Skip to content

Latest commit

 

History

History
18 lines (15 loc) · 1.04 KB

RoboSumo.md

File metadata and controls

18 lines (15 loc) · 1.04 KB

Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments

Maruan Al-Shedivat∗ CMU Trapit Bansal UMass Amherst Yuri Burda OpenAI Ilya Sutskever OpenAI Igor Mordatch OpenAI Pieter Abbeel UC Berkeley, OpenAI

Ability to continuously learn and adapt from limited experience in nonstationary environments is an important milestone on the path towards general intelligence. In this paper, we cast the problem of continuous adaptation into the learning-to-learn framework. We develop a simple gradient-based meta-learning algorithm suitable for adaptation in dynamically changing and adversarial scenarios. Additionally, we design a new multi-agent competitive environment, RoboSumo, and define iterated adaptation games for testing various aspects of continuous adaptation strategies.

We demonstrate that meta-learning enables significantly more efficient adaptation than reactive baselines in the few-shot regime. Our experiments with a population of agents that learn and compete suggest that meta-learners are the fittest.