{"ID":2849931,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.22503","arxiv_id":"2510.22503","title":"LLEMA: Evolutionary Search with LLMs for Multi-Objective Materials Discovery","abstract":"Materials discovery requires navigating vast chemical and structural spaces while satisfying multiple, often conflicting, objectives. We present LLM-guided Evolution for MAterials discovery (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/failure memories to guide subsequent generations. Evaluated on 14 realistic tasks that span electronics, energy, coatings, optics, and aerospace, LLEMA discovers candidates that are chemically plausible, thermodynamically stable, and property-aligned, achieving higher hit rates and improved Pareto front quality relative to 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, LLEMA provides a principled approach to accelerating practical materials discovery. Project website: https://scientific-discovery.github.io/llema-project/","short_abstract":"Materials discovery requires navigating vast chemical and structural spaces while satisfying multiple, often conflicting, objectives. We present LLM-guided Evolution for MAterials discovery (LLEMA), a unified framework that couples the scientific knowledge embedded in large language models with chemistry-informed evolu...","url_abs":"https://arxiv.org/abs/2510.22503","url_pdf":"https://arxiv.org/pdf/2510.22503v2","authors":"[\"Nikhil Abhyankar\",\"Sanchit Kabra\",\"Saaketh Desai\",\"Chandan K. Reddy\"]","published":"2025-10-26T02:47:15Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cond-mat.mtrl-sci\",\"cs.AI\",\"cs.NE\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}