CVT Archives and Chemical Embedding Measures for Multi-Objective Quality Diversity in Molecular Design

TL;DR

This paper introduces a novel archive method using chemical embeddings and CVT for multi-objective molecular design, improving diversity and quality over traditional grid-based approaches.

Contribution

It applies CVT archives with learned chemical embeddings to enhance diversity and efficiency in multi-objective molecular optimization.

Findings

Embedding-based CVT archives outperform grid-based archives in hypervolume.

The method achieves higher multi-objective quality diversity scores.

Nearly all native archive niches are filled with the proposed approach.

Abstract

Nonlinear optical (NLO) materials are essential for photonic technologies, yet discovering optimal NLO molecules requires balancing multiple competing objectives across vast chemical spaces. Previous work showed that Multi-Objective MAP-Elites (MOME) with grid-based archives discovers diverse, high-quality molecules for electro-optic applications. However, uniform grid partitioning wastes archive capacity on chemically infeasible regions while undersampling high-density areas. We apply MOME with Centroidal Voronoi Tessellation (CVT) archives whose cells are defined by learned embeddings from ChemBERTa-2 Multi-Task Regression reduced via UMAP, capturing chemical similarity beyond simple structural features. We investigate a four-objective NLO molecular design problem: maximizing the $\beta / \gamma$ hyperpolarizability ratio, constraining HOMO-LUMO gap and linear polarizability to target ranges, and minimizing energy per atom. Our results demonstrate that embedding-based measures in CVT archives yield significantly higher median global hypervolume and multi-objective quality diversity scores, while filling nearly all native archive niches.