# EDAR: Learning Environment-Dependent Action Representations for Robotic Manipulation

- Source: arXiv cs.RO
- Published: 2026-10-19T04:00:00.000Z
- RoboHot: https://robohot.news/en/items/b87ehppaoq46a6ygmu4die58s
- Original: https://arxiv.org/abs/2607.11427

## Summary

EDAR (Environment-Dependent Action Representation) anchors action tokens in both executable control structure and expected visual consequences. By coupling motor commands with their environment-conditioned effects, it enables the learned action space to capture interaction semantics rather than merely command-level patterns.