LLEMA: Accelerating Materials Design via LLM-Guided Evolutionary Search
Nikhil Abhyankar, Sanchit Kabra, Saaketh Desai, Chandan K. Reddy
Abstract
Materials discovery requires navigating vast chemical and structural spaces while satisfying multiple, often conflicting, objectives. We present LLM-guided Evolution for Materials design (**LLEMA**), a unified framework that couples the scientific knowledge embedded in large language models with chemistry-informed evolutionary rules and memory-based refinement. At each iteration, an LLM proposes crystallographically specified candidates under explicit property constraints; a surrogate-augmented oracle estimates physicochemical properties; and a multi-objective scorer updates success and failure memories to guide subsequent generations. Evaluated on **14 realistic tasks** spanning electronics, energy, coatings, optics, and aerospace, AlgName discovers candidates that are chemically plausible, thermodynamically stable, and property-aligned, achieving higher hit rates and stronger Pareto fronts than generative and LLM-only baselines. Ablation studies confirm the importance of rule-guided generation, memory-based refinement, and surrogate prediction. By enforcing synthesizability and multi-objective trade-offs, AlgName provides a principled approach to accelerating practical materials discovery. Project website: https://scientific-discovery.github.io/llema-project/
BibTeX
@inproceedings{
abhyankar2026accelerating,
title={Accelerating Materials Design via {LLM}-Guided Evolutionary Search},
author={Nikhil Abhyankar and Sanchit Kabra and Saaketh Desai and Chandan K. Reddy},
booktitle={The Fourteenth International Conference on Learning Representations},
year={2026},
url={https://openreview.net/forum?id=TIqzhBvCNB}
}