arXiv 2610.22448 · cs.LG · 2026-10-15
Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards
Ruida Wang, Gerald Friedland, Tong Zhang, Chris Kong
Amazon
已归档 · 2026-10 本站收录于 2026-10-24
一句话要点
Amazon 的 Lean4AgenticRL 给软件工程和终端任务的智能体强化学习加上多层过程奖励：把工具调用级、工作流阶段级和最终结果奖励统一建立在 Lean4 形式化之上，可验证，弥补只用最终结果奖励训练效率低的问题。
Amazon's Lean4AgenticRL adds verifiable multi-level process rewards to agentic RL for software-engineering and terminal tasks, grounding tool-call, workflow-stage and outcome rewards in a unified Lean4 formalization instead of relying on outcome rewards alone.
阅读 arXiv 原文 PDF Hugging Face 讨论 代码仓库 项目页
主题与学科
同主题：强化学习
- Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation ▲214
- Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight ▲189
- LoGRA: Scaling LLM Reinforcement Learning with Low-Rank Gradient Sketches ▲152
- Sharpening Tax in Post-Training ▲111
- RULER: Instance-aware Rubric Rewards for SVG Generation ▲104
同主题：智能体
- Raven: The Harness of Harnesses for Composable Agentic Intelligence 2026-09-27 · ▲748
- False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents 2026-09-30 · ▲642
- Omni-IO Skills: Harnessing Your Agent Omni-Native 2026-09-25 · ▲485
- BizAgentBench: Unveiling Hidden Challenges in Deploying Long-Horizon Agents for Real-World Business 2026-10-14 · ▲388
- Any Time, Any Space: A World-State-Centric Framework for Agentic Video Generation 2026-10-14 · ▲364
同机构：Amazon
- ALoDLM: Adaptively Looped Diffusion Language Models 2026-10-03
- Beyond the Timeline: Augmenting Long-Video Memory with Grounded Entity Biographies 2026-09-29
- Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence 2026-09-28
- LEGO-Anything: Coding Agents for 3D Scene Reconstruction 2026-09-28
数据说明
元数据来自 arXiv API（已核实）；点赞来自 Hugging Face 每日论文；引用来自 Semantic Scholar。最近刷新 2026-10-25。
要点由本站根据摘要自写，不转载摘要原文，请阅读原文。