STRIDE: A Self-Reflective Agent Framework for Reliable Automatic Equation Discovery

In three linesSTRIDE is a self-reflective agent framework for LLM-based symbolic equation discovery. It improves reliability by coordinating data-aware generation, mixed-fitting evaluation, critic-executor repair, and diversity-preserving semantic memory. Experiments on symbolic regression benchmarks show gains in accuracy, OOD robustness, and structural recovery across multiple LLM backbones.

Summary generated by Claude — human-verified