, we present a comprehensive survey of text-to-video generative models, tracing their development from early GANs and VAEs to hybrid Diffusion-Transformer (DiT) architectures, detailing how these models work, what limitations they addressed in their predecessors, and why shifts toward new architectural paradigms were necessary to overcome challenges in quality, coherence, and control. We provide a systematic account of the datasets, which the surveyed text-to-video models were trained and evaluated on, and, to support reproducibility and assess the accessibility of training such models, we detail their training configurations, including their hardware specifications, GPU counts, batch sizes, learning rates, optimizers, epochs, and other key hyperparameters. Further, we outline the evaluation metrics commonly used for evaluating such models and present their performance across standard benchmarks, while also discussing the limitations of these metrics and the emerging shift toward more holistic, perception-aligned evaluation strategies. Finally, drawing from our analysis, we outline the current open challenges and propose a few promising future directions, laying out a perspective for future researchers to explore and build upon in advancing T2V research and applications.</p> - https://arxiv.org/abs/2510.04999 - Mon, 06 Oct 2025 16:39:05 +0000 + Drive&Gen: Co-Evaluating End-to-End Driving and Video Generation Models + https://arxiv.org/abs/2510.06209 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06209.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Jiahao Wang, Zhenpei Yang, Yijing Bai, Yingwei Li, Yuliang Zou, Bo Sun, Abhijit Kundu, Jose Lezama, Luna Yue Huang, Zehao Zhu, Jyh-Jing Hwang, Dragomir Anguelov, Mingxing Tan, Chiyu Max Jiang</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Recent advances in generative models have sparked exciting new possibilities in the field of autonomous vehicles. Specifically, video generation models are now being explored as controllable virtual testing environments. Simultaneously, end-to-end (E2E) driving models have emerged as a streamlined alternative to conventional modular autonomous driving systems, gaining popularity for their simplicity and scalability. However, the application of these techniques to simulation and planning raises important questions. First, while video generation models can generate increasingly realistic videos, can these videos faithfully adhere to the specified conditions and be realistic enough for E2E autonomous planner evaluation? Second, given that data is crucial for understanding and controlling E2E planners, how can we gain deeper insights into their biases and improve their ability to generalize to out-of-distribution scenarios? In this work, we bridge the gap between the driving models and generative world models (Drive&Gen) to address these questions. We propose novel statistical measures leveraging E2E drivers to evaluate the realism of generated videos. By exploiting the controllability of the video generation model, we conduct targeted experiments to investigate distribution gaps affecting E2E planner performance. Finally, we show that synthetic data produced by the video generation model offers a cost-effective alternative to real-world data collection. This synthetic data effectively improves E2E model generalization beyond existing Operational Design Domains, facilitating the expansion of autonomous vehicle services into new operational contexts.</p> + https://arxiv.org/abs/2510.06209 + Tue, 07 Oct 2025 17:58:32 +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> 10</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 + 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> 2</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 - NorMuon: Making Muon more efficient and scalable - https://arxiv.org/abs/2510.05491 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05491.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zichong Li, Liming Liu, Chen Liang, Weizhu Chen, Tuo Zhao</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> The choice of optimizer significantly impacts the training efficiency and computational costs of large language models (LLMs). Recently, the Muon optimizer has demonstrated promising results by orthogonalizing parameter updates, improving optimization geometry through better conditioning. Despite Muon's emergence as a candidate successor to Adam, the potential for jointly leveraging their strengths has not been systematically explored. In this work, we bridge this gap by proposing NorMuon (Neuron-wise Normalized Muon), an optimizer that synergistically combines orthogonalization with neuron-level adaptive learning rates. Our analysis reveals that while Muon effectively reduces condition numbers, the resulting updates exhibit highly non-uniform neuron norms, causing certain neurons to dominate the optimization process. NorMuon addresses this imbalance by maintaining second-order momentum statistics for each neuron and applying row-wise normalization after orthogonalization, ensuring balanced parameter utilization while preserving Muon's conditioning benefits. To enable practical deployment at scale, we develop an efficient distributed implementation under the FSDP2 framework that strategically distributes orthogonalization computations across devices. Experiments across multiple model scales demonstrate that NorMuon consistently outperforms both Adam and Muon, achieving 21.74% better training efficiency than Adam and 11.31% improvement over Muon on 1.1 B pretraining setting, while maintaining a comparable memory footprint to Muon. Our findings suggest that orthogonalization and adaptive learning rates are complementary rather than competing approaches, opening new avenues for optimizer design in large-scale deep learning.</p> - https://arxiv.org/abs/2510.05491 - Tue, 07 Oct 2025 01:13:41 +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> 12</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 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> 17</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 + Search-R3: Unifying Reasoning and Embedding Generation in Large Language Models + https://arxiv.org/abs/2510.07048 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07048.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yuntao Gui, James Cheng</p><p><b>Upvotes:</b> 0</p><p><b>Summary:</b> Despite their remarkable natural language understanding capabilities, Large Language Models (LLMs) have been underutilized for retrieval tasks. We present Search-R3, a novel framework that addresses this limitation by adapting LLMs to generate search embeddings as a direct output of their reasoning process. Our approach exploits LLMs' chain-of-thought capabilities, allowing them to produce more effective embeddings by reasoning step-by-step through complex semantic analyses. We implement this through three complementary mechanisms. (1) a supervised learning stage enables the model's ability to produce quality embeddings, (2) a reinforcement learning (RL) methodology that optimizes embedding generation alongside reasoning, and (3) a specialized RL environment that efficiently handles evolving embedding representations without requiring complete corpus re-encoding at each training iteration. Our extensive evaluations on diverse benchmarks demonstrate that Search-R3 significantly outperforms prior methods by unifying the reasoning and embedding generation processes. This integrated post-training approach represents a substantial advancement in handling complex knowledge-intensive tasks that require both sophisticated reasoning and effective information retrieval. Project page: https://github.com/ytgui/Search-R3</p> + https://arxiv.org/abs/2510.07048 + Wed, 08 Oct 2025 14:16:20 +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> 13</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 + 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> 25</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 - D^3QE: Learning Discrete Distribution Discrepancy-aware Quantization Error for Autoregressive-Generated Image Detection - https://arxiv.org/abs/2510.05891 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.05891.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yanran Zhang, Bingyao Yu, Yu Zheng, Wenzhao Zheng, Yueqi Duan, Lei Chen, Jie Zhou, Jiwen Lu</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> The emergence of visual autoregressive (AR) models has revolutionized image generation while presenting new challenges for synthetic image detection. Unlike previous GAN or diffusion-based methods, AR models generate images through discrete token prediction, exhibiting both marked improvements in image synthesis quality and unique characteristics in their vector-quantized representations. In this paper, we propose to leverage Discrete Distribution Discrepancy-aware Quantization Error (D^3QE) for autoregressive-generated image detection that exploits the distinctive patterns and the frequency distribution bias of the codebook existing in real and fake images. We introduce a discrete distribution discrepancy-aware transformer that integrates dynamic codebook frequency statistics into its attention mechanism, fusing semantic features and quantization error latent. To evaluate our method, we construct a comprehensive dataset termed ARForensics covering 7 mainstream visual AR models. Experiments demonstrate superior detection accuracy and strong generalization of D^3QE across different AR models, with robustness to real-world perturbations. Code is available at https://github.com/Zhangyr2022/D3QE{https://github.com/Zhangyr2022/D3QE}.</p> - https://arxiv.org/abs/2510.05891 - Tue, 07 Oct 2025 13:02:27 +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> 26</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 - 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> 39</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 + GyroSwin: 5D Surrogates for Gyrokinetic Plasma Turbulence Simulations + https://arxiv.org/abs/2510.07314 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07314.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Fabian Paischer, Gianluca Galletti, William Hornsby, Paul Setinek, Lorenzo Zanisi, Naomi Carey, Stanislas Pamela, Johannes Brandstetter</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> Nuclear fusion plays a pivotal role in the quest for reliable and sustainable energy production. A major roadblock to viable fusion power is understanding plasma turbulence, which significantly impairs plasma confinement, and is vital for next-generation reactor design. Plasma turbulence is governed by the nonlinear gyrokinetic equation, which evolves a 5D distribution function over time. Due to its high computational cost, reduced-order models are often employed in practice to approximate turbulent transport of energy. However, they omit nonlinear effects unique to the full 5D dynamics. To tackle this, we introduce GyroSwin, the first scalable 5D neural surrogate that can model 5D nonlinear gyrokinetic simulations, thereby capturing the physical phenomena neglected by reduced models, while providing accurate estimates of turbulent heat transport.GyroSwin (i) extends hierarchical Vision Transformers to 5D, (ii) introduces cross-attention and integration modules for latent 3Dleftrightarrow5D interactions between electrostatic potential fields and the distribution function, and (iii) performs channelwise mode separation inspired by nonlinear physics. We demonstrate that GyroSwin outperforms widely used reduced numerics on heat flux prediction, captures the turbulent energy cascade, and reduces the cost of fully resolved nonlinear gyrokinetics by three orders of magnitude while remaining physically verifiable. GyroSwin shows promising scaling laws, tested up to one billion parameters, paving the way for scalable neural surrogates for gyrokinetic simulations of plasma turbulence.</p> + https://arxiv.org/abs/2510.07314 + Wed, 08 Oct 2025 17:59:10 +0000 - FinLFQA: Evaluating Attributed Text Generation of LLMs in Financial Long-Form Question Answering - https://arxiv.org/abs/2510.06426 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06426.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yitao Long, Tiansheng Hu, Yilun Zhao, Arman Cohan, Chen Zhao</p><p><b>Upvotes:</b> 0</p><p><b>Summary:</b> Large Language Models (LLMs) frequently hallucinate to long-form questions, producing plausible yet factually incorrect answers. A common mitigation strategy is to provide attribution to LLM outputs. However, existing benchmarks primarily focus on simple attribution that retrieves supporting textual evidence as references. We argue that in real-world scenarios such as financial applications, attribution goes beyond reference retrieval. We introduce FinLFQA, a benchmark designed to evaluate the ability of LLMs to generate long-form answers to complex financial questions with reliable and nuanced attributions. FinLFQA evaluates three critical aspects of attribution through human annotations: (1) supporting evidence extracted from financial reports, (2) intermediate numerical reasoning steps, and (3) domain-specific financial knowledge that informs the reasoning process. We further provide an automatic evaluation framework covering both answer quality and attribution quality. Through extensive experiments on eight LLMs across multiple attribution-generation paradigms, we find that fine-grained metrics are important to distinguish model capabilities, that end-to-end generation achieves comparable performance to post-hoc approaches, and that iterative refinement only helps when guided by external feedback.</p> - https://arxiv.org/abs/2510.06426 - Tue, 07 Oct 2025 20:06:15 +0000 + Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs + https://arxiv.org/abs/2510.07429 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07429.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Wang Wei, Tiankai Yang, Hongjie Chen, Yue Zhao, Franck Dernoncourt, Ryan A. Rossi, Hoda Eldardiry</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Efficient use of large language models (LLMs) is critical for deployment at scale: without adaptive routing, systems either overpay for strong models or risk poor performance from weaker ones. Selecting the right LLM for each query is fundamentally an online decision problem: models differ in strengths, prices fluctuate, and users value accuracy and cost differently. Yet most routers are trained offline with labels for all candidate models, an assumption that breaks in deployment, where only the outcome of the chosen model is observed. We bridge this gap with BaRP, a Bandit-feedback Routing with Preferences approach that trains under the same partial-feedback restriction as deployment, while supporting preference-tunable inference: operators can dial the performance/cost trade-off at test time without retraining. Framed as a contextual bandit over prompt features and a user preference vector, our method simulates an online feedback setting during training and adapts its routing decisions to each new prompt, rather than depending on full-information offline supervision. Comprehensive experiments show that our method consistently outperforms strong offline routers by at least 12.46% and the largest LLM by at least 2.45%, and generalizes robustly for unseen tasks.</p> + https://arxiv.org/abs/2510.07429 + Wed, 08 Oct 2025 18:24:59 +0000 - PuzzlePlex: Benchmarking Foundation Models on Reasoning and Planning with Puzzles - https://arxiv.org/abs/2510.06475 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06475.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yitao Long, Yuru Jiang, Hongjun Liu, Yilun Zhao, Jingchen Sun, Yiqiu Shen, Chen Zhao, Arman Cohan, Dennis Shasha</p><p><b>Upvotes:</b> 0</p><p><b>Summary:</b> This work investigates the reasoning and planning capabilities of foundation models and their scalability in complex, dynamic environments. We introduce PuzzlePlex, a benchmark designed to assess these capabilities through a diverse set of puzzles. PuzzlePlex consists of 15 types of puzzles, including deterministic and stochastic games of varying difficulty, as well as single-player and two-player scenarios. The PuzzlePlex framework provides a comprehensive environment for each game, and supports extensibility to generate more challenging instances as foundation models evolve. Additionally, we implement customized game-playing strategies for comparison. Building on this benchmark, we develop fine-grained metrics to measure performance and conduct an in-depth analysis of frontier foundation models across two settings: instruction-based and code-based. Furthermore, we systematically investigate their scaling limits. Our findings show that reasoning models outperform others in instruction-based settings, while code-based execution presents greater challenges but offers a scalable and efficient alternative. PuzzlePlex enables targeted evaluation and guides future improvements in reasoning, planning, and generalization for foundation models.</p> - https://arxiv.org/abs/2510.06475 - Tue, 07 Oct 2025 21:24:29 +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> 37</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 - 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> 14</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 + 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> 15</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 - 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> 58</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 + OpenRubrics: Towards Scalable Synthetic Rubric Generation for Reward Modeling and LLM Alignment + https://arxiv.org/abs/2510.07743 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07743.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Tianci Liu, Ran Xu, Tony Yu, Ilgee Hong, Carl Yang, Tuo Zhao, Haoyu Wang</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Reward modeling lies at the core of reinforcement learning from human feedback (RLHF), yet most existing reward models rely on scalar or pairwise judgments that fail to capture the multifaceted nature of human preferences. Recent studies have explored rubrics-as-rewards (RaR) that uses structured natural language criteria that capture multiple dimensions of response quality. However, producing rubrics that are both reliable and scalable remains a key challenge. In this work, we introduce OpenRubrics, a diverse, large-scale collection of (prompt, rubric) pairs for training rubric-generation and rubric-based reward models. To elicit discriminative and comprehensive evaluation signals, we introduce Contrastive Rubric Generation (CRG), which derives both hard rules (explicit constraints) and principles (implicit qualities) by contrasting preferred and rejected responses. We further improve reliability by enforcing preference-label consistency via rejection sampling to remove noisy rubrics. Across multiple reward-modeling benchmarks, our rubric-based reward model, Rubric-RM, surpasses strong size-matched baselines by 6.8%. These gains transfer to policy models on instruction-following and biomedical benchmarks. Our results show that rubrics provide scalable alignment signals that narrow the gap between costly human evaluation and automated reward modeling, enabling a new principle-driven paradigm for LLM alignment.</p> + https://arxiv.org/abs/2510.07743 + Thu, 09 Oct 2025 03:31:26 +0000 - Code Agent can be an End-to-end System Hacker: Benchmarking Real-world Threats of Computer-use Agent - https://arxiv.org/abs/2510.06607 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06607.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Weidi Luo, Qiming Zhang, Tianyu Lu, Xiaogeng Liu, Bin Hu, Hung-Chun Chiu, Siyuan Ma, Yizhe Zhang, Xusheng Xiao, Yinzhi Cao, Zhen Xiang, Chaowei Xiao</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Computer-use agent (CUA) frameworks, powered by large language models (LLMs) or multimodal LLMs (MLLMs), are rapidly maturing as assistants that can perceive context, reason, and act directly within software environments. Among their most critical applications is operating system (OS) control. As CUAs in the OS domain become increasingly embedded in daily operations, it is imperative to examine their real-world security implications, specifically whether CUAs can be misused to perform realistic, security-relevant attacks. Existing works exhibit four major limitations: Missing attacker-knowledge model on tactics, techniques, and procedures (TTP), Incomplete coverage for end-to-end kill chains, unrealistic environment without multi-host and encrypted user credentials, and unreliable judgment dependent on LLM-as-a-Judge. To address these gaps, we propose AdvCUA, the first benchmark aligned with real-world TTPs in MITRE ATT&CK Enterprise Matrix, which comprises 140 tasks, including 40 direct malicious tasks, 74 TTP-based malicious tasks, and 26 end-to-end kill chains, systematically evaluates CUAs under a realistic enterprise OS security threat in a multi-host environment sandbox by hard-coded evaluation. We evaluate the existing five mainstream CUAs, including ReAct, AutoGPT, Gemini CLI, Cursor CLI, and Cursor IDE based on 8 foundation LLMs. The results demonstrate that current frontier CUAs do not adequately cover OS security-centric threats. These capabilities of CUAs reduce dependence on custom malware and deep domain expertise, enabling even inexperienced attackers to mount complex enterprise intrusions, which raises social concern about the responsibility and security of CUAs.</p> - https://arxiv.org/abs/2510.06607 - Wed, 08 Oct 2025 03:35:23 +0000 + GCPO: When Contrast Fails, Go Gold + https://arxiv.org/abs/2510.07790 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07790.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hao Wu, Wei Liu</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Reinforcement learning has been widely applied to enhance the reasoning capabilities of large language models. Extending the inference limits of smaller models has become a prominent research focus. However, algorithms such as Group Relative Policy Optimization (GRPO) suffer from a clear drawback: the upper bound of a model's rollout responses is entirely determined by the model itself, preventing the acquisition of knowledge from samples that are either all incorrect or all correct. In this paper, we introduce Group Contrastive Policy Optimization (GCPO), a method that incorporates external standard reference answers. When the model cannot solve a problem, the reference answer supplies the correct response, steering the model toward an unequivocally accurate update direction. This approach offers two main advantages: (1) it improves training efficiency by fully utilizing every sample; (2) it enables the model to emulate the problem solving strategy of the reference answer during training, thereby enhancing generalization in reasoning. GCPO achieves outstanding results across multiple benchmark datasets, yielding substantial improvements over the baseline model. Our code is available at: https://github.com/AchoWu/GCPO.</p> + https://arxiv.org/abs/2510.07790 + Thu, 09 Oct 2025 05:09:06 +0000 - Heptapod: Language Modeling on Visual Signals - https://arxiv.org/abs/2510.06673 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06673.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Yongxin Zhu, Jiawei Chen, Yuanzhe Chen, Zhuo Chen, Dongya Jia, Jian Cong, Xiaobin Zhuang, Yuping Wang, Yuxuan Wang</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> We introduce Heptapod, an image autoregressive model that adheres to the foundational principles of language modeling. Heptapod employs causal attention, eliminates reliance on CFG, and eschews the trend of semantic tokenizers. Our key innovation is next 2D distribution prediction: a causal Transformer with reconstruction-focused visual tokenizer, learns to predict the distribution over the entire 2D spatial grid of images at each timestep. This learning objective unifies the sequential modeling of autoregressive framework with the holistic self-supervised learning of masked autoencoding, enabling the model to capture comprehensive image semantics via generative training. On the ImageNet generation benchmark, Heptapod achieves an FID of 2.70, significantly outperforming previous causal autoregressive approaches. We hope our work inspires a principled rethinking of language modeling on visual signals and beyond.</p> - https://arxiv.org/abs/2510.06673 - Wed, 08 Oct 2025 05:54:46 +0000 + A^2Search: Ambiguity-Aware Question Answering with Reinforcement Learning + https://arxiv.org/abs/2510.07958 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07958.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Fengji Zhang, Xinyao Niu, Chengyang Ying, Guancheng Lin, Zhongkai Hao, Zhou Fan, Chengen Huang, Jacky Keung, Bei Chen, Junyang Lin</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Recent advances in Large Language Models (LLMs) and Reinforcement Learning (RL) have led to strong performance in open-domain question answering (QA). However, existing models still struggle with questions that admit multiple valid answers. Standard QA benchmarks, which typically assume a single gold answer, overlook this reality and thus produce inappropriate training signals. Existing attempts to handle ambiguity often rely on costly manual annotation, which is difficult to scale to multi-hop datasets such as HotpotQA and MuSiQue. In this paper, we present A^2Search, an annotation-free, end-to-end training framework to recognize and handle ambiguity. At its core is an automated pipeline that detects ambiguous questions and gathers alternative answers via trajectory sampling and evidence verification. The model is then optimized with RL using a carefully designed AnsF1 reward, which naturally accommodates multiple answers. Experiments on eight open-domain QA benchmarks demonstrate that A^2Search achieves new state-of-the-art performance. With only a single rollout, A^2Search-7B yields an average AnsF1@1 score of 48.4% across four multi-hop benchmarks, outperforming all strong baselines, including the substantially larger ReSearch-32B (46.2%). Extensive analyses further show that A^2Search resolves ambiguity and generalizes across benchmarks, highlighting that embracing ambiguity is essential for building more reliable QA systems. Our code, data, and model weights can be found at https://github.com/zfj1998/A2Search</p> + https://arxiv.org/abs/2510.07958 + Thu, 09 Oct 2025 08:53:31 +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> 29</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 + 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> 9</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 - 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> 14</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 + Recycling Pretrained Checkpoints: Orthogonal Growth of Mixture-of-Experts for Efficient Large Language Model Pre-Training + https://arxiv.org/abs/2510.08008 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08008.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Ruizhe Wang, Yucheng Ding, Xiao Liu, Yaoxiang Wang, Peng Cheng, Baining Guo, Zhengjun Zha, Yeyun Gong</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> The rapidly increasing computational cost of pretraining Large Language Models necessitates more efficient approaches. Numerous computational costs have been invested in existing well-trained checkpoints, but many of them remain underutilized due to engineering constraints or limited model capacity. To efficiently reuse this "sunk" cost, we propose to recycle pretrained checkpoints by expanding their parameter counts and continuing training. We propose orthogonal growth method well-suited for converged Mixture-of-Experts model: interpositional layer copying for depth growth and expert duplication with injected noise for width growth. To determine the optimal timing for such growth across checkpoints sequences, we perform comprehensive scaling experiments revealing that the final accuracy has a strong positive correlation with the amount of sunk cost, indicating that greater prior investment leads to better performance. We scale our approach to models with 70B parameters and over 1T training tokens, achieving 10.66% accuracy gain over training from scratch under the same additional compute budget. Our checkpoint recycling approach establishes a foundation for economically efficient large language model pretraining.</p> + https://arxiv.org/abs/2510.08008 + Thu, 09 Oct 2025 09:45:45 +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> 8</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 + 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> 18</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 - Online Generic Event Boundary Detection - https://arxiv.org/abs/2510.06855 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06855.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Hyungrok Jung, Daneul Kim, Seunggyun Lim, Jeany Son, Jonghyun Choi</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Generic Event Boundary Detection (GEBD) aims to interpret long-form videos through the lens of human perception. However, current GEBD methods require processing complete video frames to make predictions, unlike humans processing data online and in real-time. To bridge this gap, we introduce a new task, Online Generic Event Boundary Detection (On-GEBD), aiming to detect boundaries of generic events immediately in streaming videos. This task faces unique challenges of identifying subtle, taxonomy-free event changes in real-time, without the access to future frames. To tackle these challenges, we propose a novel On-GEBD framework, Estimator, inspired by Event Segmentation Theory (EST) which explains how humans segment ongoing activity into events by leveraging the discrepancies between predicted and actual information. Our framework consists of two key components: the Consistent Event Anticipator (CEA), and the Online Boundary Discriminator (OBD). Specifically, the CEA generates a prediction of the future frame reflecting current event dynamics based solely on prior frames. Then, the OBD measures the prediction error and adaptively adjusts the threshold using statistical tests on past errors to capture diverse, subtle event transitions. Experimental results demonstrate that Estimator outperforms all baselines adapted from recent online video understanding models and achieves performance comparable to prior offline-GEBD methods on the Kinetics-GEBD and TAPOS datasets.</p> - https://arxiv.org/abs/2510.06855 - Wed, 08 Oct 2025 10:23:45 +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> 21</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 - M3Retrieve: Benchmarking Multimodal Retrieval for Medicine - https://arxiv.org/abs/2510.06888 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06888.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Arkadeep Acharya, Akash Ghosh, Pradeepika Verma, Kitsuchart Pasupa, Sriparna Saha, Priti Singh</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> With the increasing use of RetrievalAugmented Generation (RAG), strong retrieval models have become more important than ever. In healthcare, multimodal retrieval models that combine information from both text and images offer major advantages for many downstream tasks such as question answering, cross-modal retrieval, and multimodal summarization, since medical data often includes both formats. However, there is currently no standard benchmark to evaluate how well these models perform in medical settings. To address this gap, we introduce M3Retrieve, a Multimodal Medical Retrieval Benchmark. M3Retrieve, spans 5 domains,16 medical fields, and 4 distinct tasks, with over 1.2 Million text documents and 164K multimodal queries, all collected under approved licenses. We evaluate leading multimodal retrieval models on this benchmark to explore the challenges specific to different medical specialities and to understand their impact on retrieval performance. By releasing M3Retrieve, we aim to enable systematic evaluation, foster model innovation, and accelerate research toward building more capable and reliable multimodal retrieval systems for medical applications. The dataset and the baselines code are available in this github page https://github.com/AkashGhosh/M3Retrieve.</p> - https://arxiv.org/abs/2510.06888 - Wed, 08 Oct 2025 11:08:47 +0000 + Memory Retrieval and Consolidation in Large Language Models through Function Tokens + https://arxiv.org/abs/2510.08203 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08203.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Shaohua Zhang, Yuan Lin, Hang Li</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> The remarkable success of large language models (LLMs) stems from their ability to consolidate vast amounts of knowledge into the memory during pre-training and to retrieve it from the memory during inference, enabling advanced capabilities such as knowledge memorization, instruction-following and reasoning. However, the mechanisms of memory retrieval and consolidation in LLMs remain poorly understood. In this paper, we propose the function token hypothesis to explain the workings of LLMs: During inference, function tokens activate the most predictive features from context and govern next token prediction (memory retrieval). During pre-training, predicting the next tokens (usually content tokens) that follow function tokens increases the number of learned features of LLMs and updates the model parameters (memory consolidation). Function tokens here roughly correspond to function words in linguistics, including punctuation marks, articles, prepositions, and conjunctions, in contrast to content tokens. We provide extensive experimental evidence supporting this hypothesis. Using bipartite graph analysis, we show that a small number of function tokens activate the majority of features. Case studies further reveal how function tokens activate the most predictive features from context to direct next token prediction. We also find that during pre-training, the training loss is dominated by predicting the next content tokens following function tokens, which forces the function tokens to select the most predictive features from context.</p> + https://arxiv.org/abs/2510.08203 + Thu, 09 Oct 2025 13:31:20 +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> 31</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 + 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> 18</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 - Revisiting the Uniform Information Density Hypothesis in LLM Reasoning Traces - https://arxiv.org/abs/2510.06953 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.06953.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Minju Gwak, Guijin Son, Jaehyung Kim</p><p><b>Upvotes:</b> 4</p><p><b>Summary:</b> The Uniform Information Density (UID) hypothesis suggests that effective communication maintains a stable flow of information. In this work, we revisit this principle in the context of large language model (LLM) reasoning traces, asking whether step-level uniformity reflects reasoning quality. To this end, we propose an entropy-based stepwise information density metric and introduce two complementary measures of uniformity, local and global uniformity scores. Across the experiments on six different reasoning benchmarks, we find that step-level uniformity not only provides a strong theoretical lens but also yields practical performance benefits; for example, selecting reasoning traces with more uniform information density at the step-level improves accuracy by 10-32\% relative gains over baselines at AIME2025. Our analysis further reveals that correct reasoning traces tend to avoid sharp information density spikes, while incorrect traces exhibit irregular information bursts. These results demonstrate that UID-inspired information density measures outperform alternative internal signals as predictors of reasoning quality. Results highlight the uniformity of the information density as a robust diagnostic and selection criterion for building more reliable and accurate reasoning systems.</p> - https://arxiv.org/abs/2510.06953 - Wed, 08 Oct 2025 12:37:04 +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 - 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> 15</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 + SViM3D: Stable Video Material Diffusion for Single Image 3D Generation + https://arxiv.org/abs/2510.08271 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08271.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Andreas Engelhardt, Mark Boss, Vikram Voletti, Chun-Han Yao, Hendrik P. A. Lensch, Varun Jampani</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> We present Stable Video Materials 3D (SViM3D), a framework to predict multi-view consistent physically based rendering (PBR) materials, given a single image. Recently, video diffusion models have been successfully used to reconstruct 3D objects from a single image efficiently. However, reflectance is still represented by simple material models or needs to be estimated in additional steps to enable relighting and controlled appearance edits. We extend a latent video diffusion model to output spatially varying PBR parameters and surface normals jointly with each generated view based on explicit camera control. This unique setup allows for relighting and generating a 3D asset using our model as neural prior. We introduce various mechanisms to this pipeline that improve quality in this ill-posed setting. We show state-of-the-art relighting and novel view synthesis performance on multiple object-centric datasets. Our method generalizes to diverse inputs, enabling the generation of relightable 3D assets useful in AR/VR, movies, games and other visual media.</p> + https://arxiv.org/abs/2510.08271 + Thu, 09 Oct 2025 14:29:47 +0000 - Beyond Monolingual Assumptions: A Survey of Code-Switched NLP in the Era of Large Language Models - https://arxiv.org/abs/2510.07037 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07037.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Rajvee Sheth, Samridhi Raj Sinha, Mahavir Patil, Himanshu Beniwal, Mayank Singh</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Code-switching (CSW), the alternation of languages and scripts within a single utterance, remains a fundamental challenge for multiling ual NLP, even amidst the rapid advances of large language models (LLMs). Most LLMs still struggle with mixed-language inputs, limited CSW datasets, and evaluation biases, hindering deployment in multilingual societies. This survey provides the first comprehensive analysis of CSW-aware LLM research, reviewing unique_references studies spanning five research areas, 12 NLP tasks, 30+ datasets, and 80+ languages. We classify recent advances by architecture, training strategy, and evaluation methodology, outlining how LLMs have reshaped CSW modeling and what challenges persist. The paper concludes with a roadmap emphasizing the need for inclusive datasets, fair evaluation, and linguistically grounded models to achieve truly multilingual intelligence. A curated collection of all resources is maintained at https://github.com/lingo-iitgn/awesome-code-mixing/.</p> - https://arxiv.org/abs/2510.07037 - Wed, 08 Oct 2025 14:04:14 +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 - U-Bench: A Comprehensive Understanding of U-Net through 100-Variant Benchmarking - https://arxiv.org/abs/2510.07041 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07041.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Fenghe Tang, Chengqi Dong, Wenxin Ma, Zikang Xu, Heqin Zhu, Zihang Jiang, Rongsheng Wang, Yuhao Wang, Chenxu Wu, Shaohua Kevin Zhou</p><p><b>Upvotes:</b> 2</p><p><b>Summary:</b> Over the past decade, U-Net has been the dominant architecture in medical image segmentation, leading to the development of thousands of U-shaped variants. Despite its widespread adoption, there is still no comprehensive benchmark to systematically evaluate their performance and utility, largely because of insufficient statistical validation and limited consideration of efficiency and generalization across diverse datasets. To bridge this gap, we present U-Bench, the first large-scale, statistically rigorous benchmark that evaluates 100 U-Net variants across 28 datasets and 10 imaging modalities. Our contributions are threefold: (1) Comprehensive Evaluation: U-Bench evaluates models along three key dimensions: statistical robustness, zero-shot generalization, and computational efficiency. We introduce a novel metric, U-Score, which jointly captures the performance-efficiency trade-off, offering a deployment-oriented perspective on model progress. (2) Systematic Analysis and Model Selection Guidance: We summarize key findings from the large-scale evaluation and systematically analyze the impact of dataset characteristics and architectural paradigms on model performance. Based on these insights, we propose a model advisor agent to guide researchers in selecting the most suitable models for specific datasets and tasks. (3) Public Availability: We provide all code, models, protocols, and weights, enabling the community to reproduce our results and extend the benchmark with future methods. In summary, U-Bench not only exposes gaps in previous evaluations but also establishes a foundation for fair, reproducible, and practically relevant benchmarking in the next decade of U-Net-based segmentation models. The project can be accessed at: https://fenghetan9.github.io/ubench. Code is available at: https://github.com/FengheTan9/U-Bench.</p> - https://arxiv.org/abs/2510.07041 - Wed, 08 Oct 2025 14:06:17 +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> 18</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 - 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> 10</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 + 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> 47</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 - 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> 12</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 + 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> 8</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 - MLE-Smith: Scaling MLE Tasks with Automated Multi-Agent Pipeline - https://arxiv.org/abs/2510.07307 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07307.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Rushi Qiang, Yuchen Zhuang, Anikait Singh, Percy Liang, Chao Zhang, Sherry Yang, Bo Dai</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> While Language Models (LMs) have made significant progress in automating machine learning engineering (MLE), the acquisition of high-quality MLE training data is significantly constrained. Current MLE benchmarks suffer from low scalability and limited applicability because they rely on static, manually curated tasks, demanding extensive time and manual effort to produce. We introduce MLE-Smith, a fully automated multi-agent pipeline, to transform raw datasets into competition-style MLE challenges through an efficient generate-verify-execute paradigm for scaling MLE tasks with verifiable quality, real-world usability, and rich diversity. The proposed multi-agent pipeline in MLE-Smith drives structured task design and standardized refactoring, coupled with a hybrid verification mechanism that enforces strict structural rules and high-level semantic soundness. It further validates empirical solvability and real-world fidelity through interactive execution. We apply MLE-Smith to 224 of real-world datasets and generate 606 tasks spanning multiple categories, objectives, and modalities, demonstrating that MLE-Smith can work effectively across a wide range of real-world datasets. Evaluation on the generated tasks shows that the performance of eight mainstream and cutting-edge LLMs on MLE-Smith tasks is strongly correlated with their performance on carefully human-designed tasks, highlighting the effectiveness of the MLE-Smith to scaling up MLE tasks, while maintaining task quality.</p> - https://arxiv.org/abs/2510.07307 - Wed, 08 Oct 2025 17:57:19 +0000 + Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency + https://arxiv.org/abs/2510.08431 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08431.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Kaiwen Zheng, Yuji Wang, Qianli Ma, Huayu Chen, Jintao Zhang, Yogesh Balaji, Jianfei Chen, Ming-Yu Liu, Jun Zhu, Qinsheng Zhang</p><p><b>Upvotes:</b> 7</p><p><b>Summary:</b> This work represents the first effort to scale up continuous-time consistency distillation to general application-level image and video diffusion models. Although continuous-time consistency model (sCM) is theoretically principled and empirically powerful for accelerating academic-scale diffusion, its applicability to large-scale text-to-image and video tasks remains unclear due to infrastructure challenges in Jacobian-vector product (JVP) computation and the limitations of standard evaluation benchmarks. We first develop a parallelism-compatible FlashAttention-2 JVP kernel, enabling sCM training on models with over 10 billion parameters and high-dimensional video tasks. Our investigation reveals fundamental quality limitations of sCM in fine-detail generation, which we attribute to error accumulation and the "mode-covering" nature of its forward-divergence objective. To remedy this, we propose the score-regularized continuous-time consistency model (rCM), which incorporates score distillation as a long-skip regularizer. This integration complements sCM with the "mode-seeking" reverse divergence, effectively improving visual quality while maintaining high generation diversity. Validated on large-scale models (Cosmos-Predict2, Wan2.1) up to 14B parameters and 5-second videos, rCM matches or surpasses the state-of-the-art distillation method DMD2 on quality metrics while offering notable advantages in diversity, all without GAN tuning or extensive hyperparameter searches. The distilled models generate high-fidelity samples in only 1sim4 steps, accelerating diffusion sampling by 15timessim50times. These results position rCM as a practical and theoretically grounded framework for advancing large-scale diffusion distillation.</p> + https://arxiv.org/abs/2510.08431 + Thu, 09 Oct 2025 16:45:30 +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> 28</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 + 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> 21</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 - WristWorld: Generating Wrist-Views via 4D World Models for Robotic Manipulation - https://arxiv.org/abs/2510.07313 - <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.07313.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zezhong Qian, Xiaowei Chi, Yuming Li, Shizun Wang, Zhiyuan Qin, Xiaozhu Ju, Sirui Han, Shanghang Zhang</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> Wrist-view observations are crucial for VLA models as they capture fine-grained hand-object interactions that directly enhance manipulation performance. Yet large-scale datasets rarely include such recordings, resulting in a substantial gap between abundant anchor views and scarce wrist views. Existing world models cannot bridge this gap, as they require a wrist-view first frame and thus fail to generate wrist-view videos from anchor views alone. Amid this gap, recent visual geometry models such as VGGT emerge with geometric and cross-view priors that make it possible to address extreme viewpoint shifts. Inspired by these insights, we propose WristWorld, the first 4D world model that generates wrist-view videos solely from anchor views. WristWorld operates in two stages: (i) Reconstruction, which extends VGGT and incorporates our Spatial Projection Consistency (SPC) Loss to estimate geometrically consistent wrist-view poses and 4D point clouds; (ii) Generation, which employs our video generation model to synthesize temporally coherent wrist-view videos from the reconstructed perspective. Experiments on Droid, Calvin, and Franka Panda demonstrate state-of-the-art video generation with superior spatial consistency, while also improving VLA performance, raising the average task completion length on Calvin by 3.81% and closing 42.4% of the anchor-wrist view gap.</p> - https://arxiv.org/abs/2510.07313 - Wed, 08 Oct 2025 17:59:08 +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> 13</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 - 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> 27</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 + 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> 13</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 - 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> 17</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 + 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> 89</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 + + + R2RGEN: Real-to-Real 3D Data Generation for Spatially Generalized Manipulation + https://arxiv.org/abs/2510.08547 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08547.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xiuwei Xu, Angyuan Ma, Hankun Li, Bingyao Yu, Zheng Zhu, Jie Zhou, Jiwen Lu</p><p><b>Upvotes:</b> 1</p><p><b>Summary:</b> Towards the aim of generalized robotic manipulation, spatial generalization is the most fundamental capability that requires the policy to work robustly under different spatial distribution of objects, environment and agent itself. To achieve this, substantial human demonstrations need to be collected to cover different spatial configurations for training a generalized visuomotor policy via imitation learning. Prior works explore a promising direction that leverages data generation to acquire abundant spatially diverse data from minimal source demonstrations. However, most approaches face significant sim-to-real gap and are often limited to constrained settings, such as fixed-base scenarios and predefined camera viewpoints. In this paper, we propose a real-to-real 3D data generation framework (R2RGen) that directly augments the pointcloud observation-action pairs to generate real-world data. R2RGen is simulator- and rendering-free, thus being efficient and plug-and-play. Specifically, given a single source demonstration, we introduce an annotation mechanism for fine-grained parsing of scene and trajectory. A group-wise augmentation strategy is proposed to handle complex multi-object compositions and diverse task constraints. We further present camera-aware processing to align the distribution of generated data with real-world 3D sensor. Empirically, R2RGen substantially enhances data efficiency on extensive experiments and demonstrates strong potential for scaling and application on mobile manipulation.</p> + https://arxiv.org/abs/2510.08547 + Thu, 09 Oct 2025 17:55:44 +0000 + + + Entropy Regularizing Activation: Boosting Continuous Control, Large Language Models, and Image Classification with Activation as Entropy Constraints + https://arxiv.org/abs/2510.08549 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08549.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Zilin Kang, Chonghua Liao, Tingqiang Xu, Huazhe Xu</p><p><b>Upvotes:</b> 5</p><p><b>Summary:</b> We propose ERA, a new paradigm that constrains the sampling entropy above given thresholds by applying specially designed activations to the outputs of models. Our approach demonstrates broad effectiveness across different domains: 1) for large language models(LLMs), boosting the AIME 2025 score for Qwen2.5-Math-7B by 37.4%; 2) for continuous control reinforcement learning agents, improving performance by more than 30% over strong baselines such as SAC on the challenging HumanoidBench; 3) for image classification, enhancing ImageNet top-1 accuracy by 0.69% for ResNet-50. These gains are achieved with a computational overhead of less than 7%. Our work validates output activation as a powerful tool for entropy control, opening a new direction for designing simpler and more robust algorithms.</p> + https://arxiv.org/abs/2510.08549 + Thu, 09 Oct 2025 17:56:17 +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> 21</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 + + + 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> 36</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 + + + DexNDM: Closing the Reality Gap for Dexterous In-Hand Rotation via Joint-Wise Neural Dynamics Model + https://arxiv.org/abs/2510.08556 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08556.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Xueyi Liu, He Wang, Li Yi</p><p><b>Upvotes:</b> 3</p><p><b>Summary:</b> Achieving generalized in-hand object rotation remains a significant challenge in robotics, largely due to the difficulty of transferring policies from simulation to the real world. The complex, contact-rich dynamics of dexterous manipulation create a "reality gap" that has limited prior work to constrained scenarios involving simple geometries, limited object sizes and aspect ratios, constrained wrist poses, or customized hands. We address this sim-to-real challenge with a novel framework that enables a single policy, trained in simulation, to generalize to a wide variety of objects and conditions in the real world. The core of our method is a joint-wise dynamics model that learns to bridge the reality gap by effectively fitting limited amount of real-world collected data and then adapting the sim policy's actions accordingly. The model is highly data-efficient and generalizable across different whole-hand interaction distributions by factorizing dynamics across joints, compressing system-wide influences into low-dimensional variables, and learning each joint's evolution from its own dynamic profile, implicitly capturing these net effects. We pair this with a fully autonomous data collection strategy that gathers diverse, real-world interaction data with minimal human intervention. Our complete pipeline demonstrates unprecedented generality: a single policy successfully rotates challenging objects with complex shapes (e.g., animals), high aspect ratios (up to 5.33), and small sizes, all while handling diverse wrist orientations and rotation axes. Comprehensive real-world evaluations and a teleoperation application for complex tasks validate the effectiveness and robustness of our approach. Website: https://meowuu7.github.io/DexNDM/</p> + https://arxiv.org/abs/2510.08556 + Thu, 09 Oct 2025 17:59:11 +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> 121</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 + + + SciVideoBench: Benchmarking Scientific Video Reasoning in Large Multimodal Models + https://arxiv.org/abs/2510.08559 + <p><img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2510.08559.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /></p><p><b>Authors:</b> Andong Deng, Taojiannan Yang, Shoubin Yu, Lincoln Spencer, Mohit Bansal, Chen Chen, Serena Yeung-Levy, Xiaohan Wang</p><p><b>Upvotes:</b> 7</p><p><b>Summary:</b> Large Multimodal Models (LMMs) have achieved remarkable progress across various capabilities; however, complex video reasoning in the scientific domain remains a significant and challenging frontier. Current video benchmarks predominantly target general scenarios where perception/recognition is heavily relied on, while with relatively simple reasoning tasks, leading to saturation and thus failing to effectively evaluate advanced multimodal cognitive skills. To address this critical gap, we introduce SciVideoBench, a rigorous benchmark specifically designed to assess advanced video reasoning in scientific contexts. SciVideoBench consists of 1,000 carefully crafted multiple-choice questions derived from cutting-edge scientific experimental videos spanning over 25 specialized academic subjects and verified by a semi-automatic system. Each question demands sophisticated domain-specific knowledge, precise spatiotemporal perception, and intricate logical reasoning, effectively challenging models' higher-order cognitive abilities. Our evaluation highlights significant performance deficits in state-of-the-art proprietary and open-source LMMs, including Gemini 2.5 Pro and Qwen2.5-VL, indicating substantial room for advancement in video reasoning capabilities. Detailed analyses of critical factors such as reasoning complexity and visual grounding provide valuable insights and clear direction for future developments in LMMs, driving the evolution of truly capable multimodal AI co-scientists. We hope SciVideoBench could fit the interests of the community and help to push the boundary of cutting-edge AI for border science.</p> + https://arxiv.org/abs/2510.08559 + Thu, 09 Oct 2025 17:59:23 +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