← 研踪 PaperTrailarXiv:2610.35659 · v1
Environment Edits as a Corrective Signal for Reward Learning
首次提交 2026-10-27 00:04
版本更新 2026-10-27 00:04
雷达观测 2026-10-29 00:35
摘要
Robot corrections allow people to adjust a robot's behavior, but a single correction changes many features. Also, preferences about object attributes and relationships between objects cannot be expressed by moving the robot. We propose environment edits as a complementary form of feedback. The robot's motion is held constant as the human edits the scene, so the only explanation for the edit is the moved object. We introduce Feature Learning from Environmental and Robot-Relational Edit Traces (FERRET), which learns preferences from both robot and environment corrections. We show that environment edits are able to disambiguate features and learn preferences that robot corrections alone cannot.
作者备注：Accepted to CoRL 2026 Workshop on Learning from Corrections
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-10-27 00:04 · 最新观测版本
Environment Edits as a Corrective Signal for Reward Learning
Robot corrections allow people to adjust a robot's behavior, but a single correction changes many features. Also, preferences about object attributes and relationships between objects cannot be expressed by moving the robot. We propose environment edits as a complementary form of feedback. The robot's motion is held constant as the human edits the scene, so the only explanation for the edit is the moved object. We introduce Feature Learning from Environmental and Robot-Relational Edit Traces (FERRET), which learns preferences from both robot and environment corrections. We show that environment edits are able to disambiguate features and learn preferences that robot corrections alone cannot.
查看此版本解读 →查看此版本原文 ↗