ChatSR: Conversational Symbolic Regression

Yanjie Li, Weijun Li, Lina Yu, Min Wu, Jingyi Liu, Shu Wei, Yusong Deng

Keywords: Symbolic Regression, Multi-modal Large Language Models, Scientific discovery

TL;DR: A large language model for symbolic regression that can fit the data by generating desired expressions based on natural language prompts.

Abstract: Formulas are the language of communication between humans and nature. It is an important research topic of artificial intelligence to find expressions from observed data to reflect the relationship between each variable in the data, which is called a symbolic regression problem. The existing symbolic regression methods directly generate expressions according to the given observation data, but we cannot require the algorithm to generate expressions that meet specific requirements according to the known prior knowledge. For example, the expression needs to contain the symbol `$\sin$' or be periodicity, and so on. Even if it can, it often requires very complex operations, which is very inconvenient. In this paper, based on multi-modal large language models, we propose ChatSR, a conversational symbolic regression method that can generate expressions that meet the requirements simply by describing the requirements with natural language instructions. By experimenting on the test datasets, we can demonstrate that ChatSR leads the state-of-the-art baselines in fitting performance. More notably, ChatSR can well understand the prior knowledge contained in natural language prompts, and can further improve the quality of generated expressions according to the prior knowledge. In addition, it is exciting that ChatSR has good zero-shot capability.

Primary Area: foundation or frontier models, including LLMs

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 7284