Feynman

The 'Feynman' dataset is a benchmark that contains a collection of mathematical equations derived from physics problems, used to evaluate the performance of models in symbolic regression and equation discovery tasks.

Papers using Feynman (3)

Data-Efficient Symbolic Regression via Foundation Model Distillation2025 · 1 citesSymMatika: Structure-Aware Symbolic Discovery2025 · 1 citesInitialization Schemes for Kolmogorov-Arnold Networks: An Empirical Study2025