# Learning Agile Locomotion via Adversarial Training

arXiv: https://arxiv.org/abs/2008.00603 · HTML: https://mlsanity.com/paper/2008.00603/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Aug 2020.

## Built on by (most cited first, 6 of 6)

- [Learning and Adapting Agile Locomotion Skills by Transferring Experience](https://mlsanity.com/paper/2304.09834/) (2023)
- [Deep Reinforcement Learning Versus Evolution Strategies: A Comparative Survey](https://mlsanity.com/paper/2110.01411/) (2021)
- [Rethinking Robustness Assessment: Adversarial Attacks on Learning-based Quadrupedal Locomotion Controllers](https://mlsanity.com/paper/2405.12424/) (2024)
- [Learning Vision-based Pursuit-Evasion Robot Policies](https://mlsanity.com/paper/2308.16185/) (2023)
- [GraspARL: Dynamic Grasping via Adversarial Reinforcement Learning](https://mlsanity.com/paper/2203.02119/) (2022)
- [Digital Red Queen: Adversarial Program Evolution in Core War with LLMs](https://mlsanity.com/paper/2601.03335/) (2026)

## Abstract (arXiv)

Developing controllers for agile locomotion is a long-standing challenge for legged robots. Reinforcement learning (RL) and Evolution Strategy (ES) hold the promise of automating the design process of such controllers. However, dedicated and careful human effort is required to design training environments to promote agility. In this paper, we present a multi-agent learning system, in which a quadruped robot (protagonist) learns to chase another robot (adversary) while the latter learns to escape. We find that this adversarial training process not only encourages agile behaviors but also effectively alleviates the laborious environment design effort. In contrast to prior works that used only one adversary, we find that training an ensemble of adversaries, each of which specializes in a different escaping strategy, is essential for the protagonist to master agility. Through extensive experiments, we show that the locomotion controller learned with adversarial training significantly outperforms carefully designed baselines.