AI RESEARCH
ReaComp: Compiling LLM Reasoning into Symbolic Solvers for Efficient Program Synthesis
arXiv CS.CL
•
ArXi:2605.05485v1 Announce Type: new LLMs can solve program synthesis tasks but remain inefficient and unreliable on hard instances requiring large combinatorial search. Given a small set of reasoning traces, we use coding agents to compile them into reusable symbolic program synthesizers over constrained DSLs.