![[2602.18008] NIMMGen: Learning Neural-Integrated Mechanistic Digital Twins with LLMs](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.18008] NIMMGen: Learning Neural-Integrated Mechanistic Digital Twins with LLMs
Summary
The paper introduces NIMMGen, a framework for learning neural-integrated mechanistic models using large language models (LLMs), addressing challenges in model reliability and effectiveness.
Why It Matters
This research is significant as it tackles the limitations of existing LLM-based mechanistic modeling approaches, providing a more realistic evaluation framework and enhancing the practical applicability of these models in scientific and policy contexts.
Key Takeaways
- NIMMGen improves the reliability of LLM-generated mechanistic models.
- The framework evaluates models under realistic conditions, addressing oversimplifications in previous studies.
- Experiments demonstrate NIMMGen's strong performance across diverse scientific datasets.
Computer Science > Machine Learning arXiv:2602.18008 (cs) [Submitted on 20 Feb 2026] Title:NIMMGen: Learning Neural-Integrated Mechanistic Digital Twins with LLMs Authors:Zihan Guan, Rituparna Datta, Mengxuan Hu, Shunshun Liu, Aiying Zhang, Prasanna Balachandran, Sheng Li, Anil Vullikanti View a PDF of the paper titled NIMMGen: Learning Neural-Integrated Mechanistic Digital Twins with LLMs, by Zihan Guan and 7 other authors View PDF HTML (experimental) Abstract:Mechanistic models encode scientific knowledge about dynamical systems and are widely used in downstream scientific and policy applications. Recent work has explored LLM-based agentic frameworks to automatically construct mechanistic models from data; however, existing problem settings substantially oversimplify real-world conditions, leaving it unclear whether LLM-generated mechanistic models are reliable in practice. To address this gap, we introduce the Neural-Integrated Mechanistic Modeling (NIMM) evaluation framework, which evaluates LLM-generated mechanistic models under realistic settings with partial observations and diversified task objectives. Our evaluation reveals fundamental challenges in current baselines, ranging from model effectiveness to code-level correctness. Motivated by these findings, we design NIMMgen, an agentic framework for neural-integrated mechanistic modeling that enhances code correctness and practical validity through iterative refinement. Experiments across three datasets from divers...
