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.

Source: arXiv:2610.35659v1 - http://arxiv.org/abs/2610.35659v1 PDF: https://arxiv.org/pdf/2610.35659v1 Original Link: http://arxiv.org/abs/2610.35659v1