Modular large language model agents for multi-task computational materials science
Open accessPublished version of a preprint
Abstract
The integration of large language models (LLMs) with domain-specific computational tools offers a promising pathway to streamline and enhance materials science workflows. This paper presents MatSciAgent, a multi-agent framework capable of performing key tasks such as materials data retrieval, continuum simulation, crystal structure generation, and molecular dynamics simulation. At the core of the framework is the master agent, which interprets the user’s natural language query, identifies the task type, and delegates it to a corresponding task-specific agent equipped with appropriate computational tools. Leveraging databases such as the Materials Project and MatWeb, the framework retrieves and summarizes materials data with grounded, factual responses—addressing a key limitation of vanilla LLMs. In cases where the target material is not found in existing databases, a generative task-specific agent can propose plausible crystal structures. For simulation tasks, dedicated agents extract relevant parameters from the user query to conduct continuum simulations (e.g., Cellular Automata and Monte Carlo Annealing) and atomistic simulations (e.g., Molecular Dynamics) using both established software and custom code. The modular design of these agents and their associated tools enables seamless extensibility, allowing the framework to evolve as new capabilities are integrated.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
These papers typically split the workflow into role-specific agents for planning, structure retrieval, simulation execution, analysis, and quality control.
“These papers typically split the workflow into role-specific agents for planning, structure retrieval, simulation execution, analysis, and quality control [18,19,21,25,26,35-38,41,42].”
From From Scientific Copilots to Tool-Grounded Autonomy: AI Agents in Simulation-Driven Materials Discovery (2026) · cites this paper as [41]
The Introduction and §2.1 (page not available) describe a master agent that routes queries to task-specific agents for materials retrieval, structure generation, and simulations. They do not describe separate planning, analysis, or quality-control agents; the paper’s reliability analysis evaluates the system rather than assigning quality control to an agent. (LLM, conf=0.86)
Experimental — accuracy not yet measuredOther papers were more demonstrative, emphasizing workflow feasibility, design navigation, or case-based success rather than tightly standardized metrics.
“Other papers were more demonstrative, emphasizing workflow feasibility, design navigation, or case-based success rather than tightly standardized metrics [17,18,21,22,35,37-39,41].”
From From Scientific Copilots to Tool-Grounded Autonomy: AI Agents in Simulation-Driven Materials Discovery (2026) · cites this paper as [41]
The Conclusion describes case studies demonstrating automated materials-science workflows, supporting the characterization of the paper as demonstrative. But §2.13 and Table 5 also report repeated-run reliability tests and quantitative consistency metrics, so the contrast with metrics needs qualification; the paper does not establish a comparison with other papers’ metric standardization. (LLM, conf=0.83)
Experimental — accuracy not yet measuredOther studies act more as discovery managers around retrieval, ranking, surrogate models, or protocol navigation.
“Others act more as discovery managers around retrieval, ranking, surrogate models, or protocol navigation [19,20,26,30,41].”
From From Scientific Copilots to Tool-Grounded Autonomy: AI Agents in Simulation-Driven Materials Discovery (2026) · cites this paper as [41]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 46 · Cited by 19
Research works printed but not linked: 10
- Brown, T. et al. Language models are few-shot learners. Advances in neural information processing systems 33, 1877-1901 (2020).
- Meda, R. S. & Farimani, A. B. Bapulm: Binding affinity prediction us- ing language models (2024). URL https://arxiv.org/abs/2411.04150. 2411.04150. · arXiv
- Ock, J., Vinchurkar, T., Jadhav, Y. & Farimani, A. B. Adsorb-agent: Au- tonomous identification of stable adsorption configurations via large lan- guage model agent (2024). URL https://arxiv.org/abs/2410.16658. 2410.16658. · arXiv
- Jadhav, Y. & Farimani, A. B. Large language model agent as a mechanical designer (2024). URL https://arxiv.org/abs/2404.17525. 2404.17525. · arXiv
- Liu, F. et al. Exploring and evaluating hallucinations in llm-powered code generation (2024). URL https://arxiv.org/abs/2404.00971. 2404. 00971. · arXiv
- Radford, A. & Narasimhan, K. Improving language understanding by gen- erative pre-training (2018). URL https://api.semanticscholar.org/ CorpusID:49313245.
- Vaswani, A. Attention is all you need. Advances in Neural Information Processing Systems (2017).
- Ling, C. et al. Domain specialization as the key to make large language models disruptive: A comprehensive survey (2024). URL https://arxiv. org/abs/2305.18703. 2305.18703.
- Li, H. et al. Blade: Enhancing black-box large language models with small domain-specific models (2024). URL https://arxiv.org/abs/ 2403.18365. 2403.18365.
- Lennard-Jones, J. E. On the forces between atoms and ions. Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathe- matical and Physical Character 109, 584-597 (1925).
- Authors
- Chaudhari, A., Ock, J., & Barati Farimani, A.
- Affiliations
- Carnegie Mellon University (US)
- University of Nebraska–Lincoln (US)
- Corresponding author
- Ock, Janghoon
- Venue
- Communications Materials
- Published
- 2026-03-26
- Volume
- 7
- Issue
- 1
- ISSN
- 2662-4443 (ISSN-L)
- OpenAlex
- W7140770361
- Cited by
- 19
- References
- 46
- Citations by year
- 2026: 17 · 2025: 2
- FWCI
- 22.85
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Communications Materials (publishedVersion, cc-by-nc-nd)
- ChemRxiv (acceptedVersion, cc-by-nc)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Social Sciences › Decision Sciences › Information Systems and Management › Scientific Computing and Data Management
- Keywords
- large language models, multi-agent systems, computational materials science, crystal structure generation, molecular dynamics simulation, continuum simulation, parameter extraction, Materials Project, LLM agents, workflow automation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
