外观
论文arXiv 2607.282513 个月前
基于关键性世界模型的具身 AI 策略级递归自我改进
原文链接：Policy-Level Recursive Self-Improvement for Embodied AI with a Criticality World Model（arXiv 2607.28251 · 发布日期：2026-07-30）
Policy-Level Recursive Self-Improvement for Embodied AI with a Criticality World Model
针对具身 AI 任务微调易停滞，提出以关键性世界模型引导场景收集的递归自改进，并在五个具身领域验证。
核心要点
- 均匀收集场景在后期较少遇到剩余失败，导致任务特定微调进入平台期。
- 执行结果训练关键性世界模型，用风险分数引导场景收集并训练专门策略。
- 加权重采样近似校正收集偏差，推理时风险门控在专门策略与冻结名义策略间选择。
- 在五个具身领域，相比基线，系统将运动与操作失败率降59--67%，VLA降8--25%。
验证范围与局限：验证覆盖五个具身领域，包括运动、操作和 VLA 基准；摘要未说明真机或仿真。
英文摘要 (Abstract)
Despite advances in policy pretraining, embodied AI systems can plateau during task-specific fine-tuning as uniform scenario collection encounters fewer of the remaining failures. We address this problem with a policy-level recursive self-improvement loop: execution outcomes train a criticality world model, whose risk scores guide scenario collection for specialist training. To account for the resulting shift in scenario frequencies, weighted resampling approximately corrects the collection bias. Because specialized updates can weaken nominal behavior, a risk-based gate selects between the specialist and a frozen nominal policy at inference. We extend this procedure to reinforcement learning and behavior cloning. A sampling analysis characterizes the correction, while a controlled study examines the trade-off between critical and nominal coverage. Across five embodied domains, the composed systems reduce failure rates by 59--67% for locomotion and manipulation and 8--25% for VLA benchmarks relative to their respective baselines.
作者：Linxuan He, Yuying Tian, Lingxiang Fan, Jiaqi Pi, Yinqiao Lu, Shang Su, Mengkai Shi, Shuo Feng
来源
本页事实对应的公开页面，以原文为准。