Publication
Shape Your Body: Value Gradients for Multi-Embodiment Robot Design
Nico Bohlinger; Jan Peters
In: Computing Research Repository eprint Journal (CoRR), Vol. abs/2606.00702, Pages 1-27, arXiv, 2026.
Abstract
We propose to turn generalist multi-embodiment value functions into
reusable models for robot design. Instead of running a new reinforcement learn-
ing co-design loop for each robot, we first train an embodiment-aware policy and
value function across many robot designs. After training, the frozen value function
is used as a differentiable surrogate to optimize candidate embodiments through
value gradients. We evaluate our approach across different robot design settings,
from perturbed single robots to held-out robots across morphology classes, with
single models trained on up to 50 robots and design spaces of over 1100 con-
tinuous embodiment parameters. Beyond optimizing complete embodiments, we
show that value gradients can identify performance-limiting design and control
parameters, enabling both the optimization and the analysis of new robot designs.