PySR v2.0 Released: High-Performance Symbolic Regression for Discovering Math Laws
MilesCranmer · x · 2026-08-20
PySR v2.0 is releasing soon, with a beta version currently available for testing. PySR is a high-performance symbolic regression library built on Python and Julia, designed to discover interpretable symbolic expressions from data instead of producing black-box models.
Key features include:
- Interpretable by Design: Finds human-readable mathematical equations.
- Production Ready: Mature, highly optimized parallel evolutionary algorithms.
- Extremely Customizable: Allows configuration of operators, loss functions, complexity, and optimizers.
- Neural Net Interpretation: Supports "Symbolic Distillation" to convert neural networks into analytic equations, aiding in the interpretation of deep learning models (e.g., in N-body problems).
It is particularly useful for scientific discovery on low-dimensional datasets and as a tool for explaining deep neural networks.
More from Research
- SAI Lab Launches SAI Arena to Evaluate AI Verification Systems — ChenhaoTan · 2026-08-20
- Solving Sim-to-Real challenges with Domain Randomization — ShawnHymel · 2026-08-20
- AI-Driven 'Exercise Pill' Passes Human Trial, Revolutionizing Drug Discovery — TansuYegen · 2026-08-20
- Melanie Mitchell: AI is 'Alien Intelligence' Requiring New Metrics — MelMitchell1 · 2026-08-20
- Humyn Labs: Turning Human Experience into Physical AI Training Data — kimmonismus · 2026-08-20
- OpenMaMMUT Collection: Weights on DataComp and Re-LAION Available — wightmanr · 2026-08-20