Zero-Shot Reinforcement Learning from Low Quality Data

TL;DR

This paper investigates how zero-shot reinforcement learning performs with small, homogeneous datasets and introduces conservative methods that improve performance in low-quality data scenarios, outperforming existing approaches.

Contribution

The paper proposes conservative modifications to zero-shot RL algorithms that enhance performance on small, homogeneous datasets, a setting less explored in prior work.

Findings

Conservative zero-shot RL outperforms non-conservative methods on low-quality datasets.

Proposed methods match or exceed performance of baselines with task exposure.

Conservative approaches do not degrade performance on high-quality datasets.

Abstract

Zero-shot reinforcement learning (RL) promises to provide agents that can perform any task in an environment after an offline, reward-free pre-training phase. Methods leveraging successor measures and successor features have shown strong performance in this setting, but require access to large heterogenous datasets for pre-training which cannot be expected for most real problems. Here, we explore how the performance of zero-shot RL methods degrades when trained on small homogeneous datasets, and propose fixes inspired by conservatism, a well-established feature of performant single-task offline RL algorithms. We evaluate our proposals across various datasets, domains and tasks, and show that conservative zero-shot RL algorithms outperform their non-conservative counterparts on low quality datasets, and perform no worse on high quality datasets. Somewhat surprisingly, our proposals also outperform baselines that get to see the task during training. Our code is available via https://enjeeneer.io/projects/zero-shot-rl/ .