Benchmarking Machine Learning Potentials for Crystal Structure Relaxation

Kowen Woo, Prashant Govindarajan, Sarath Chandar

Track: Track 1: Original Research/Position/Education/Attention Track

Keywords: ML potentials, crystals, structure relaxation, benchmark

TL;DR: We benchmark ML interatomic potentials on RL-generated crystals for reliable structure relaxation.

Abstract: High-throughput materials discovery workflows require rapid and accurate relaxation of crystal structures to identify thermodynamically stable phases among thousands to millions of candidate structures. Yet current machine learning interatomic potential (MLIP) benchmarks focus predominantly on energy prediction rather than structure relaxation, creating a critical evaluation gap for models designed to accelerate optimization. Additionally, these benchmarks are trained on datasets consisting mainly of known stable or near-stable materials, thus failing to capture the challenges of unexplored chemical spaces. We address these limitations by introducing a benchmark that evaluates state-of-the-art MLIPs and a one-shot relaxation model on structure relaxation with crystals generated via a reinforcement learning pipeline. We compare energy lowering and average maximum force computed via DFT, as well as relaxation runtime. We also contrast direct force-prediction strategies against conservative energy-differentiation approaches to determine which paradigm delivers superior relaxation performance. Our results indicate that there is a clear disconnect between MLIP energy prediction and force convergence in relaxation, challenging current benchmarking approaches.

Submission Number: 442