bot: update RSS feed

2026-02-08 00:06:53 +00:00
parent 43ccf9a613
commit 4be83be3b1
1 changed files with 27 additions and 27 deletions
--- a/feed.xml
+++ b/feed.xml
@@ -7,7 +7,7 @@
    <docs>http://www.rssboard.org/rss-specification</docs>
    <generator>python-feedgen</generator>
    <language>en</language>
-    <lastBuildDate>Sat, 07 Feb 2026 00:04:34 +0000</lastBuildDate>
+    <lastBuildDate>Sun, 08 Feb 2026 00:06:53 +0000</lastBuildDate>
    <item>
      <title>Thinking in Frames: How Visual Context and Test-Time Scaling Empower Video Reasoning</title>
      <link>https://arxiv.org/abs/2601.21037</link>
@@ -25,14 +25,14 @@
    <item>
      <title>Retrieval-Infused Reasoning Sandbox: A Benchmark for Decoupling Retrieval and Reasoning Capabilities</title>
      <link>https://arxiv.org/abs/2601.21937</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2601.21937.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shuangshuang Ying, Zheyu Wang, Yunjian Peng, Jin Chen, Yuhao Wu, Hongbin Lin, Dingyu He, Siyi Liu, Gengchen Yu, YinZhu Piao, Yuchen Wu, Xin Gui, Zhongyuan Peng, Xin Li, Xeron Du, Libo Qin, YiXin Cao, Ge Zhang, Stephen Huang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 14&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Despite strong performance on existing benchmarks, it remains unclear whether large language models can reason over genuinely novel scientific information. Most evaluations score end-to-end RAG pipelines, where reasoning is confounded with retrieval and toolchain choices, and the signal is further contaminated by parametric memorization and open-web volatility. We introduce DeR2, a controlled deep-research sandbox that isolates document-grounded reasoning while preserving core difficulties of deep search: multi-step synthesis, denoising, and evidence-based conclusion making. DeR2 decouples evidence access from reasoning via four regimes--Instruction-only, Concepts (gold concepts without documents), Related-only (only relevant documents), and Full-set (relevant documents plus topically related distractors)--yielding interpretable regime gaps that operationalize retrieval loss vs. reasoning loss and enable fine-grained error attribution. To prevent parametric leakage, we apply a two-phase validation that requires parametric failure without evidence while ensuring oracle-concept solvability. To ensure reproducibility, each instance provides a frozen document library (drawn from 2023-2025 theoretical papers) with expert-annotated concepts and validated rationales. Experiments across a diverse set of state-of-the-art foundation models reveal substantial variation and significant headroom: some models exhibit mode-switch fragility, performing worse with the Full-set than with Instruction-only, while others show structural concept misuse, correctly naming concepts but failing to execute them as procedures.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2601.21937.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shuangshuang Ying, Zheyu Wang, Yunjian Peng, Jin Chen, Yuhao Wu, Hongbin Lin, Dingyu He, Siyi Liu, Gengchen Yu, YinZhu Piao, Yuchen Wu, Xin Gui, Zhongyuan Peng, Xin Li, Xeron Du, Libo Qin, YiXin Cao, Ge Zhang, Stephen Huang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 17&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Despite strong performance on existing benchmarks, it remains unclear whether large language models can reason over genuinely novel scientific information. Most evaluations score end-to-end RAG pipelines, where reasoning is confounded with retrieval and toolchain choices, and the signal is further contaminated by parametric memorization and open-web volatility. We introduce DeR2, a controlled deep-research sandbox that isolates document-grounded reasoning while preserving core difficulties of deep search: multi-step synthesis, denoising, and evidence-based conclusion making. DeR2 decouples evidence access from reasoning via four regimes--Instruction-only, Concepts (gold concepts without documents), Related-only (only relevant documents), and Full-set (relevant documents plus topically related distractors)--yielding interpretable regime gaps that operationalize retrieval loss vs. reasoning loss and enable fine-grained error attribution. To prevent parametric leakage, we apply a two-phase validation that requires parametric failure without evidence while ensuring oracle-concept solvability. To ensure reproducibility, each instance provides a frozen document library (drawn from 2023-2025 theoretical papers) with expert-annotated concepts and validated rationales. Experiments across a diverse set of state-of-the-art foundation models reveal substantial variation and significant headroom: some models exhibit mode-switch fragility, performing worse with the Full-set than with Instruction-only, while others show structural concept misuse, correctly naming concepts but failing to execute them as procedures.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2601.21937</guid>
      <pubDate>Thu, 29 Jan 2026 16:26:19 +0000</pubDate>
    </item>
    <item>
      <title>CAR-bench: Evaluating the Consistency and Limit-Awareness of LLM Agents under Real-World Uncertainty</title>
      <link>https://arxiv.org/abs/2601.22027</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2601.22027.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Johannes Kirmayr, Lukas Stappen, Elisabeth André&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 60&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Existing benchmarks for Large Language Model (LLM) agents focus on task completion under idealistic settings but overlook reliability in real-world, user-facing applications. In domains, such as in-car voice assistants, users often issue incomplete or ambiguous requests, creating intrinsic uncertainty that agents must manage through dialogue, tool use, and policy adherence. We introduce CAR-bench, a benchmark for evaluating consistency, uncertainty handling, and capability awareness in multi-turn, tool-using LLM agents in an in-car assistant domain. The environment features an LLM-simulated user, domain policies, and 58 interconnected tools spanning navigation, productivity, charging, and vehicle control. Beyond standard task completion, CAR-bench introduces Hallucination tasks that test agents' limit-awareness under missing tools or information, and Disambiguation tasks that require resolving uncertainty through clarification or internal information gathering. Baseline results reveal large gaps between occasional and consistent success on all task types. Even frontier reasoning LLMs achieve less than 50% consistent pass rate on Disambiguation tasks due to premature actions, and frequently violate policies or fabricate information to satisfy user requests in Hallucination tasks, underscoring the need for more reliable and self-aware LLM agents in real-world settings.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2601.22027.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Johannes Kirmayr, Lukas Stappen, Elisabeth André&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 69&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Existing benchmarks for Large Language Model (LLM) agents focus on task completion under idealistic settings but overlook reliability in real-world, user-facing applications. In domains, such as in-car voice assistants, users often issue incomplete or ambiguous requests, creating intrinsic uncertainty that agents must manage through dialogue, tool use, and policy adherence. We introduce CAR-bench, a benchmark for evaluating consistency, uncertainty handling, and capability awareness in multi-turn, tool-using LLM agents in an in-car assistant domain. The environment features an LLM-simulated user, domain policies, and 58 interconnected tools spanning navigation, productivity, charging, and vehicle control. Beyond standard task completion, CAR-bench introduces Hallucination tasks that test agents' limit-awareness under missing tools or information, and Disambiguation tasks that require resolving uncertainty through clarification or internal information gathering. Baseline results reveal large gaps between occasional and consistent success on all task types. Even frontier reasoning LLMs achieve less than 50% consistent pass rate on Disambiguation tasks due to premature actions, and frequently violate policies or fabricate information to satisfy user requests in Hallucination tasks, underscoring the need for more reliable and self-aware LLM agents in real-world settings.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2601.22027</guid>
      <pubDate>Thu, 29 Jan 2026 17:33:42 +0000</pubDate>
    </item>
@@ -67,7 +67,7 @@
    <item>
      <title>DASH: Faster Shampoo via Batched Block Preconditioning and Efficient Inverse-Root Solvers</title>
      <link>https://arxiv.org/abs/2602.02016</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.02016.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Ionut-Vlad Modoranu, Philip Zmushko, Erik Schultheis, Mher Safaryan, Dan Alistarh&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 2&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Shampoo is one of the leading approximate second-order optimizers: a variant of it has won the MLCommons AlgoPerf competition, and it has been shown to produce models with lower activation outliers that are easier to compress. Yet, applying Shampoo currently comes at the cost of significant computational slowdown, due to its expensive internal operations. In this paper, we take a significant step to address this shortcoming by proposing \method (for Distributed Accelerated SHampoo), a faster implementation of Distributed Shampoo based on two main new techniques: First, we show that preconditioner blocks can be stacked into 3D tensors to significantly improve GPU utilization; second, we introduce the Newton-DB iteration and the Chebyshev polynomial approximations as novel and faster approaches for computing the inverse matrix roots required by Shampoo. Along with these algorithmic contributions, we provide a first in-depth analysis of how matrix scaling critically affects Shampoo convergence. On the practical side, our GPU-aware implementation achieves up to 4.83times faster optimizer steps compared to the well-optimized Distributed Shampoo, while Newton-DB attains the lowest validation perplexity per iteration among all tested methods. Our code is available at https://github.com/IST-DASLab/DASH.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.02016.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Ionut-Vlad Modoranu, Philip Zmushko, Erik Schultheis, Mher Safaryan, Dan Alistarh&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 6&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Shampoo is one of the leading approximate second-order optimizers: a variant of it has won the MLCommons AlgoPerf competition, and it has been shown to produce models with lower activation outliers that are easier to compress. Yet, applying Shampoo currently comes at the cost of significant computational slowdown, due to its expensive internal operations. In this paper, we take a significant step to address this shortcoming by proposing \method (for Distributed Accelerated SHampoo), a faster implementation of Distributed Shampoo based on two main new techniques: First, we show that preconditioner blocks can be stacked into 3D tensors to significantly improve GPU utilization; second, we introduce the Newton-DB iteration and the Chebyshev polynomial approximations as novel and faster approaches for computing the inverse matrix roots required by Shampoo. Along with these algorithmic contributions, we provide a first in-depth analysis of how matrix scaling critically affects Shampoo convergence. On the practical side, our GPU-aware implementation achieves up to 4.83times faster optimizer steps compared to the well-optimized Distributed Shampoo, while Newton-DB attains the lowest validation perplexity per iteration among all tested methods. Our code is available at https://github.com/IST-DASLab/DASH.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.02016</guid>
      <pubDate>Mon, 02 Feb 2026 12:14:45 +0000</pubDate>
    </item>
@@ -88,21 +88,21 @@
    <item>
      <title>MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents</title>
      <link>https://arxiv.org/abs/2602.02474</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.02474.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Haozhen Zhang, Quanyu Long, Jianzhu Bao, Tao Feng, Weizhi Zhang, Haodong Yue, Wenya Wang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 23&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Most Large Language Model (LLM) agent memory systems rely on a small set of static, hand-designed operations for extracting memory. These fixed procedures hard-code human priors about what to store and how to revise memory, making them rigid under diverse interaction patterns and inefficient on long histories. To this end, we present MemSkill, which reframes these operations as learnable and evolvable memory skills, structured and reusable routines for extracting, consolidating, and pruning information from interaction traces. Inspired by the design philosophy of agent skills, MemSkill employs a controller that learns to select a small set of relevant skills, paired with an LLM-based executor that produces skill-guided memories. Beyond learning skill selection, MemSkill introduces a designer that periodically reviews hard cases where selected skills yield incorrect or incomplete memories, and evolves the skill set by proposing refinements and new skills. Together, MemSkill forms a closed-loop procedure that improves both the skill-selection policy and the skill set itself. Experiments on LoCoMo, LongMemEval, HotpotQA, and ALFWorld demonstrate that MemSkill improves task performance over strong baselines and generalizes well across settings. Further analyses shed light on how skills evolve, offering insights toward more adaptive, self-evolving memory management for LLM agents.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.02474.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Haozhen Zhang, Quanyu Long, Jianzhu Bao, Tao Feng, Weizhi Zhang, Haodong Yue, Wenya Wang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 42&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Most Large Language Model (LLM) agent memory systems rely on a small set of static, hand-designed operations for extracting memory. These fixed procedures hard-code human priors about what to store and how to revise memory, making them rigid under diverse interaction patterns and inefficient on long histories. To this end, we present MemSkill, which reframes these operations as learnable and evolvable memory skills, structured and reusable routines for extracting, consolidating, and pruning information from interaction traces. Inspired by the design philosophy of agent skills, MemSkill employs a controller that learns to select a small set of relevant skills, paired with an LLM-based executor that produces skill-guided memories. Beyond learning skill selection, MemSkill introduces a designer that periodically reviews hard cases where selected skills yield incorrect or incomplete memories, and evolves the skill set by proposing refinements and new skills. Together, MemSkill forms a closed-loop procedure that improves both the skill-selection policy and the skill set itself. Experiments on LoCoMo, LongMemEval, HotpotQA, and ALFWorld demonstrate that MemSkill improves task performance over strong baselines and generalizes well across settings. Further analyses shed light on how skills evolve, offering insights toward more adaptive, self-evolving memory management for LLM agents.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.02474</guid>
      <pubDate>Mon, 02 Feb 2026 18:53:28 +0000</pubDate>
    </item>
    <item>
      <title>LatentMem: Customizing Latent Memory for Multi-Agent Systems</title>
      <link>https://arxiv.org/abs/2602.03036</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.03036.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Muxin Fu, Guibin Zhang, Xiangyuan Xue, Yafu Li, Zefeng He, Siyuan Huang, Xiaoye Qu, Yu Cheng, Yang Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 8&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language model (LLM)-powered multi-agent systems (MAS) demonstrate remarkable collective intelligence, wherein multi-agent memory serves as a pivotal mechanism for continual adaptation. However, existing multi-agent memory designs remain constrained by two fundamental bottlenecks: (i) memory homogenization arising from the absence of role-aware customization, and (ii) information overload induced by excessively fine-grained memory entries. To address these limitations, we propose LatentMem, a learnable multi-agent memory framework designed to customize agent-specific memories in a token-efficient manner. Specifically, LatentMem comprises an experience bank that stores raw interaction trajectories in a lightweight form, and a memory composer that synthesizes compact latent memories conditioned on retrieved experience and agent-specific contexts. Further, we introduce Latent Memory Policy Optimization (LMPO), which propagates task-level optimization signals through latent memories to the composer, encouraging it to produce compact and high-utility representations. Extensive experiments across diverse benchmarks and mainstream MAS frameworks show that LatentMem achieves a performance gain of up to 19.36% over vanilla settings and consistently outperforms existing memory architectures, without requiring any modifications to the underlying frameworks.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.03036.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Muxin Fu, Guibin Zhang, Xiangyuan Xue, Yafu Li, Zefeng He, Siyuan Huang, Xiaoye Qu, Yu Cheng, Yang Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 12&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language model (LLM)-powered multi-agent systems (MAS) demonstrate remarkable collective intelligence, wherein multi-agent memory serves as a pivotal mechanism for continual adaptation. However, existing multi-agent memory designs remain constrained by two fundamental bottlenecks: (i) memory homogenization arising from the absence of role-aware customization, and (ii) information overload induced by excessively fine-grained memory entries. To address these limitations, we propose LatentMem, a learnable multi-agent memory framework designed to customize agent-specific memories in a token-efficient manner. Specifically, LatentMem comprises an experience bank that stores raw interaction trajectories in a lightweight form, and a memory composer that synthesizes compact latent memories conditioned on retrieved experience and agent-specific contexts. Further, we introduce Latent Memory Policy Optimization (LMPO), which propagates task-level optimization signals through latent memories to the composer, encouraging it to produce compact and high-utility representations. Extensive experiments across diverse benchmarks and mainstream MAS frameworks show that LatentMem achieves a performance gain of up to 19.36% over vanilla settings and consistently outperforms existing memory architectures, without requiring any modifications to the underlying frameworks.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.03036</guid>
      <pubDate>Tue, 03 Feb 2026 03:03:16 +0000</pubDate>
    </item>
    <item>
      <title>Accurate Failure Prediction in Agents Does Not Imply Effective Failure Prevention</title>
      <link>https://arxiv.org/abs/2602.03338</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.03338.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Rakshith Vasudev, Melisa Russak, Dan Bikel, Waseem Alshikh&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 23&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Proactive interventions by LLM critic models are often assumed to improve reliability, yet their effects at deployment time are poorly understood. We show that a binary LLM critic with strong offline accuracy (AUROC 0.94) can nevertheless cause severe performance degradation, inducing a 26 percentage point (pp) collapse on one model while affecting another by near zero pp. This variability demonstrates that LLM critic accuracy alone is insufficient to determine whether intervention is safe.   We identify a disruption-recovery tradeoff: interventions may recover failing trajectories but also disrupt trajectories that would have succeeded. Based on this insight, we propose a pre-deployment test that uses a small pilot of 50 tasks to estimate whether intervention is likely to help or harm, without requiring full deployment. Across benchmarks, the test correctly anticipates outcomes: intervention degrades performance on high-success tasks (0 to -26 pp), while yielding a modest improvement on the high-failure ALFWorld benchmark (+2.8 pp, p=0.014). The primary value of our framework is therefore identifying when not to intervene, preventing severe regressions before deployment.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.03338.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Rakshith Vasudev, Melisa Russak, Dan Bikel, Waseem Alshikh&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 25&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Proactive interventions by LLM critic models are often assumed to improve reliability, yet their effects at deployment time are poorly understood. We show that a binary LLM critic with strong offline accuracy (AUROC 0.94) can nevertheless cause severe performance degradation, inducing a 26 percentage point (pp) collapse on one model while affecting another by near zero pp. This variability demonstrates that LLM critic accuracy alone is insufficient to determine whether intervention is safe.   We identify a disruption-recovery tradeoff: interventions may recover failing trajectories but also disrupt trajectories that would have succeeded. Based on this insight, we propose a pre-deployment test that uses a small pilot of 50 tasks to estimate whether intervention is likely to help or harm, without requiring full deployment. Across benchmarks, the test correctly anticipates outcomes: intervention degrades performance on high-success tasks (0 to -26 pp), while yielding a modest improvement on the high-failure ALFWorld benchmark (+2.8 pp, p=0.014). The primary value of our framework is therefore identifying when not to intervene, preventing severe regressions before deployment.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.03338</guid>
      <pubDate>Tue, 03 Feb 2026 10:02:50 +0000</pubDate>
    </item>
@@ -137,14 +137,14 @@
    <item>
      <title>Privileged Information Distillation for Language Models</title>
      <link>https://arxiv.org/abs/2602.04942</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.04942.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Emiliano Penaloza, Dheeraj Vattikonda, Nicolas Gontier, Alexandre Lacoste, Laurent Charlin, Massimo Caccia&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 15&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Training-time privileged information (PI) can enable language models to succeed on tasks they would otherwise fail, making it a powerful tool for reinforcement learning in hard, long-horizon settings. However, transferring capabilities learned with PI to policies that must act without it at inference time remains a fundamental challenge. We study this problem in the context of distilling frontier models for multi-turn agentic environments, where closed-source systems typically hide their internal reasoning and expose only action trajectories. This breaks standard distillation pipelines, since successful behavior is observable but the reasoning process is not. For this, we introduce π-Distill, a joint teacher-student objective that trains a PI-conditioned teacher and an unconditioned student simultaneously using the same model. Additionally, we also introduce On-Policy Self-Distillation (OPSD), an alternative approach that trains using Reinforcement Learning (RL) with a reverse KL-penalty between the student and the PI-conditioned teacher. We show that both of these algorithms effectively distill frontier agents using action-only PI. Specifically we find that π-Distill and in some cases OPSD, outperform industry standard practices (Supervised finetuning followed by RL) that assume access to full Chain-of-Thought supervision across multiple agentic benchmarks, models, and forms of PI. We complement our results with extensive analysis that characterizes the factors enabling effective learning with PI, focusing primarily on π-Distill and characterizing when OPSD is competitive.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.04942.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Emiliano Penaloza, Dheeraj Vattikonda, Nicolas Gontier, Alexandre Lacoste, Laurent Charlin, Massimo Caccia&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 20&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Training-time privileged information (PI) can enable language models to succeed on tasks they would otherwise fail, making it a powerful tool for reinforcement learning in hard, long-horizon settings. However, transferring capabilities learned with PI to policies that must act without it at inference time remains a fundamental challenge. We study this problem in the context of distilling frontier models for multi-turn agentic environments, where closed-source systems typically hide their internal reasoning and expose only action trajectories. This breaks standard distillation pipelines, since successful behavior is observable but the reasoning process is not. For this, we introduce π-Distill, a joint teacher-student objective that trains a PI-conditioned teacher and an unconditioned student simultaneously using the same model. Additionally, we also introduce On-Policy Self-Distillation (OPSD), an alternative approach that trains using Reinforcement Learning (RL) with a reverse KL-penalty between the student and the PI-conditioned teacher. We show that both of these algorithms effectively distill frontier agents using action-only PI. Specifically we find that π-Distill and in some cases OPSD, outperform industry standard practices (Supervised finetuning followed by RL) that assume access to full Chain-of-Thought supervision across multiple agentic benchmarks, models, and forms of PI. We complement our results with extensive analysis that characterizes the factors enabling effective learning with PI, focusing primarily on π-Distill and characterizing when OPSD is competitive.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.04942</guid>
      <pubDate>Wed, 04 Feb 2026 18:46:17 +0000</pubDate>
    </item>
    <item>
      <title>Reinforced Attention Learning</title>
      <link>https://arxiv.org/abs/2602.04884</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.04884.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Bangzheng Li, Jianmo Ni, Chen Qu, Ian Miao, Liu Yang, Xingyu Fu, Muhao Chen, Derek Zhiyuan Cheng&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 11&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Post-training with Reinforcement Learning (RL) has substantially improved reasoning in Large Language Models (LLMs) via test-time scaling. However, extending this paradigm to Multimodal LLMs (MLLMs) through verbose rationales yields limited gains for perception and can even degrade performance.   We propose Reinforced Attention Learning (RAL), a policy-gradient framework that directly optimizes internal attention distributions rather than output token sequences. By shifting optimization from what to generate to where to attend, RAL promotes effective information allocation and improved grounding in complex multimodal inputs. Experiments across diverse image and video benchmarks show consistent gains over GRPO and other baselines. We further introduce On-Policy Attention Distillation, demonstrating that transferring latent attention behaviors yields stronger cross-modal alignment than standard knowledge distillation. Our results position attention policies as a principled and general alternative for multimodal post-training.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.04884.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Bangzheng Li, Jianmo Ni, Chen Qu, Ian Miao, Liu Yang, Xingyu Fu, Muhao Chen, Derek Zhiyuan Cheng&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 13&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Post-training with Reinforcement Learning (RL) has substantially improved reasoning in Large Language Models (LLMs) via test-time scaling. However, extending this paradigm to Multimodal LLMs (MLLMs) through verbose rationales yields limited gains for perception and can even degrade performance.   We propose Reinforced Attention Learning (RAL), a policy-gradient framework that directly optimizes internal attention distributions rather than output token sequences. By shifting optimization from what to generate to where to attend, RAL promotes effective information allocation and improved grounding in complex multimodal inputs. Experiments across diverse image and video benchmarks show consistent gains over GRPO and other baselines. We further introduce On-Policy Attention Distillation, demonstrating that transferring latent attention behaviors yields stronger cross-modal alignment than standard knowledge distillation. Our results position attention policies as a principled and general alternative for multimodal post-training.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.04884</guid>
      <pubDate>Wed, 04 Feb 2026 18:59:52 +0000</pubDate>
    </item>
@@ -165,28 +165,28 @@
    <item>
      <title>Towards Reducible Uncertainty Modeling for Reliable Large Language Model Agents</title>
      <link>https://arxiv.org/abs/2602.05073</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05073.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Changdae Oh, Seongheon Park, To Eun Kim, Jiatong Li, Wendi Li, Samuel Yeh, Xuefeng Du, Hamed Hassani, Paul Bogdan, Dawn Song, Sharon Li&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 7&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Uncertainty quantification (UQ) for large language models (LLMs) is a key building block for safety guardrails of daily LLM applications. Yet, even as LLM agents are increasingly deployed in highly complex tasks, most UQ research still centers on single-turn question-answering. We argue that UQ research must shift to realistic settings with interactive agents, and that a new principled framework for agent UQ is needed. This paper presents the first general formulation of agent UQ that subsumes broad classes of existing UQ setups. Under this formulation, we show that prior works implicitly treat LLM UQ as an uncertainty accumulation process, a viewpoint that breaks down for interactive agents in an open world. In contrast, we propose a novel perspective, a conditional uncertainty reduction process, that explicitly models reducible uncertainty over an agent's trajectory by highlighting "interactivity" of actions. From this perspective, we outline a conceptual framework to provide actionable guidance for designing UQ in LLM agent setups. Finally, we conclude with practical implications of the agent UQ in frontier LLM development and domain-specific applications, as well as open remaining problems.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05073.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Changdae Oh, Seongheon Park, To Eun Kim, Jiatong Li, Wendi Li, Samuel Yeh, Xuefeng Du, Hamed Hassani, Paul Bogdan, Dawn Song, Sharon Li&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 9&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Uncertainty quantification (UQ) for large language models (LLMs) is a key building block for safety guardrails of daily LLM applications. Yet, even as LLM agents are increasingly deployed in highly complex tasks, most UQ research still centers on single-turn question-answering. We argue that UQ research must shift to realistic settings with interactive agents, and that a new principled framework for agent UQ is needed. This paper presents the first general formulation of agent UQ that subsumes broad classes of existing UQ setups. Under this formulation, we show that prior works implicitly treat LLM UQ as an uncertainty accumulation process, a viewpoint that breaks down for interactive agents in an open world. In contrast, we propose a novel perspective, a conditional uncertainty reduction process, that explicitly models reducible uncertainty over an agent's trajectory by highlighting "interactivity" of actions. From this perspective, we outline a conceptual framework to provide actionable guidance for designing UQ in LLM agent setups. Finally, we conclude with practical implications of the agent UQ in frontier LLM development and domain-specific applications, as well as open remaining problems.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05073</guid>
      <pubDate>Wed, 04 Feb 2026 21:47:40 +0000</pubDate>
    </item>
    <item>
      <title>SocialVeil: Probing Social Intelligence of Language Agents under Communication Barriers</title>
      <link>https://arxiv.org/abs/2602.05115</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05115.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Keyang Xuan, Pengda Wang, Chongrui Ye, Haofei Yu, Tal August, Jiaxuan You&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 14&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language models (LLMs) are increasingly evaluated in interactive environments to test their social intelligence. However, existing benchmarks often assume idealized communication between agents, limiting our ability to diagnose whether LLMs can maintain and repair interactions in more realistic, imperfect settings. To close this gap, we present SocialVeil, a social learning environment that can simulate social interaction under cognitive-difference-induced communication barriers. Grounded in a systematic literature review of communication challenges in human interaction, SocialVeil introduces three representative types of such disruption, semantic vagueness, sociocultural mismatch, and emotional interference. We also introduce two barrier-aware evaluation metrics, unresolved confusion and mutual understanding, to evaluate interaction quality under impaired communication. Experiments across 720 scenarios and four frontier LLMs show that barriers consistently impair performance, with mutual understanding reduced by over 45\% on average, and confusion elevated by nearly 50\%. Human evaluations validate the fidelity of these simulated barriers (ICCapprox0.78, Pearson rapprox0.80). We further demonstrate that adaptation strategies (Repair Instruction and Interactive learning) only have a modest effect far from barrier-free performance. This work takes a step toward bringing social interaction environments closer to real-world communication, opening opportunities for exploring the social intelligence of LLM agents.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05115.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Keyang Xuan, Pengda Wang, Chongrui Ye, Haofei Yu, Tal August, Jiaxuan You&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 16&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language models (LLMs) are increasingly evaluated in interactive environments to test their social intelligence. However, existing benchmarks often assume idealized communication between agents, limiting our ability to diagnose whether LLMs can maintain and repair interactions in more realistic, imperfect settings. To close this gap, we present SocialVeil, a social learning environment that can simulate social interaction under cognitive-difference-induced communication barriers. Grounded in a systematic literature review of communication challenges in human interaction, SocialVeil introduces three representative types of such disruption, semantic vagueness, sociocultural mismatch, and emotional interference. We also introduce two barrier-aware evaluation metrics, unresolved confusion and mutual understanding, to evaluate interaction quality under impaired communication. Experiments across 720 scenarios and four frontier LLMs show that barriers consistently impair performance, with mutual understanding reduced by over 45\% on average, and confusion elevated by nearly 50\%. Human evaluations validate the fidelity of these simulated barriers (ICCapprox0.78, Pearson rapprox0.80). We further demonstrate that adaptation strategies (Repair Instruction and Interactive learning) only have a modest effect far from barrier-free performance. This work takes a step toward bringing social interaction environments closer to real-world communication, opening opportunities for exploring the social intelligence of LLM agents.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05115</guid>
      <pubDate>Wed, 04 Feb 2026 23:04:25 +0000</pubDate>
    </item>
    <item>
      <title>Semantic Search over 9 Million Mathematical Theorems</title>
      <link>https://arxiv.org/abs/2602.05216</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05216.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Luke Alexander, Eric Leonen, Sophie Szeto, Artemii Remizov, Ignacio Tejeda, Giovanni Inchiostro, Vasily Ilin&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 16&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Searching for mathematical results remains difficult: most existing tools retrieve entire papers, while mathematicians and theorem-proving agents often seek a specific theorem, lemma, or proposition that answers a query. While semantic search has seen rapid progress, its behavior on large, highly technical corpora such as research-level mathematical theorems remains poorly understood. In this work, we introduce and study semantic theorem retrieval at scale over a unified corpus of 9.2 million theorem statements extracted from arXiv and seven other sources, representing the largest publicly available corpus of human-authored, research-level theorems. We represent each theorem with a short natural-language description as a retrieval representation and systematically analyze how representation context, language model choice, embedding model, and prompting strategy affect retrieval quality. On a curated evaluation set of theorem-search queries written by professional mathematicians, our approach substantially improves both theorem-level and paper-level retrieval compared to existing baselines, demonstrating that semantic theorem search is feasible and effective at web scale. The theorem search tool is available at https://huggingface.co/spaces/uw-math-ai/theorem-search{this link}, and the dataset is available at https://huggingface.co/datasets/uw-math-ai/TheoremSearch{this link}.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05216.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Luke Alexander, Eric Leonen, Sophie Szeto, Artemii Remizov, Ignacio Tejeda, Giovanni Inchiostro, Vasily Ilin&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 17&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Searching for mathematical results remains difficult: most existing tools retrieve entire papers, while mathematicians and theorem-proving agents often seek a specific theorem, lemma, or proposition that answers a query. While semantic search has seen rapid progress, its behavior on large, highly technical corpora such as research-level mathematical theorems remains poorly understood. In this work, we introduce and study semantic theorem retrieval at scale over a unified corpus of 9.2 million theorem statements extracted from arXiv and seven other sources, representing the largest publicly available corpus of human-authored, research-level theorems. We represent each theorem with a short natural-language description as a retrieval representation and systematically analyze how representation context, language model choice, embedding model, and prompting strategy affect retrieval quality. On a curated evaluation set of theorem-search queries written by professional mathematicians, our approach substantially improves both theorem-level and paper-level retrieval compared to existing baselines, demonstrating that semantic theorem search is feasible and effective at web scale. The theorem search tool is available at https://huggingface.co/spaces/uw-math-ai/theorem-search{this link}, and the dataset is available at https://huggingface.co/datasets/uw-math-ai/TheoremSearch{this link}.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05216</guid>
      <pubDate>Thu, 05 Feb 2026 02:16:20 +0000</pubDate>
    </item>
    <item>
      <title>CoPE: Clipped RoPE as A Scalable Free Lunch for Long Context LLMs</title>
      <link>https://arxiv.org/abs/2602.05258</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05258.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Haoran Li, Sucheng Ren, Alan Yuille, Feng Wang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 4&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Rotary Positional Embedding (RoPE) is a key component of context scaling in Large Language Models (LLMs). While various methods have been proposed to adapt RoPE to longer contexts, their guiding principles generally fall into two categories: (1) out-of-distribution (OOD) mitigation, which scales RoPE frequencies to accommodate unseen positions, and (2) Semantic Modeling, which posits that the attention scores computed with RoPE should always prioritize semantically similar tokens. In this work, we unify these seemingly distinct objectives through a minimalist intervention, namely CoPE: soft clipping lowfrequency components of RoPE. CoPE not only eliminates OOD outliers and refines semantic signals, but also prevents spectral leakage caused by hard clipping. Extensive experiments demonstrate that simply applying our soft clipping strategy to RoPE yields significant performance gains that scale up to 256k context length, validating our theoretical analysis and establishing CoPE as a new state-of-the-art for length generalization. Our code, data, and models are available at https://github.com/hrlics/CoPE.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05258.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Haoran Li, Sucheng Ren, Alan Yuille, Feng Wang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 5&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Rotary Positional Embedding (RoPE) is a key component of context scaling in Large Language Models (LLMs). While various methods have been proposed to adapt RoPE to longer contexts, their guiding principles generally fall into two categories: (1) out-of-distribution (OOD) mitigation, which scales RoPE frequencies to accommodate unseen positions, and (2) Semantic Modeling, which posits that the attention scores computed with RoPE should always prioritize semantically similar tokens. In this work, we unify these seemingly distinct objectives through a minimalist intervention, namely CoPE: soft clipping lowfrequency components of RoPE. CoPE not only eliminates OOD outliers and refines semantic signals, but also prevents spectral leakage caused by hard clipping. Extensive experiments demonstrate that simply applying our soft clipping strategy to RoPE yields significant performance gains that scale up to 256k context length, validating our theoretical analysis and establishing CoPE as a new state-of-the-art for length generalization. Our code, data, and models are available at https://github.com/hrlics/CoPE.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05258</guid>
      <pubDate>Thu, 05 Feb 2026 03:31:14 +0000</pubDate>
    </item>
@@ -207,21 +207,21 @@
    <item>
      <title>ProAct: Agentic Lookahead in Interactive Environments</title>
      <link>https://arxiv.org/abs/2602.05327</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05327.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Yangbin Yu, Mingyu Yang, Junyou Li, Yiming Gao, Feiyu Liu, Yijun Yang, Zichuan Lin, Jiafei Lyu, Yicheng Liu, Zhicong Lu, Deheng Ye, Jie Jiang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 19&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Existing Large Language Model (LLM) agents struggle in interactive environments requiring long-horizon planning, primarily due to compounding errors when simulating future states. To address this, we propose ProAct, a framework that enables agents to internalize accurate lookahead reasoning through a two-stage training paradigm. First, we introduce Grounded LookAhead Distillation (GLAD), where the agent undergoes supervised fine-tuning on trajectories derived from environment-based search. By compressing complex search trees into concise, causal reasoning chains, the agent learns the logic of foresight without the computational overhead of inference-time search. Second, to further refine decision accuracy, we propose the Monte-Carlo Critic (MC-Critic), a plug-and-play auxiliary value estimator designed to enhance policy-gradient algorithms like PPO and GRPO. By leveraging lightweight environment rollouts to calibrate value estimates, MC-Critic provides a low-variance signal that facilitates stable policy optimization without relying on expensive model-based value approximation. Experiments on both stochastic (e.g., 2048) and deterministic (e.g., Sokoban) environments demonstrate that ProAct significantly improves planning accuracy. Notably, a 4B parameter model trained with ProAct outperforms all open-source baselines and rivals state-of-the-art closed-source models, while demonstrating robust generalization to unseen environments. The codes and models are available at https://github.com/GreatX3/ProAct&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05327.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Yangbin Yu, Mingyu Yang, Junyou Li, Yiming Gao, Feiyu Liu, Yijun Yang, Zichuan Lin, Jiafei Lyu, Yicheng Liu, Zhicong Lu, Deheng Ye, Jie Jiang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 22&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Existing Large Language Model (LLM) agents struggle in interactive environments requiring long-horizon planning, primarily due to compounding errors when simulating future states. To address this, we propose ProAct, a framework that enables agents to internalize accurate lookahead reasoning through a two-stage training paradigm. First, we introduce Grounded LookAhead Distillation (GLAD), where the agent undergoes supervised fine-tuning on trajectories derived from environment-based search. By compressing complex search trees into concise, causal reasoning chains, the agent learns the logic of foresight without the computational overhead of inference-time search. Second, to further refine decision accuracy, we propose the Monte-Carlo Critic (MC-Critic), a plug-and-play auxiliary value estimator designed to enhance policy-gradient algorithms like PPO and GRPO. By leveraging lightweight environment rollouts to calibrate value estimates, MC-Critic provides a low-variance signal that facilitates stable policy optimization without relying on expensive model-based value approximation. Experiments on both stochastic (e.g., 2048) and deterministic (e.g., Sokoban) environments demonstrate that ProAct significantly improves planning accuracy. Notably, a 4B parameter model trained with ProAct outperforms all open-source baselines and rivals state-of-the-art closed-source models, while demonstrating robust generalization to unseen environments. The codes and models are available at https://github.com/GreatX3/ProAct&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05327</guid>
      <pubDate>Thu, 05 Feb 2026 05:45:16 +0000</pubDate>
    </item>
    <item>
      <title>Spider-Sense: Intrinsic Risk Sensing for Efficient Agent Defense with Hierarchical Adaptive Screening</title>
      <link>https://arxiv.org/abs/2602.05386</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05386.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Zhenxiong Yu, Zhi Yang, Zhiheng Jin, Shuhe Wang, Heng Zhang, Yanlin Fei, Lingfeng Zeng, Fangqi Lou, Shuo Zhang, Tu Hu, Jingping Liu, Rongze Chen, Xingyu Zhu, Kunyi Wang, Chaofa Yuan, Xin Guo, Zhaowei Liu, Feipeng Zhang, Jie Huang, Huacan Wang, Ronghao Chen, Liwen Zhang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 56&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; As large language models (LLMs) evolve into autonomous agents, their real-world applicability has expanded significantly, accompanied by new security challenges. Most existing agent defense mechanisms adopt a mandatory checking paradigm, in which security validation is forcibly triggered at predefined stages of the agent lifecycle. In this work, we argue that effective agent security should be intrinsic and selective rather than architecturally decoupled and mandatory. We propose Spider-Sense framework, an event-driven defense framework based on Intrinsic Risk Sensing (IRS), which allows agents to maintain latent vigilance and trigger defenses only upon risk perception. Once triggered, the Spider-Sense invokes a hierarchical defence mechanism that trades off efficiency and precision: it resolves known patterns via lightweight similarity matching while escalating ambiguous cases to deep internal reasoning, thereby eliminating reliance on external models. To facilitate rigorous evaluation, we introduce S^2Bench, a lifecycle-aware benchmark featuring realistic tool execution and multi-stage attacks. Extensive experiments demonstrate that Spider-Sense achieves competitive or superior defense performance, attaining the lowest Attack Success Rate (ASR) and False Positive Rate (FPR), with only a marginal latency overhead of 8.3\%.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05386.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Zhenxiong Yu, Zhi Yang, Zhiheng Jin, Shuhe Wang, Heng Zhang, Yanlin Fei, Lingfeng Zeng, Fangqi Lou, Shuo Zhang, Tu Hu, Jingping Liu, Rongze Chen, Xingyu Zhu, Kunyi Wang, Chaofa Yuan, Xin Guo, Zhaowei Liu, Feipeng Zhang, Jie Huang, Huacan Wang, Ronghao Chen, Liwen Zhang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 63&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; As large language models (LLMs) evolve into autonomous agents, their real-world applicability has expanded significantly, accompanied by new security challenges. Most existing agent defense mechanisms adopt a mandatory checking paradigm, in which security validation is forcibly triggered at predefined stages of the agent lifecycle. In this work, we argue that effective agent security should be intrinsic and selective rather than architecturally decoupled and mandatory. We propose Spider-Sense framework, an event-driven defense framework based on Intrinsic Risk Sensing (IRS), which allows agents to maintain latent vigilance and trigger defenses only upon risk perception. Once triggered, the Spider-Sense invokes a hierarchical defence mechanism that trades off efficiency and precision: it resolves known patterns via lightweight similarity matching while escalating ambiguous cases to deep internal reasoning, thereby eliminating reliance on external models. To facilitate rigorous evaluation, we introduce S^2Bench, a lifecycle-aware benchmark featuring realistic tool execution and multi-stage attacks. Extensive experiments demonstrate that Spider-Sense achieves competitive or superior defense performance, attaining the lowest Attack Success Rate (ASR) and False Positive Rate (FPR), with only a marginal latency overhead of 8.3\%.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05386</guid>
      <pubDate>Thu, 05 Feb 2026 07:11:05 +0000</pubDate>
    </item>
    <item>
      <title>Late-to-Early Training: LET LLMs Learn Earlier, So Faster and Better</title>
      <link>https://arxiv.org/abs/2602.05393</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05393.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Ji Zhao, Yufei Gu, Shitong Shao, Xun Zhou, Liang Xiang, Zeke Xie&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 3&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; As Large Language Models (LLMs) achieve remarkable empirical success through scaling model and data size, pretraining has become increasingly critical yet computationally prohibitive, hindering rapid development. Despite the availability of numerous pretrained LLMs developed at significant computational expense, a fundamental real-world question remains underexplored: Can we leverage existing small pretrained models to accelerate the training of larger models? In this paper, we propose a Late-to-Early Training (LET) paradigm that enables LLMs to explicitly learn later knowledge in earlier steps and earlier layers. The core idea is to guide the early layers of an LLM during early training using representations from the late layers of a pretrained (i.e. late training phase) model. We identify two key mechanisms that drive LET's effectiveness: late-to-early-step learning and late-to-early-layer learning. These mechanisms significantly accelerate training convergence while robustly enhancing both language modeling capabilities and downstream task performance, enabling faster training with superior performance. Extensive experiments on 1.4B and 7B parameter models demonstrate LET's efficiency and effectiveness. Notably, when training a 1.4B LLM on the Pile dataset, our method achieves up to 1.6times speedup with nearly 5\% improvement in downstream task accuracy compared to standard training, even when using a pretrained model with 10times fewer parameters than the target model.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05393.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Ji Zhao, Yufei Gu, Shitong Shao, Xun Zhou, Liang Xiang, Zeke Xie&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 4&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; As Large Language Models (LLMs) achieve remarkable empirical success through scaling model and data size, pretraining has become increasingly critical yet computationally prohibitive, hindering rapid development. Despite the availability of numerous pretrained LLMs developed at significant computational expense, a fundamental real-world question remains underexplored: Can we leverage existing small pretrained models to accelerate the training of larger models? In this paper, we propose a Late-to-Early Training (LET) paradigm that enables LLMs to explicitly learn later knowledge in earlier steps and earlier layers. The core idea is to guide the early layers of an LLM during early training using representations from the late layers of a pretrained (i.e. late training phase) model. We identify two key mechanisms that drive LET's effectiveness: late-to-early-step learning and late-to-early-layer learning. These mechanisms significantly accelerate training convergence while robustly enhancing both language modeling capabilities and downstream task performance, enabling faster training with superior performance. Extensive experiments on 1.4B and 7B parameter models demonstrate LET's efficiency and effectiveness. Notably, when training a 1.4B LLM on the Pile dataset, our method achieves up to 1.6times speedup with nearly 5\% improvement in downstream task accuracy compared to standard training, even when using a pretrained model with 10times fewer parameters than the target model.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05393</guid>
      <pubDate>Thu, 05 Feb 2026 07:19:34 +0000</pubDate>
    </item>
@@ -235,7 +235,7 @@
    <item>
      <title>Multi-Task GRPO: Reliable LLM Reasoning Across Tasks</title>
      <link>https://arxiv.org/abs/2602.05547</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05547.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shyam Sundhar Ramesh, Xiaotong Ji, Matthieu Zimmer, Sangwoong Yoon, Zhiyong Wang, Haitham Bou Ammar, Aurelien Lucchi, Ilija Bogunovic&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 6&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; RL-based post-training with GRPO is widely used to improve large language models on individual reasoning tasks. However, real-world deployment requires reliable performance across diverse tasks. A straightforward multi-task adaptation of GRPO often leads to imbalanced outcomes, with some tasks dominating optimization while others stagnate. Moreover, tasks can vary widely in how frequently prompts yield zero advantages (and thus zero gradients), which further distorts their effective contribution to the optimization signal. To address these issues, we propose a novel Multi-Task GRPO (MT-GRPO) algorithm that (i) dynamically adapts task weights to explicitly optimize worst-task performance and promote balanced progress across tasks, and (ii) introduces a ratio-preserving sampler to ensure task-wise policy gradients reflect the adapted weights. Experiments on both 3-task and 9-task settings show that MT-GRPO consistently outperforms baselines in worst-task accuracy. In particular, MT-GRPO achieves 16-28% and 6% absolute improvement on worst-task performance over standard GRPO and DAPO, respectively, while maintaining competitive average accuracy. Moreover, MT-GRPO requires 50% fewer training steps to reach 50% worst-task accuracy in the 3-task setting, demonstrating substantially improved efficiency in achieving reliable performance across tasks.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05547.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shyam Sundhar Ramesh, Xiaotong Ji, Matthieu Zimmer, Sangwoong Yoon, Zhiyong Wang, Haitham Bou Ammar, Aurelien Lucchi, Ilija Bogunovic&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 7&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; RL-based post-training with GRPO is widely used to improve large language models on individual reasoning tasks. However, real-world deployment requires reliable performance across diverse tasks. A straightforward multi-task adaptation of GRPO often leads to imbalanced outcomes, with some tasks dominating optimization while others stagnate. Moreover, tasks can vary widely in how frequently prompts yield zero advantages (and thus zero gradients), which further distorts their effective contribution to the optimization signal. To address these issues, we propose a novel Multi-Task GRPO (MT-GRPO) algorithm that (i) dynamically adapts task weights to explicitly optimize worst-task performance and promote balanced progress across tasks, and (ii) introduces a ratio-preserving sampler to ensure task-wise policy gradients reflect the adapted weights. Experiments on both 3-task and 9-task settings show that MT-GRPO consistently outperforms baselines in worst-task accuracy. In particular, MT-GRPO achieves 16-28% and 6% absolute improvement on worst-task performance over standard GRPO and DAPO, respectively, while maintaining competitive average accuracy. Moreover, MT-GRPO requires 50% fewer training steps to reach 50% worst-task accuracy in the 3-task setting, demonstrating substantially improved efficiency in achieving reliable performance across tasks.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05547</guid>
      <pubDate>Thu, 05 Feb 2026 11:06:37 +0000</pubDate>
    </item>
@@ -249,28 +249,28 @@
    <item>
      <title>Reinforcement World Model Learning for LLM-based Agents</title>
      <link>https://arxiv.org/abs/2602.05842</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05842.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Xiao Yu, Baolin Peng, Ruize Xu, Yelong Shen, Pengcheng He, Suman Nath, Nikhil Singh, Jiangfeng Gao, Zhou Yu&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 11&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language models (LLMs) have achieved strong performance in language-centric tasks. However, in agentic settings, LLMs often struggle to anticipate action consequences and adapt to environment dynamics, highlighting the need for world-modeling capabilities in LLM-based agents. We propose Reinforcement World Model Learning (RWML), a self-supervised method that learns action-conditioned world models for LLM-based agents on textual states using sim-to-real gap rewards. Our method aligns simulated next states produced by the model with realized next states observed from the environment, encouraging consistency between internal world simulations and actual environment dynamics in a pre-trained embedding space. Unlike next-state token prediction, which prioritizes token-level fidelity (i.e., reproducing exact wording) over semantic equivalence and can lead to model collapse, our method provides a more robust training signal and is empirically less susceptible to reward hacking than LLM-as-a-judge. We evaluate our method on ALFWorld and τ^2 Bench and observe significant gains over the base model, despite being entirely self-supervised. When combined with task-success rewards, our method outperforms direct task-success reward RL by 6.9 and 5.7 points on ALFWorld and τ^2 Bench respectively, while matching the performance of expert-data training.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05842.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Xiao Yu, Baolin Peng, Ruize Xu, Yelong Shen, Pengcheng He, Suman Nath, Nikhil Singh, Jiangfeng Gao, Zhou Yu&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 17&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Large language models (LLMs) have achieved strong performance in language-centric tasks. However, in agentic settings, LLMs often struggle to anticipate action consequences and adapt to environment dynamics, highlighting the need for world-modeling capabilities in LLM-based agents. We propose Reinforcement World Model Learning (RWML), a self-supervised method that learns action-conditioned world models for LLM-based agents on textual states using sim-to-real gap rewards. Our method aligns simulated next states produced by the model with realized next states observed from the environment, encouraging consistency between internal world simulations and actual environment dynamics in a pre-trained embedding space. Unlike next-state token prediction, which prioritizes token-level fidelity (i.e., reproducing exact wording) over semantic equivalence and can lead to model collapse, our method provides a more robust training signal and is empirically less susceptible to reward hacking than LLM-as-a-judge. We evaluate our method on ALFWorld and τ^2 Bench and observe significant gains over the base model, despite being entirely self-supervised. When combined with task-success rewards, our method outperforms direct task-success reward RL by 6.9 and 5.7 points on ALFWorld and τ^2 Bench respectively, while matching the performance of expert-data training.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05842</guid>
      <pubDate>Thu, 05 Feb 2026 16:30:08 +0000</pubDate>
    </item>
    <item>
      <title>BABE: Biology Arena BEnchmark</title>
      <link>https://arxiv.org/abs/2602.05857</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05857.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Junting Zhou, Jin Chen, Linfeng Hao, Denghui Cao, Zheyu Wang, Qiguang Chen, Chaoyou Fu, Jiaze Chen, Yuchen Wu, Ge Zhang, Mingxuan Wang, Wenhao Huang, Tong Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 5&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; The rapid evolution of large language models (LLMs) has expanded their capabilities from basic dialogue to advanced scientific reasoning. However, existing benchmarks in biology often fail to assess a critical skill required of researchers: the ability to integrate experimental results with contextual knowledge to derive meaningful conclusions. To address this gap, we introduce BABE(Biology Arena BEnchmark), a comprehensive benchmark designed to evaluate the experimental reasoning capabilities of biological AI systems. BABE is uniquely constructed from peer-reviewed research papers and real-world biological studies, ensuring that tasks reflect the complexity and interdisciplinary nature of actual scientific inquiry. BABE challenges models to perform causal reasoning and cross-scale inference. Our benchmark provides a robust framework for assessing how well AI systems can reason like practicing scientists, offering a more authentic measure of their potential to contribute to biological research.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05857.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Junting Zhou, Jin Chen, Linfeng Hao, Denghui Cao, Zheyu Wang, Qiguang Chen, Chaoyou Fu, Jiaze Chen, Yuchen Wu, Ge Zhang, Mingxuan Wang, Wenhao Huang, Tong Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 8&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; The rapid evolution of large language models (LLMs) has expanded their capabilities from basic dialogue to advanced scientific reasoning. However, existing benchmarks in biology often fail to assess a critical skill required of researchers: the ability to integrate experimental results with contextual knowledge to derive meaningful conclusions. To address this gap, we introduce BABE(Biology Arena BEnchmark), a comprehensive benchmark designed to evaluate the experimental reasoning capabilities of biological AI systems. BABE is uniquely constructed from peer-reviewed research papers and real-world biological studies, ensuring that tasks reflect the complexity and interdisciplinary nature of actual scientific inquiry. BABE challenges models to perform causal reasoning and cross-scale inference. Our benchmark provides a robust framework for assessing how well AI systems can reason like practicing scientists, offering a more authentic measure of their potential to contribute to biological research.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05857</guid>
      <pubDate>Thu, 05 Feb 2026 16:39:20 +0000</pubDate>
    </item>
    <item>
      <title>Pathwise Test-Time Correction for Autoregressive Long Video Generation</title>
      <link>https://arxiv.org/abs/2602.05871</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05871.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Xunzhi Xiang, Zixuan Duan, Guiyu Zhang, Haiyu Zhang, Zhe Gao, Junta Wu, Shaofeng Zhang, Tengfei Wang, Qi Fan, Chunchao Guo&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 2&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Distilled autoregressive diffusion models facilitate real-time short video synthesis but suffer from severe error accumulation during long-sequence generation. While existing Test-Time Optimization (TTO) methods prove effective for images or short clips, we identify that they fail to mitigate drift in extended sequences due to unstable reward landscapes and the hypersensitivity of distilled parameters. To overcome these limitations, we introduce Test-Time Correction (TTC), a training-free alternative. Specifically, TTC utilizes the initial frame as a stable reference anchor to calibrate intermediate stochastic states along the sampling trajectory. Extensive experiments demonstrate that our method seamlessly integrates with various distilled models, extending generation lengths with negligible overhead while matching the quality of resource-intensive training-based methods on 30-second benchmarks.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05871.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Xunzhi Xiang, Zixuan Duan, Guiyu Zhang, Haiyu Zhang, Zhe Gao, Junta Wu, Shaofeng Zhang, Tengfei Wang, Qi Fan, Chunchao Guo&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 3&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Distilled autoregressive diffusion models facilitate real-time short video synthesis but suffer from severe error accumulation during long-sequence generation. While existing Test-Time Optimization (TTO) methods prove effective for images or short clips, we identify that they fail to mitigate drift in extended sequences due to unstable reward landscapes and the hypersensitivity of distilled parameters. To overcome these limitations, we introduce Test-Time Correction (TTC), a training-free alternative. Specifically, TTC utilizes the initial frame as a stable reference anchor to calibrate intermediate stochastic states along the sampling trajectory. Extensive experiments demonstrate that our method seamlessly integrates with various distilled models, extending generation lengths with negligible overhead while matching the quality of resource-intensive training-based methods on 30-second benchmarks.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05871</guid>
      <pubDate>Thu, 05 Feb 2026 16:50:39 +0000</pubDate>
    </item>
    <item>
      <title>Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations</title>
      <link>https://arxiv.org/abs/2602.05885</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05885.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Wei Liu, Jiawei Xu, Yingru Li, Longtao Zheng, Tianjian Li, Qian Liu, Junxian He&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 18&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; High-quality kernel is critical for scalable AI systems, and enabling LLMs to generate such code would advance AI development. However, training LLMs for this task requires sufficient data, a robust environment, and the process is often vulnerable to reward hacking and lazy optimization. In these cases, models may hack training rewards and prioritize trivial correctness over meaningful speedup. In this paper, we systematically study reinforcement learning (RL) for kernel generation. We first design KernelGYM, a robust distributed GPU environment that supports reward hacking check, data collection from multi-turn interactions and long-term RL training. Building on KernelGYM, we investigate effective multi-turn RL methods and identify a biased policy gradient issue caused by self-inclusion in GRPO. To solve this, we propose Turn-level Reinforce-Leave-One-Out (TRLOO) to provide unbiased advantage estimation for multi-turn RL. To alleviate lazy optimization, we incorporate mismatch correction for training stability and introduce Profiling-based Rewards (PR) and Profiling-based Rejection Sampling (PRS) to overcome the issue. The trained model, Dr.Kernel-14B, reaches performance competitive with Claude-4.5-Sonnet in Kernelbench. Finally, we study sequential test-time scaling for Dr.Kernel-14B. On the KernelBench Level-2 subset, 31.6% of the generated kernels achieve at least a 1.2x speedup over the Torch reference, surpassing Claude-4.5-Sonnet (26.7%) and GPT-5 (28.6%). When selecting the best candidate across all turns, this 1.2x speedup rate further increases to 47.8%. All resources, including environment, training code, models, and dataset, are included in https://www.github.com/hkust-nlp/KernelGYM.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05885.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Wei Liu, Jiawei Xu, Yingru Li, Longtao Zheng, Tianjian Li, Qian Liu, Junxian He&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 19&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; High-quality kernel is critical for scalable AI systems, and enabling LLMs to generate such code would advance AI development. However, training LLMs for this task requires sufficient data, a robust environment, and the process is often vulnerable to reward hacking and lazy optimization. In these cases, models may hack training rewards and prioritize trivial correctness over meaningful speedup. In this paper, we systematically study reinforcement learning (RL) for kernel generation. We first design KernelGYM, a robust distributed GPU environment that supports reward hacking check, data collection from multi-turn interactions and long-term RL training. Building on KernelGYM, we investigate effective multi-turn RL methods and identify a biased policy gradient issue caused by self-inclusion in GRPO. To solve this, we propose Turn-level Reinforce-Leave-One-Out (TRLOO) to provide unbiased advantage estimation for multi-turn RL. To alleviate lazy optimization, we incorporate mismatch correction for training stability and introduce Profiling-based Rewards (PR) and Profiling-based Rejection Sampling (PRS) to overcome the issue. The trained model, Dr.Kernel-14B, reaches performance competitive with Claude-4.5-Sonnet in Kernelbench. Finally, we study sequential test-time scaling for Dr.Kernel-14B. On the KernelBench Level-2 subset, 31.6% of the generated kernels achieve at least a 1.2x speedup over the Torch reference, surpassing Claude-4.5-Sonnet (26.7%) and GPT-5 (28.6%). When selecting the best candidate across all turns, this 1.2x speedup rate further increases to 47.8%. All resources, including environment, training code, models, and dataset, are included in https://www.github.com/hkust-nlp/KernelGYM.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05885</guid>
      <pubDate>Thu, 05 Feb 2026 17:01:09 +0000</pubDate>
    </item>
@@ -284,21 +284,21 @@
    <item>
      <title>SAGE: Benchmarking and Improving Retrieval for Deep Research Agents</title>
      <link>https://arxiv.org/abs/2602.05975</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05975.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Tiansheng Hu, Yilun Zhao, Canyu Zhang, Arman Cohan, Chen Zhao&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 9&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Deep research agents have emerged as powerful systems for addressing complex queries. Meanwhile, LLM-based retrievers have demonstrated strong capability in following instructions or reasoning. This raises a critical question: can LLM-based retrievers effectively contribute to deep research agent workflows? To investigate this, we introduce SAGE, a benchmark for scientific literature retrieval comprising 1,200 queries across four scientific domains, with a 200,000 paper retrieval corpus.We evaluate six deep research agents and find that all systems struggle with reasoning-intensive retrieval. Using DR Tulu as backbone, we further compare BM25 and LLM-based retrievers (i.e., ReasonIR and gte-Qwen2-7B-instruct) as alternative search tools. Surprisingly, BM25 significantly outperforms LLM-based retrievers by approximately 30%, as existing agents generate keyword-oriented sub-queries. To improve performance, we propose a corpus-level test-time scaling framework that uses LLMs to augment documents with metadata and keywords, making retrieval easier for off-the-shelf retrievers. This yields 8% and 2% gains on short-form and open-ended questions, respectively.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05975.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Tiansheng Hu, Yilun Zhao, Canyu Zhang, Arman Cohan, Chen Zhao&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 10&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Deep research agents have emerged as powerful systems for addressing complex queries. Meanwhile, LLM-based retrievers have demonstrated strong capability in following instructions or reasoning. This raises a critical question: can LLM-based retrievers effectively contribute to deep research agent workflows? To investigate this, we introduce SAGE, a benchmark for scientific literature retrieval comprising 1,200 queries across four scientific domains, with a 200,000 paper retrieval corpus.We evaluate six deep research agents and find that all systems struggle with reasoning-intensive retrieval. Using DR Tulu as backbone, we further compare BM25 and LLM-based retrievers (i.e., ReasonIR and gte-Qwen2-7B-instruct) as alternative search tools. Surprisingly, BM25 significantly outperforms LLM-based retrievers by approximately 30%, as existing agents generate keyword-oriented sub-queries. To improve performance, we propose a corpus-level test-time scaling framework that uses LLMs to augment documents with metadata and keywords, making retrieval easier for off-the-shelf retrievers. This yields 8% and 2% gains on short-form and open-ended questions, respectively.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05975</guid>
      <pubDate>Thu, 05 Feb 2026 18:25:24 +0000</pubDate>
    </item>
    <item>
      <title>RISE-Video: Can Video Generators Decode Implicit World Rules?</title>
      <link>https://arxiv.org/abs/2602.05986</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05986.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Mingxin Liu, Shuran Ma, Shibei Meng, Xiangyu Zhao, Zicheng Zhang, Shaofeng Zhang, Zhihang Zhong, Peixian Chen, Haoyu Cao, Xing Sun, Haodong Duan, Xue Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 23&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; While generative video models have achieved remarkable visual fidelity, their capacity to internalize and reason over implicit world rules remains a critical yet under-explored frontier. To bridge this gap, we present RISE-Video, a pioneering reasoning-oriented benchmark for Text-Image-to-Video (TI2V) synthesis that shifts the evaluative focus from surface-level aesthetics to deep cognitive reasoning. RISE-Video comprises 467 meticulously human-annotated samples spanning eight rigorous categories, providing a structured testbed for probing model intelligence across diverse dimensions, ranging from commonsense and spatial dynamics to specialized subject domains. Our framework introduces a multi-dimensional evaluation protocol consisting of four metrics: Reasoning Alignment, Temporal Consistency, Physical Rationality, and Visual Quality. To further support scalable evaluation, we propose an automated pipeline leveraging Large Multimodal Models (LMMs) to emulate human-centric assessment. Extensive experiments on 11 state-of-the-art TI2V models reveal pervasive deficiencies in simulating complex scenarios under implicit constraints, offering critical insights for the advancement of future world-simulating generative models.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.05986.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Mingxin Liu, Shuran Ma, Shibei Meng, Xiangyu Zhao, Zicheng Zhang, Shaofeng Zhang, Zhihang Zhong, Peixian Chen, Haoyu Cao, Xing Sun, Haodong Duan, Xue Yang&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 26&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; While generative video models have achieved remarkable visual fidelity, their capacity to internalize and reason over implicit world rules remains a critical yet under-explored frontier. To bridge this gap, we present RISE-Video, a pioneering reasoning-oriented benchmark for Text-Image-to-Video (TI2V) synthesis that shifts the evaluative focus from surface-level aesthetics to deep cognitive reasoning. RISE-Video comprises 467 meticulously human-annotated samples spanning eight rigorous categories, providing a structured testbed for probing model intelligence across diverse dimensions, ranging from commonsense and spatial dynamics to specialized subject domains. Our framework introduces a multi-dimensional evaluation protocol consisting of four metrics: Reasoning Alignment, Temporal Consistency, Physical Rationality, and Visual Quality. To further support scalable evaluation, we propose an automated pipeline leveraging Large Multimodal Models (LMMs) to emulate human-centric assessment. Extensive experiments on 11 state-of-the-art TI2V models reveal pervasive deficiencies in simulating complex scenarios under implicit constraints, offering critical insights for the advancement of future world-simulating generative models.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.05986</guid>
      <pubDate>Thu, 05 Feb 2026 18:36:10 +0000</pubDate>
    </item>
    <item>
      <title>Context Forcing: Consistent Autoregressive Video Generation with Long Context</title>
      <link>https://arxiv.org/abs/2602.06028</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06028.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shuo Chen, Cong Wei, Sun Sun, Ping Nie, Kai Zhou, Ge Zhang, Ming-Hsuan Yang, Wenhu Chen&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 25&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Recent approaches to real-time long video generation typically employ streaming tuning strategies, attempting to train a long-context student using a short-context (memoryless) teacher. In these frameworks, the student performs long rollouts but receives supervision from a teacher limited to short 5-second windows. This structural discrepancy creates a critical student-teacher mismatch: the teacher's inability to access long-term history prevents it from guiding the student on global temporal dependencies, effectively capping the student's context length. To resolve this, we propose Context Forcing, a novel framework that trains a long-context student via a long-context teacher. By ensuring the teacher is aware of the full generation history, we eliminate the supervision mismatch, enabling the robust training of models capable of long-term consistency. To make this computationally feasible for extreme durations (e.g., 2 minutes), we introduce a context management system that transforms the linearly growing context into a Slow-Fast Memory architecture, significantly reducing visual redundancy. Extensive results demonstrate that our method enables effective context lengths exceeding 20 seconds -- 2 to 10 times longer than state-of-the-art methods like LongLive and Infinite-RoPE. By leveraging this extended context, Context Forcing preserves superior consistency across long durations, surpassing state-of-the-art baselines on various long video evaluation metrics.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06028.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Shuo Chen, Cong Wei, Sun Sun, Ping Nie, Kai Zhou, Ge Zhang, Ming-Hsuan Yang, Wenhu Chen&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 29&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Recent approaches to real-time long video generation typically employ streaming tuning strategies, attempting to train a long-context student using a short-context (memoryless) teacher. In these frameworks, the student performs long rollouts but receives supervision from a teacher limited to short 5-second windows. This structural discrepancy creates a critical student-teacher mismatch: the teacher's inability to access long-term history prevents it from guiding the student on global temporal dependencies, effectively capping the student's context length. To resolve this, we propose Context Forcing, a novel framework that trains a long-context student via a long-context teacher. By ensuring the teacher is aware of the full generation history, we eliminate the supervision mismatch, enabling the robust training of models capable of long-term consistency. To make this computationally feasible for extreme durations (e.g., 2 minutes), we introduce a context management system that transforms the linearly growing context into a Slow-Fast Memory architecture, significantly reducing visual redundancy. Extensive results demonstrate that our method enables effective context lengths exceeding 20 seconds -- 2 to 10 times longer than state-of-the-art methods like LongLive and Infinite-RoPE. By leveraging this extended context, Context Forcing preserves superior consistency across long durations, surpassing state-of-the-art baselines on various long video evaluation metrics.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.06028</guid>
      <pubDate>Thu, 05 Feb 2026 18:58:01 +0000</pubDate>
    </item>
@@ -319,21 +319,21 @@
    <item>
      <title>InterPrior: Scaling Generative Control for Physics-Based Human-Object Interactions</title>
      <link>https://arxiv.org/abs/2602.06035</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06035.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Sirui Xu, Samuel Schulter, Morteza Ziyadi, Xialin He, Xiaohan Fei, Yu-Xiong Wang, Liangyan Gui&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 15&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Humans rarely plan whole-body interactions with objects at the level of explicit whole-body movements. High-level intentions, such as affordance, define the goal, while coordinated balance, contact, and manipulation can emerge naturally from underlying physical and motor priors. Scaling such priors is key to enabling humanoids to compose and generalize loco-manipulation skills across diverse contexts while maintaining physically coherent whole-body coordination. To this end, we introduce InterPrior, a scalable framework that learns a unified generative controller through large-scale imitation pretraining and post-training by reinforcement learning. InterPrior first distills a full-reference imitation expert into a versatile, goal-conditioned variational policy that reconstructs motion from multimodal observations and high-level intent. While the distilled policy reconstructs training behaviors, it does not generalize reliably due to the vast configuration space of large-scale human-object interactions. To address this, we apply data augmentation with physical perturbations, and then perform reinforcement learning finetuning to improve competence on unseen goals and initializations. Together, these steps consolidate the reconstructed latent skills into a valid manifold, yielding a motion prior that generalizes beyond the training data, e.g., it can incorporate new behaviors such as interactions with unseen objects. We further demonstrate its effectiveness for user-interactive control and its potential for real robot deployment.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06035.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Sirui Xu, Samuel Schulter, Morteza Ziyadi, Xialin He, Xiaohan Fei, Yu-Xiong Wang, Liangyan Gui&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 18&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Humans rarely plan whole-body interactions with objects at the level of explicit whole-body movements. High-level intentions, such as affordance, define the goal, while coordinated balance, contact, and manipulation can emerge naturally from underlying physical and motor priors. Scaling such priors is key to enabling humanoids to compose and generalize loco-manipulation skills across diverse contexts while maintaining physically coherent whole-body coordination. To this end, we introduce InterPrior, a scalable framework that learns a unified generative controller through large-scale imitation pretraining and post-training by reinforcement learning. InterPrior first distills a full-reference imitation expert into a versatile, goal-conditioned variational policy that reconstructs motion from multimodal observations and high-level intent. While the distilled policy reconstructs training behaviors, it does not generalize reliably due to the vast configuration space of large-scale human-object interactions. To address this, we apply data augmentation with physical perturbations, and then perform reinforcement learning finetuning to improve competence on unseen goals and initializations. Together, these steps consolidate the reconstructed latent skills into a valid manifold, yielding a motion prior that generalizes beyond the training data, e.g., it can incorporate new behaviors such as interactions with unseen objects. We further demonstrate its effectiveness for user-interactive control and its potential for real robot deployment.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.06035</guid>
      <pubDate>Thu, 05 Feb 2026 18:59:27 +0000</pubDate>
    </item>
    <item>
      <title>DFlash: Block Diffusion for Flash Speculative Decoding</title>
      <link>https://arxiv.org/abs/2602.06036</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06036.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Jian Chen, Yesheng Liang, Zhijian Liu&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 10&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast draft model whose outputs are verified in parallel by the target LLM; however, existing methods still rely on autoregressive drafting, which remains sequential and limits practical speedups. Diffusion LLMs offer a promising alternative by enabling parallel generation, but current diffusion models typically underperform compared with autoregressive models. In this paper, we introduce DFlash, a speculative decoding framework that employs a lightweight block diffusion model for parallel drafting. By generating draft tokens in a single forward pass and conditioning the draft model on context features extracted from the target model, DFlash enables efficient drafting with high-quality outputs and higher acceptance rates. Experiments show that DFlash achieves over 6x lossless acceleration across a range of models and tasks, delivering up to 2.5x higher speedup than the state-of-the-art speculative decoding method EAGLE-3.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06036.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Jian Chen, Yesheng Liang, Zhijian Liu&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 27&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast draft model whose outputs are verified in parallel by the target LLM; however, existing methods still rely on autoregressive drafting, which remains sequential and limits practical speedups. Diffusion LLMs offer a promising alternative by enabling parallel generation, but current diffusion models typically underperform compared with autoregressive models. In this paper, we introduce DFlash, a speculative decoding framework that employs a lightweight block diffusion model for parallel drafting. By generating draft tokens in a single forward pass and conditioning the draft model on context features extracted from the target model, DFlash enables efficient drafting with high-quality outputs and higher acceptance rates. Experiments show that DFlash achieves over 6x lossless acceleration across a range of models and tasks, delivering up to 2.5x higher speedup than the state-of-the-art speculative decoding method EAGLE-3.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.06036</guid>
      <pubDate>Thu, 05 Feb 2026 18:59:30 +0000</pubDate>
    </item>
    <item>
      <title>SwimBird: Eliciting Switchable Reasoning Mode in Hybrid Autoregressive MLLMs</title>
      <link>https://arxiv.org/abs/2602.06040</link>
-      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06040.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Jintao Tong, Shilin Yan, Hongwei Xue, Xiaojun Tang, Kunyu Shi, Guannan Zhang, Ruixuan Li, Yixiong Zou&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 9&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Multimodal Large Language Models (MLLMs) have made remarkable progress in multimodal perception and reasoning by bridging vision and language. However, most existing MLLMs perform reasoning primarily with textual CoT, which limits their effectiveness on vision-intensive tasks. Recent approaches inject a fixed number of continuous hidden states as "visual thoughts" into the reasoning process and improve visual performance, but often at the cost of degraded text-based logical reasoning. We argue that the core limitation lies in a rigid, pre-defined reasoning pattern that cannot adaptively choose the most suitable thinking modality for different user queries. We introduce SwimBird, a reasoning-switchable MLLM that dynamically switches among three reasoning modes conditioned on the input: (1) text-only reasoning, (2) vision-only reasoning (continuous hidden states as visual thoughts), and (3) interleaved vision-text reasoning. To enable this capability, we adopt a hybrid autoregressive formulation that unifies next-token prediction for textual thoughts with next-embedding prediction for visual thoughts, and design a systematic reasoning-mode curation strategy to construct SwimBird-SFT-92K, a diverse supervised fine-tuning dataset covering all three reasoning patterns. By enabling flexible, query-adaptive mode selection, SwimBird preserves strong textual logic while substantially improving performance on vision-dense tasks. Experiments across diverse benchmarks covering textual reasoning and challenging visual understanding demonstrate that SwimBird achieves state-of-the-art results and robust gains over prior fixed-pattern multimodal reasoning methods.&lt;/p&gt;</description>
+      <description>&lt;p&gt;&lt;img src="https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2602.06040.png" alt="Paper thumbnail" style="max-width: 300px; height: auto;" /&gt;&lt;/p&gt;&lt;p&gt;&lt;b&gt;Authors:&lt;/b&gt; Jintao Tong, Shilin Yan, Hongwei Xue, Xiaojun Tang, Kunyu Shi, Guannan Zhang, Ruixuan Li, Yixiong Zou&lt;/p&gt;&lt;p&gt;&lt;b&gt;Upvotes:&lt;/b&gt; 10&lt;/p&gt;&lt;p&gt;&lt;b&gt;Summary:&lt;/b&gt; Multimodal Large Language Models (MLLMs) have made remarkable progress in multimodal perception and reasoning by bridging vision and language. However, most existing MLLMs perform reasoning primarily with textual CoT, which limits their effectiveness on vision-intensive tasks. Recent approaches inject a fixed number of continuous hidden states as "visual thoughts" into the reasoning process and improve visual performance, but often at the cost of degraded text-based logical reasoning. We argue that the core limitation lies in a rigid, pre-defined reasoning pattern that cannot adaptively choose the most suitable thinking modality for different user queries. We introduce SwimBird, a reasoning-switchable MLLM that dynamically switches among three reasoning modes conditioned on the input: (1) text-only reasoning, (2) vision-only reasoning (continuous hidden states as visual thoughts), and (3) interleaved vision-text reasoning. To enable this capability, we adopt a hybrid autoregressive formulation that unifies next-token prediction for textual thoughts with next-embedding prediction for visual thoughts, and design a systematic reasoning-mode curation strategy to construct SwimBird-SFT-92K, a diverse supervised fine-tuning dataset covering all three reasoning patterns. By enabling flexible, query-adaptive mode selection, SwimBird preserves strong textual logic while substantially improving performance on vision-dense tasks. Experiments across diverse benchmarks covering textual reasoning and challenging visual understanding demonstrate that SwimBird achieves state-of-the-art results and robust gains over prior fixed-pattern multimodal reasoning methods.&lt;/p&gt;</description>
      <guid isPermaLink="false">https://arxiv.org/abs/2602.06040</guid>
      <pubDate>Thu, 05 Feb 2026 18:59:51 +0000</pubDate>
    </item>