Renmin University of China

Vision-language-action models usually predict a whole future action chunk in one pass, and adjacent chunks lack explicit temporal consistency constraints, so actions can jump at chunk boundaries. ODEAction recasts generation as continuous evolution from the current action: it learns an action transition field and integrates it with an ordinary differential equation to produce subsequent actions, instead of jointly predicting a fixed-length chunk. Average success rate rises from 96.85% to 98.25% on LIBERO and from 15.28% to 24.72% on EBench, a 61.8% relative gain; on five RoboTwin tasks it reaches 84.4% average success against the 82.4% baseline.

Vision-language-action models usually predict a whole future action chunk in one pass, and adjacent chunks lack explicit temporal consistency constraints, so actions can jump at chunk boundaries. ODEAction recasts generation as continuous evolution from the current action: it learns an action transition field and integrates it with an ordinary differential equation to produce subsequent actions, instead of jointly predicting a fixed-length chunk. Average success rate rises from 96.85% to 98.25% on LIBERO and from 15.28% to 24.72% on EBench, a 61.8% relative gain; on five RoboTwin tasks it reaches 84.4% average success against the 82.4% baseline.

Long-horizon LLM agents interact with their environment through complete turns but generate each action token by token, and existing reinforcement learning either gives a whole trajectory one shared advantage or treats every token as an action. MPO splits credit assignment into two scales: turn-level generalized advantage estimation carries credit across interactions, a mean-preserving local correction refines it within each turn, and the policy update uses complete-turn importance ratios. Across search, tool-use and automation tasks, MPO reports the highest score on all nine metrics at both Qwen3-4B-Thinking and Qwen3-30B-A3B-Instruct scales, including a 4.33-point BrowseComp-Plus pass@1 gain over the strongest non-MPO baseline at 4B.

Long-horizon LLM agents interact with their environment through complete turns but generate each action token by token, and existing reinforcement learning either gives a whole trajectory one shared advantage or treats every token as an action. MPO splits credit assignment into two scales: turn-level generalized advantage estimation carries credit across interactions, a mean-preserving local correction refines it within each turn, and the policy update uses complete-turn importance ratios. Across search, tool-use and automation tasks, MPO reports the highest score on all nine metrics at both Qwen3-4B-Thinking and Qwen3-30B-A3B-Instruct scales, including a 4.33-point BrowseComp-Plus pass@1 gain over the strongest non-MPO baseline at 4B.

Agentic reinforcement learning shifts the reasoning trajectories of large language models, and trajectory-based confidence estimators trained on a base model lose much of their ranking ability after that shift. StaR-TTA treats trajectory stability as a label-free reliability signal: it learns the link between stability and correctness on labeled base trajectories, then uses stability rankings from unlabeled RL trajectories to correct confidence at test time. Across four QA benchmarks it reaches an average AUROC of 0.7208, with gains on unseen datasets, different model families and scales, and both PPO and GRPO.

Agentic reinforcement learning shifts the reasoning trajectories of large language models, and trajectory-based confidence estimators trained on a base model lose much of their ranking ability after that shift. StaR-TTA treats trajectory stability as a label-free reliability signal: it learns the link between stability and correctness on labeled base trajectories, then uses stability rankings from unlabeled RL trajectories to correct confidence at test time. Across four QA benchmarks it reaches an average AUROC of 0.7208, with gains on unseen datasets, different model families and scales, and both PPO and GRPO.

Interactive video world models draw frames from a memory bank stamped with capture times, and a frozen generator reads the newest timestamp as truth, so after an agent undoes an action the only view of the present is an older frame judged outdated. ReDate keeps every frame and re-dates retrieved tokens by valid time read from the action log, so a reverted region reads as one that has just reached a new state, with no training. On FAR's 352 reverted steps the current-state share rises from 0.36 to 0.62 and the outdated share falls from 0.27 to 0.09; on HY-WorldPlay it rises from 0.003 to 0.19.

Interactive video world models draw frames from a memory bank stamped with capture times, and a frozen generator reads the newest timestamp as truth, so after an agent undoes an action the only view of the present is an older frame judged outdated. ReDate keeps every frame and re-dates retrieved tokens by valid time read from the action log, so a reverted region reads as one that has just reached a new state, with no training. On FAR's 352 reverted steps the current-state share rises from 0.36 to 0.62 and the outdated share falls from 0.27 to 0.09; on HY-WorldPlay it rises from 0.003 to 0.19.

When two humanoid robots carry one freely supported rigid payload through constrained corridors, team coordination is written as executable code rather than fixed by task decomposition, with a multimodal foundation model supplying motion primitives, semantic objectives, temporary roles and bounded parameters. A verifier judges each physical transition; rejected ones trigger structured revision from the latest accepted checkpoint, and verified experience is retained for reuse. Across 15 base-task runs with three foundation-model backbones, 8 tasks completed, with every backbone achieving end-to-end success.

When two humanoid robots carry one freely supported rigid payload through constrained corridors, team coordination is written as executable code rather than fixed by task decomposition, with a multimodal foundation model supplying motion primitives, semantic objectives, temporary roles and bounded parameters. A verifier judges each physical transition; rejected ones trigger structured revision from the latest accepted checkpoint, and verified experience is retained for reuse. Across 15 base-task runs with three foundation-model backbones, 8 tasks completed, with every backbone achieving end-to-end success.

Across 160 classification and 81 regression datasets, tabular foundation model rankings change with the benchmark and the reliability metric, and regression models that already sit near nominal marginal coverage still under-cover in their worst feature slab, so model selection cannot rest on aggregate predictive accuracy alone.

Across 160 classification and 81 regression datasets, tabular foundation model rankings change with the benchmark and the reliability metric, and regression models that already sit near nominal marginal coverage still under-cover in their worst feature slab, so model selection cannot rest on aggregate predictive accuracy alone.

Whether one policy difference transfers depends on the target: Qwen3 targets gain positively while Qwen2.5 targets under a Qwen3 reference gain almost nothing, and swapping in a Qwen2.5-3B reference turns those gains positive.

Whether one policy difference transfers depends on the target: Qwen3 targets gain positively while Qwen2.5 targets under a Qwen3 reference gain almost nothing, and swapping in a Qwen2.5-3B reference turns those gains positive.

POrchestra lets sub-agents proactively send structured requests to the orchestrator during execution, so the orchestrator can revise their subgoal, context, tool set, or execution budget on the spot instead of waiting for the sub-agent to finish or fail. Earlier re-orchestration methods fixed each sub-agent's configuration until its lifecycle ended, delaying adaptation to dynamic environments. Across GAIA, GAIA2, and SWE-Bench Verified, task success improves by an average of 9.1% over the strongest baseline, and the companion OPOD turns that feedback into process supervision, raising small-orchestrator success by an average of 10.0% over training baselines.

POrchestra lets sub-agents proactively send structured requests to the orchestrator during execution, so the orchestrator can revise their subgoal, context, tool set, or execution budget on the spot instead of waiting for the sub-agent to finish or fail. Earlier re-orchestration methods fixed each sub-agent's configuration until its lifecycle ended, delaying adaptation to dynamic environments. Across GAIA, GAIA2, and SWE-Bench Verified, task success improves by an average of 9.1% over the strongest baseline, and the companion OPOD turns that feedback into process supervision, raising small-orchestrator success by an average of 10.0% over training baselines.

The authors introduce CATCH, a controllable testbed for reward hacking in coding RL that deliberately exposes environmental loopholes and provides execution-based gold labels by comparing success under a vulnerable evaluator with task correctness under an independent audit, while controlling the model's initial hacking tendency through supervised fine-tuning data mixtures and the difficulty of earning rewards through reward designing; experiments show CATCH produces diverse RL training trajectories with clear reward hacking, that initial models and reward difficulties jointly shape its emergence, and that a chain-of-thought monitor initially suppresses hacking but this protection erodes as the policy model learns to mislead the monitor with code comments.

The authors introduce CATCH, a controllable testbed for reward hacking in coding RL that deliberately exposes environmental loopholes and provides execution-based gold labels by comparing success under a vulnerable evaluator with task correctness under an independent audit, while controlling the model's initial hacking tendency through supervised fine-tuning data mixtures and the difficulty of earning rewards through reward designing; experiments show CATCH produces diverse RL training trajectories with clear reward hacking, that initial models and reward difficulties jointly shape its emergence, and that a chain-of-thought monitor initially suppresses hacking but this protection erodes as the policy model learns to mislead the monitor with code comments.

The work presents Form and Void Agent (FaV-A), a multimodal agent for staged positive-negative space generation: it first generates a base object, then analyzes its shape and spatial structure to identify candidate negative-space semantics, and finally produces compositional instructions for the final image generation stage; experiments and ablation analyses suggest FaV-A yields more visually coherent and semantically aligned positive-negative space compositions than direct zero-shot MLLM baselines.

The work presents Form and Void Agent (FaV-A), a multimodal agent for staged positive-negative space generation: it first generates a base object, then analyzes its shape and spatial structure to identify candidate negative-space semantics, and finally produces compositional instructions for the final image generation stage; experiments and ablation analyses suggest FaV-A yields more visually coherent and semantically aligned positive-negative space compositions than direct zero-shot MLLM baselines.