Reduce, Reuse, Recycle: Selective Reincarnation in Multi-Agent Reinforcement LearningDownload PDF

Published: 03 Mar 2023, Last Modified: 12 Apr 2023RRL 2023 OralReaders: Everyone
Keywords: reinforcement learning, multi-agent reinforcement learning, offline reinforcement learning
TL;DR: In a multi-agent RL system, selectively reincarnating only a subset of agents can either speed up or hamper training, depending on which agents are reincarnated.
Abstract: `Reincarnation' in reinforcement learning has been proposed as a formalisation of reusing prior computation from past experiments when training an agent in an environment. In this paper, we present a brief foray into the paradigm of reincarnation in the multi-agent (MA) context. We consider the case where only some agents are reincarnated, whereas the others are trained from scratch -- selective reincarnation. In the fully-cooperative MA setting with heterogeneous agents, we demonstrate that selective reincarnation can lead to higher returns than training fully from scratch, and faster convergence than training with full reincarnation. However, the choice of which agents to reincarnate in a heterogeneous system is vitally important to the outcome of the training -- in fact, a poor choice can lead to considerably worse results than the alternatives. We argue that a rich field of work exists here, and we hope that our effort catalyses further energy in bringing the topic of reincarnation to the multi-agent realm.
Track: Technical Paper
Supplementary Material: zip
Confirmation: I have read and agree with the workshop's policy on behalf of myself and my co-authors.
3 Replies