Results of the NeurIPS 2023 Neural MMO Competition on Multi-task Reinforcement Learning
This competition addresses the challenge of generalization in multi-task reinforcement learning for AI researchers, though it is incremental as it builds on existing benchmarks and methods.
The paper presents results from the NeurIPS 2023 Neural MMO Competition, where participants developed goal-conditional policies for multi-task reinforcement learning that generalize to unseen tasks, maps, and opponents, with the top solution achieving a score 4x higher than the baseline in 8 hours on a single GPU.
We present the results of the NeurIPS 2023 Neural MMO Competition, which attracted over 200 participants and submissions. Participants trained goal-conditional policies that generalize to tasks, maps, and opponents never seen during training. The top solution achieved a score 4x higher than our baseline within 8 hours of training on a single 4090 GPU. We open-source everything relating to Neural MMO and the competition under the MIT license, including the policy weights and training code for our baseline and for the top submissions.