外观
论文arXiv 2610.0791722 天前
PACE：通过用于执行的进度对齐上下文实现阶段一致的长时程机器人操作
原文链接：PACE: Stage-Consistent Long-Horizon Robot Manipulation via Progress-Aligned Context for Execution（arXiv 2610.07917 · 发布日期：2026-10-06）
PACE: Stage-Consistent Long-Horizon Robot Manipulation via Progress-Aligned Context for Execution
PACE针对长时程操作中的阶段混淆，通过进度对齐上下文重解读演示，并在LIBERO-Gen与Block Routing任务上验证提升。
核心要点
- 长时程操作中，相似视觉状态复现或演示与执行速度不同会导致阶段混淆。
- PACE将完整演示压缩为有序多模态提示令牌，并用训练期双边注意监督显露阶段结构。
- 执行时以episode本地fast-weight记忆因果编码动作-观测转移，调制提示交叉注意。
- 在三类任务上成功率分别由88.9%至94.0%、79.1%至83.3%、33.3%至73.3%。
验证范围与局限：验证范围为LIBERO-Gen Goal Chain、Spatial Combination及两步Block Routing任务。
英文摘要 (Abstract)
Demonstration-conditioned policies provide a natural interface for specifying robot behavior, yet long-horizon manipulation remains difficult when visually similar states recur across different stages or when demonstrations and executions proceed at different speeds. We identify the resulting failure mode as stage confusion and introduce Progress-Aligned Context for Execution (PACE), a stateful method that continually reinterprets a complete demonstration according to realized execution progress. PACE compresses the demonstration into ordered multimodal prompt tokens and uses training-only dual-edge attention supervision to expose its latent stage structure. During execution, an episode-local fast-weight memory causally encodes realized action-observation transitions and modulates prompt cross-attention, producing a progress-aligned context for a unified diffusion action expert without test-time stage labels or stage-specific policies. PACE improves success from 88.9% to 94.0% on LIBERO-Gen Goal Chain, from 79.1% to 83.3% on Spatial Combination, and from 33.3% to 73.3% on the two-step Block Routing tasks. Failure analysis further indicates that structured demonstration alignment and causal execution memory jointly mitigate stage confusion.
作者：Yenan Chen, Junjie Shi, Chen, Lu, Zhongxiang Zhou, Rong Xiong
来源
本页事实对应的公开页面，以原文为准。