UNIQ: Offline Inverse Q-learning for Avoiding Undesirable Demonstrations

TL;DR

This paper introduces UNIQ, a novel offline inverse Q-learning method designed to learn policies that avoid undesirable behaviors by maximizing the distance from undesirable demonstrations, effectively leveraging unlabeled data.

Contribution

The paper proposes a new inverse Q-learning framework for avoiding undesirable demonstrations, with a novel training objective and algorithm that outperform existing methods.

Findings

Outperforms state-of-the-art baselines on benchmark environments

Effectively leverages unlabeled data for training

Provides a new approach to avoid undesirable behaviors in offline learning

Abstract

We address the problem of offline learning a policy that avoids undesirable demonstrations. Unlike conventional offline imitation learning approaches that aim to imitate expert or near-optimal demonstrations, our setting involves avoiding undesirable behavior (specified using undesirable demonstrations). To tackle this problem, unlike standard imitation learning where the aim is to minimize the distance between learning policy and expert demonstrations, we formulate the learning task as maximizing a statistical distance, in the space of state-action stationary distributions, between the learning policy and the undesirable policy. This significantly different approach results in a novel training objective that necessitates a new algorithm to address it. Our algorithm, UNIQ, tackles these challenges by building on the inverse Q-learning framework, framing the learning problem as a cooperative (non-adversarial) task. We then demonstrate how to efficiently leverage unlabeled data for practical training. Our method is evaluated on standard benchmark environments, where it consistently outperforms state-of-the-art baselines. The code implementation can be accessed at: https://github.com/hmhuy0/UNIQ.