Show HN: Modernizing my old PhD work in an evening with little Qwen3.6 MoE

Category: ai-ml

Tags: symbolic-regression, jax, grammar-enumeration

Score: 6.8/10 (Innovation: 6, Technical: 8, Documentation: 7, Utility: 6)

pge-jax is a JAX-native implementation of the Prioritized Grammar Enumeration algorithm for symbolic regression, automatically discovering interpretable mathematical formulas from data. It leverages GPU/TPU acceleration, JIT compilation, and automatic differentiation to efficiently enumerate and fit candidate expressions, making it a modern, high-performance tool for interpretable machine learning.

Target audience: data scientists, machine learning researchers, quantitative analysts

Repository: https://github.com/verdverm/pge-jax · Python · Apache-2.0 · 1 stars

View on Hacker News