Quality-Weighted Vendi Scores And Their Application To Diverse Experimental Design

TL;DR

This paper introduces quality-weighted Vendi scores to improve diversity in experimental design, enabling more effective exploration and discovery across scientific applications.

Contribution

It extends Vendi scores to incorporate quality, providing a novel metric that balances diversity and quality in experimental design tasks.

Findings

Achieved 70%-170% more effective discoveries than baselines.

Enabled flexible policies for balancing quality and diversity.

Applied successfully to drug, materials discovery, and reinforcement learning.

Abstract

Experimental design techniques such as active search and Bayesian optimization are widely used in the natural sciences for data collection and discovery. However, existing techniques tend to favor exploitation over exploration of the search space, which causes them to get stuck in local optima. This ``collapse" problem prevents experimental design algorithms from yielding diverse high-quality data. In this paper, we extend the Vendi scores -- a family of interpretable similarity-based diversity metrics -- to account for quality. We then leverage these quality-weighted Vendi scores to tackle experimental design problems across various applications, including drug discovery, materials discovery, and reinforcement learning. We found that quality-weighted Vendi scores allow us to construct policies for experimental design that flexibly balance quality and diversity, and ultimately assemble rich and diverse sets of high-performing data points. Our algorithms led to a 70%-170% increase in the number of effective discoveries compared to baselines.