Paper proposes agentic reasoning method for sample-efficient symbolic regression with LLMs

Arxiv CS.CL (NLP/language models)

The Deliberate Evolution method uses LLM-based agentic reasoning to guide evolutionary search, achieving sample efficiency by incorporating intermediate reasoning steps rather than only final error. Experiments on symbolic regression benchmarks demonstrate reduced number of evaluations needed to discover compact expressions.

llmsymbolic-regression

Arxiv CS.CL (NLP/language models)·Xinyu PangZhanke ZhouXuan LiFangrui LvShanshan WeiSen CuiBo HanChangshui Zhang

7 days ago