Active Embodiment Identification with Reinforcement Learning for Legged Robots

Nico Bohlinger; Jan Peters

Active Embodiment Identification with Reinforcement Learning for Legged Robots

Robotics 2026-05-11 v1

Authors: Nico Bohlinger , Jan Peters

Abstract

We present an active embodiment identification method for legged robots that jointly learns information-seeking behavior and explicit embodiment prediction. Using a history-augmented URMA architecture, the method infers joint-level and global embodiment parameters through interaction with the environment in simulation across different morphologies.

Keywords

embodied agent imitation learning robot locomotion

Cite

@article{arxiv.2605.08020,
  title  = {Active Embodiment Identification with Reinforcement Learning for Legged Robots},
  author = {Nico Bohlinger and Jan Peters},
  journal= {arXiv preprint arXiv:2605.08020},
  year   = {2026}
}

Related papers

View all related →