Improving Environment Novelty Quantification for Effective Unsupervised Environment Design

TL;DR

This paper introduces CENIE, a novel framework for quantifying environment novelty in Unsupervised Environment Design, enhancing curriculum effectiveness by combining novelty measurement with existing regret-based methods.

Contribution

It proposes a scalable, domain-agnostic method to measure environment novelty using state-action coverage and Gaussian Mixture Models, improving curriculum design for better generalization.

Findings

Augmenting regret-based UED with CENIE improves performance

CENIE effectively measures environment novelty across benchmarks

Combining regret and novelty enhances exploration and generalization

Abstract

Unsupervised Environment Design (UED) formalizes the problem of autocurricula through interactive training between a teacher agent and a student agent. The teacher generates new training environments with high learning potential, curating an adaptive curriculum that strengthens the student's ability to handle unseen scenarios. Existing UED methods mainly rely on regret, a metric that measures the difference between the agent's optimal and actual performance, to guide curriculum design. Regret-driven methods generate curricula that progressively increase environment complexity for the student but overlook environment novelty -- a critical element for enhancing an agent's generalizability. Measuring environment novelty is especially challenging due to the underspecified nature of environment parameters in UED, and existing approaches face significant limitations. To address this, this paper introduces the Coverage-based Evaluation of Novelty In Environment (CENIE) framework. CENIE proposes a scalable, domain-agnostic, and curriculum-aware approach to quantifying environment novelty by leveraging the student's state-action space coverage from previous curriculum experiences. We then propose an implementation of CENIE that models this coverage and measures environment novelty using Gaussian Mixture Models. By integrating both regret and novelty as complementary objectives for curriculum design, CENIE facilitates effective exploration across the state-action space while progressively increasing curriculum complexity. Empirical evaluations demonstrate that augmenting existing regret-based UED algorithms with CENIE achieves state-of-the-art performance across multiple benchmarks, underscoring the effectiveness of novelty-driven autocurricula for robust generalization.