Research community previously filed under “Materials Chemistry”

Machine Learning Potentials and Large Language Models for Chemical Discovery

Mostly in the field High-Pressure Physics and Machine-Learned Chemical Discovery (16% of its papers), one of 2 communities there

Papers 6,489
Elements 2
Keywords & sectors 1
Leading venue arXiv (Cornell University)

Machine Learning Potentials and Large Language Models for Chemical Discovery

This community develops computational tools that use machine learning to predict atomic interactions, simulate chemical reactions, and automate the design of new materials.

The work centers on creating interatomic potentials and force fields that allow for accurate molecular dynamics simulations of complex systems. A significant portion of the research focuses on training neural networks, including graph neural networks and deep learning models, to predict crystal structures and chemical properties. Concurrently, the community applies large language models to extract data from scientific literature, generate research hypotheses, and facilitate autonomous laboratory workflows. These methods are applied to specific challenges such as modeling amorphous silicon, transition metal catalysis, and the discovery of inorganic materials. The integration of active learning and Bayesian optimization enables the efficient exploration of chemical space, reducing the computational cost of traditional density functional theory calculations.

The community's output is most heavily concentrated in silicon research, accounting for 20% of all silicon studies tracked, and germanium research, where it represents 20% of that element's literature. Silicon also contributes the highest absolute number of papers to this group, with 820 entries, followed by carbon with 523 papers.

The community comprises 6,489 papers, with the highest publication volumes appearing in arXiv, npj Computational Materials, and ChemRxiv.

Recent work focuses on developing universal machine learning interatomic potentials for large-scale atomistic simulations and using foundation models to predict thermal conductivity. Other recent studies explore the use of knowledge graphs and large language models to generate scientific ideas and evaluate AI's ability to perform expert-level scientific tasks.

Recurring themes in the literature

Phrases that recur across this community's paper titles -- a quick map of its sub-topics, drawn straight from the titles themselves.

  • machine learning 799
  • molecular dynamics 226
  • neural network 178
  • interatomic potentials 131
  • deep learning 110
  • neural networks 98
  • artificial intelligence 91
  • graph neural 85
  • learning interatomic 83
  • active learning 82
  • density functional 81
  • large language 75
  • language models 72
  • force fields 66
  • functional theory 61
  • dynamics simulations 51
  • bayesian optimization 49
  • crystal structure 47
  • amorphous silicon 47
  • learning potentials 46
  • transition metal 44
  • force field 44
  • dynamics simulation 41
  • inverse design 40

Papers behind this description

Most cited
Nature Machine Intelligence · 2024 · 736 citations
Nature · 2023 · 930 citations
Nature · 2025 · 447 citations
Nature Communications · 2024 · 724 citations
Nature Machine Intelligence · 2023 · 908 citations
Newest
Nature Communications · 2026 · 71 citations
Nature Communications · 2026 · 77 citations
npj Computational Materials · 2025 · 56 citations
Nature Communications · 2025 · 49 citations
npj Computational Materials · 2026 · 35 citations

A sample from the 18 papers behind this description. Create a free account to see them all.