Campero, A; Raileanu, R; Küttler, H; Tenenbaum, JB; Rocktäschel, T; Grefenstette, E; (2021) Learning With AMIGo: Adversarially Motivated Intrinsic Goals. In: Proceedings of the 9th International Conference on Learning Representations (ICLR 2021). International Conference on Learning Representations: Virtual.

Preview

Text 2006.12122.pdf - Other Download (3MB) | Preview

Abstract

A key challenge for reinforcement learning (RL) consists of learning in environments with sparse extrinsic rewards. In contrast to current RL methods, humans are able to learn new skills with little or no reward by using various forms of intrinsic motivation. We propose AMIGO, a novel agent incorporating-as form of meta-learning-a goal-generating teacher that proposes Adversarially Motivated Intrinsic GOals to train a goal-conditioned “student” policy in the absence of (or alongside) environment reward. Specifically, through a simple but effective “constructively adversarial” objective, the teacher learns to propose increasingly challenging-yet achievable-goals that allow the student to learn general skills for acting in a new environment, independent of the task to be solved. We show that our method generates a natural curriculum of self-proposed goals which ultimately allows the agent to solve challenging procedurally-generated tasks where other forms of intrinsic motivation and state-of-the-art RL methods fail.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item