AI RESEARCH
Formalize, Don't Optimize: The Heuristic Trap in LLM-Generated Combinatorial Solvers
arXiv CS.AI
•
ArXi:2605.12421v1 Announce Type: new Large Language Models (LLMs) struggle to solve complex combinatorial problems through direct reasoning, so recent neuro-symbolic systems increasingly use them to synthesize executable solvers. A central design question is how the LLM should represent the solver, and whether it should also attempt to optimize search. We