Shifts symbolic regression from discrete genetic search to a continuous, embedding-driven optimization paradigm.
March 25, 2026
Original Paper
Neural Structure Embedding for Symbolic Regression via Continuous Structure Search and Coefficient Optimization
arXiv · 2603.22429
The Takeaway
By projecting symbolic structures into a continuous latent space, the framework allows for gradient-based search of mathematical equations. This significantly reduces the computational cost of navigating the combinatorial space of expressions while improving robustness to noise.
From the abstract
Symbolic regression aims to discover human-interpretable equations that explain observational data. However, existing approaches rely heavily on discrete structure search (e.g., genetic programming), which often leads to high computational cost, unstable performance, and limited scalability to large equation spaces. To address these challenges, we propose SRCO, a unified embedding-driven framework for symbolic regression that transforms symbolic structures into a continuous, optimizable represen