# Multi-embodiment Legged Robot Control as a Sequence Modeling Problem

arXiv: https://arxiv.org/abs/2212.09078 · HTML: https://mlsanity.com/paper/2212.09078/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Dec 2022.

## Built on by (most cited first, 11 of 11)

- [Cross-Embodiment Dexterous Grasping with Reinforcement Learning](https://mlsanity.com/paper/2410.02479/) (2024)
- [A Comprehensive Survey of Cross-Domain Policy Transfer for Embodied Agents](https://mlsanity.com/paper/2402.04580/) (2024)
- [OXE-AugE: A Large-Scale Robot Augmentation of OXE for Scaling Cross-Embodiment Policy Learning](https://mlsanity.com/paper/2512.13100/) (2025)
- [PEAC: Unsupervised Pre-training for Cross-Embodiment Reinforcement Learning](https://mlsanity.com/paper/2405.14073/) (2024)
- [Multi-Agent Reinforcement Learning: Methods, Applications, Visionary Prospects, and Challenges](https://mlsanity.com/paper/2305.10091/) (2023)
- [Beyond Topology: A Morphological Symmetry Graph Representation for Locomotion Policy Learning](https://mlsanity.com/paper/2512.00727/) (2025)
- [Fourier Controller Networks for Real-Time Decision-Making in Embodied Learning](https://mlsanity.com/paper/2405.19885/) (2024)
- [Adaptive Control Strategy for Quadruped Robots in Actuator Degradation Scenarios](https://mlsanity.com/paper/2312.17606/) (2023)
- [Shape Your Body: Value Gradients for Multi-Embodiment Robot Design](https://mlsanity.com/paper/2606.00702/) (2026)
- [MetaLoco: Universal Quadrupedal Locomotion with Meta-Reinforcement Learning and Motion Imitation](https://mlsanity.com/paper/2407.17502/) (2024)
- [RobotDesignGPT: Automated Robot Design Synthesis using Vision Language Models](https://mlsanity.com/paper/2601.11801/) (2026)

## Abstract (arXiv)

Robots are traditionally bounded by a fixed embodiment during their operational lifetime, which limits their ability to adapt to their surroundings. Co-optimizing control and morphology of a robot, however, is often inefficient due to the complex interplay between the controller and morphology. In this paper, we propose a learning-based control method that can inherently take morphology into consideration such that once the control policy is trained in the simulator, it can be easily deployed to robots with different embodiments in the real world. In particular, we present the Embodiment-aware Transformer (EAT), an architecture that casts this control problem as conditional sequence modeling. EAT outputs the optimal actions by leveraging a causally masked Transformer. By conditioning an autoregressive model on the desired robot embodiment, past states, and actions, our EAT model can generate future actions that best fit the current robot embodiment. Experimental results show that EAT can outperform all other alternatives in embodiment-varying tasks, and succeed in an example of real-world evolution tasks: stepping down a stair through updating the morphology alone. We hope that EAT will inspire a new push toward real-world evolution across many domains, where algorithms like EAT can blaze a trail by bridging the field of evolutionary robotics and big data sequence modeling.