---
title: "DASyR-LLM: Domain-Aware Symbolic Regression with LLMs for Kinetic Model Discovery"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.05120v1"
permalink: "https://feed.astralor.com/items/itm_61e9ec2c35bb7e8e47211f86"
published: "2026-08-05T17:48:45.000Z"
score: 72
channel: "tech"
tags: ["DASyR-LLM","Kinetic Model","Chemical Engineering","Large Language Models","Symbolic Regression"]
---

# DASyR-LLM: Domain-Aware Symbolic Regression with LLMs for Kinetic Model Discovery

## AI commentary

> This paper is valuable because it extends LLMs from writing code to making scientific hypotheses, especially for chemical fields where wet-lab experiments are costly. If you work on kinetic modeling or AI4Science, it is worth comparing with your own search strategy; even outside the field, it showcases an efficient paradigm for injecting domain knowledge into AI.

## At a glance

Kinetic model discovery in chemical engineering depends on accurate rate expressions, yet traditional symbolic regression often lacks domain knowledge and tends to explore physicochemically implausible models. This study introduces DASyR-LLM, which embeds a large language model into an iterative symbolic regression algorithm. The LLM plays dual roles: providing qualitative physicochemical critiques of candidate models and proposing new rate expressions guided by chemical knowledge. Across four in silico case studies, the framework reduces iterations needed to identify the ground-truth model by 41.7%-79.3%, with the LLM directly proposing correct model structures in over half of guided runs, while maintaining equivalent predictive performance (R²>0.98) on independent validation sets. This substantially reduces wet-lab experimental effort and paves the way for fully automated, domain-aware kinetic modeling pipelines.

## Key points

- Introduces DASyR-LLM, an LLM-guided symbolic regression framework with dual roles of critique and generation.
- Reduces iterations to identify ground-truth models by 41.7%-79.3% across four catalysis and bioprocess case studies.
- LLM directly proposes the correct model structure in over half of guided runs while maintaining R²>0.98.
- Ablation studies show both the SR component and LLM scale matter, with a smaller LLM retaining discovery efficiency.

## Full text

Kinetic model discovery is a central challenge in chemical engineering, as accurate rate expressions are essential for understanding and controlling chemical and biological processes. Symbolic regression (SR) has emerged as a powerful data-driven approach for identifying interpretable kinetic models, but usually operates without domain knowledge, often exploring physicochemically implausible models. Large language models (LLMs) offer a promising avenue for injecting domain expertise into this search. Here, we introduce an LLM-guided SR framework, embedding an LLM module within an iterative SR algorithm for automated kinetic model discovery. The LLM performs two roles at each iteration: (1) a qualitative physicochemical critique of the best SR candidates, and (2) the proposal of new candidate rate expressions guided by the SR-generated models and embedded chemical knowledge. Our framework is evaluated on four in silico case studies of increasing complexity, spanning heterogeneous catalysis and bioprocess systems. Results show the LLM-guided framework reduces iterations to identify the ground-truth model by $41.7-79.3\%$ versus a state-of-the-art SR framework, with the LLM directly proposing the correct model structure in over half of the guided runs. In practical settings, where each iteration typically requires a new wet-lab experiment, this translates into a substantial reduction in experimental effort. Predictive performance on an independent validation set is equivalent between both approaches, with $R^2>0.98$ in all case studies. Ablation studies indicate that both the SR component and the LLM scale contribute to this performance, with a reduced-size LLM largely retaining discovery efficiency. These findings demonstrate that LLMs can effectively inject domain knowledge into scientific model discovery, paving the way toward fully automated, domain-aware kinetic modelling pipelines.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_61e9ec2c35bb7e8e47211f86