Retrospective on the 2021 BASALT Competition on Learning from Human Feedback

TL;DR

This paper reviews the 2021 MineRL BASALT competition at NeurIPS 2021, which aimed to advance learning from human feedback techniques for open-world tasks in Minecraft, highlighting diverse approaches and their validation.

Contribution

It provides a retrospective analysis of the first MineRL BASALT competition, showcasing diverse algorithms, validation of task selection, and insights for future improvements.

Findings

Diverse LfHF algorithms achieved similar performance.

Different approaches excelled on different tasks.

The competition validated the task design and approach diversity.

Abstract

We held the first-ever MineRL Benchmark for Agents that Solve Almost-Lifelike Tasks (MineRL BASALT) Competition at the Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021). The goal of the competition was to promote research towards agents that use learning from human feedback (LfHF) techniques to solve open-world tasks. Rather than mandating the use of LfHF techniques, we described four tasks in natural language to be accomplished in the video game Minecraft, and allowed participants to use any approach they wanted to build agents that could accomplish the tasks. Teams developed a diverse range of LfHF algorithms across a variety of possible human feedback types. The three winning teams implemented significantly different approaches while achieving similar performance. Interestingly, their approaches performed well on different tasks, validating our choice of tasks to include in the competition. While the outcomes validated the design of our competition, we did not get as many participants and submissions as our sister competition, MineRL Diamond. We speculate about the causes of this problem and suggest improvements for future iterations of the competition.