AtomWorld: A Benchmark for Evaluating Spatial Reasoning in Large Language Models on Crystalline Materials

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lv, Taoyuze, Chen, Alexander, Xie, Fengyu, Wu, Chu, Meng, Jeffrey, Zhou, Dongzhan, Wang, Yingheng, Hoex, Bram, Zhong, Zhicheng, Xie, Tong
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918529269235712
author Lv, Taoyuze
Chen, Alexander
Xie, Fengyu
Wu, Chu
Meng, Jeffrey
Zhou, Dongzhan
Wang, Yingheng
Hoex, Bram
Zhong, Zhicheng
Xie, Tong
author_facet Lv, Taoyuze
Chen, Alexander
Xie, Fengyu
Wu, Chu
Meng, Jeffrey
Zhou, Dongzhan
Wang, Yingheng
Hoex, Bram
Zhong, Zhicheng
Xie, Tong
contents Large language models (LLMs) have shown promising potential in scientific research, enabling tasks ranging from knowledge retrieval to property prediction. Existing science benchmarks mainly focus on perceptual or knowledge-based tasks, largely ignoring the modelling tasks, a fundamental starting point for any real scientific research. For materials science, constructing and manipulating atomic structures is one of the most creative and least automated steps. In this work, we introduce AtomWorld, a benchmark designed to evaluate the abilities of LLMs on structure modifications. The benchmark includes ten fundamental actions under four widely used modelling categories, enabling verifiable evaluation metrics. We find that Claude Opus 4.6 generally performs the best. While the success rate decreases markedly with increasing modelling complexity, with particularly low success rates (below 12\% for rotation) for operations involving complex spatial relations. Our results suggest that contemporary LLMs are better suited as copilots for materials structure modelling rather than fully unsupervised autonomous scientific agents. Beyond evaluation, AtomWorld also serves as a testbed and playground for developing future structure-aware models, including reinforcement learning and agentic approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2510_04704
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AtomWorld: A Benchmark for Evaluating Spatial Reasoning in Large Language Models on Crystalline Materials
Lv, Taoyuze
Chen, Alexander
Xie, Fengyu
Wu, Chu
Meng, Jeffrey
Zhou, Dongzhan
Wang, Yingheng
Hoex, Bram
Zhong, Zhicheng
Xie, Tong
Materials Science
Artificial Intelligence
Computation and Language
Large language models (LLMs) have shown promising potential in scientific research, enabling tasks ranging from knowledge retrieval to property prediction. Existing science benchmarks mainly focus on perceptual or knowledge-based tasks, largely ignoring the modelling tasks, a fundamental starting point for any real scientific research. For materials science, constructing and manipulating atomic structures is one of the most creative and least automated steps. In this work, we introduce AtomWorld, a benchmark designed to evaluate the abilities of LLMs on structure modifications. The benchmark includes ten fundamental actions under four widely used modelling categories, enabling verifiable evaluation metrics. We find that Claude Opus 4.6 generally performs the best. While the success rate decreases markedly with increasing modelling complexity, with particularly low success rates (below 12\% for rotation) for operations involving complex spatial relations. Our results suggest that contemporary LLMs are better suited as copilots for materials structure modelling rather than fully unsupervised autonomous scientific agents. Beyond evaluation, AtomWorld also serves as a testbed and playground for developing future structure-aware models, including reinforcement learning and agentic approaches.
title AtomWorld: A Benchmark for Evaluating Spatial Reasoning in Large Language Models on Crystalline Materials
topic Materials Science
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2510.04704