Causal discovery from data assisted by large language models

Journal Article · Wed Sep 24 00:00:00 EDT 2025 · Applied Physics Letters

DOI:https://doi.org/10.1063/5.0272287· OSTI ID:2997644

Barakati, Kamyar ^[1]; Molak, Aleksander ^[2]; Nelson, Christopher T. ^[3]; Zhang, Xiaohang ^[4]; Takeuchi, Ichiro ^[4]; Kalinin, Sergei V. ^[1]

---

1. University of Tennessee, Knoxville, TN (United States)
2. CausalPython.io, Warsaw (Poland)
3. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
4. University of Maryland, College Park, MD (United States)

Knowledge-driven discovery of novel materials necessitates the development of causal models for property emergence. While in the classical physical paradigm, the causal relationships are deduced based on physical principles or via experiment, the rapid accumulation of observational data necessitates learning causal relationships between dissimilar aspects of material structure and functionalities based on observations. For this, it is essential to integrate experimental data with prior domain knowledge. Here, we demonstrate this approach by combining high-resolution scanning transmission electron microscopy data with insights derived from large language models (LLMs). By applying ChatGPT to domain-specific literature, such as arXiv papers on ferroelectrics, and combining the obtained information with data-driven causal discovery, we construct adjacency matrices for directed acyclic graphs that map the causal relationships between structural, chemical, and polarization degrees of freedom in Sm-doped BiFeO₃. This approach enables us to hypothesize how synthesis conditions influence material properties and guides experimental validation. Furthermore, the ultimate objective of this work is to develop a unified framework that integrates LLM-driven literature analysis with data-driven discovery, facilitating the precise engineering of ferroelectric materials by establishing clear connections between synthesis conditions and their resulting material properties.

• Research Organization: The Pennsylvania State University, University Park, PA (United States)
• Sponsoring Organization: National Institute of Standards and Technology; USDOE Office of Science (SC), Basic Energy Sciences (BES)
• Grant/Contract Number: SC0021118
• OSTI ID: 2997644
• Journal Information: Applied Physics Letters, Journal Name: Applied Physics Letters Journal Issue: 12 Vol. 127; ISSN 1077-3118; ISSN 0003-6951
• Publisher: AIP Publishing
• Country of Publication: United States
• Language: English

Related Subjects

Causal inference
Ferroelectric materials
Machine learning
Materials analysis
Natural language processing
Polarization
Transmission electron microscopy