3 papers
cs.LG2026
ERBench: A Benchmark and Testsuite for Equation Discovery Algorithms
Paul Kahlmeyer, Henrik Voigt, Michael Habeck +1
Equation discovery aims to automate the discovery of scientific models in the form of mathematical equations from data. Technically, equation discovery is implemented by symbolic r…
cs.LG2026
Beyond the Training Distribution: Mapping Generalization Boundaries in Neural Program Synthesis
Henrik Voigt, Michael Habeck, Joachim Giesen
Large-scale transformers achieve impressive results on program synthesis benchmarks, yet their true generalization capabilities remain obscured by data contamination and opaque tra…
cs.LG2025
Analyzing Generalization in Pre-Trained Symbolic Regression
Henrik Voigt, Paul Kahlmeyer, Kai Lawonn +2
Symbolic regression algorithms search a space of mathematical expressions for formulas that explain given data. Transformer-based models have emerged as a promising, scalable appro…