Reproduce “EvoMO-SR: Multiobjective LLM-based Evolution of Symbolic Expressions with substructure guidance” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Cristina Rossetti, Anna V. Kononova, Thomas Bäck, Fei Liu, Niki Van Stein
EvoMO-SR discovers equations by having an LLM propose and revise equation structures while an optimizer fits coefficients; evolution selects for fit and expression size, and archived substructures guide some mutations. On LSR-Synth, it has lower OOD NMSE than LLM-SR in all four domains. In Material Science OOD, median NMSE is 2.73e-6 versus 6.67e-3, while structural recovery remains low overall and varies by metric and domain.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.