{"id":"arxiv/2601.21008","url":"https://arxiv.org/abs/2601.21008","tags":["cs.LG","cs.AI","math.OC"],"score":0,"title":"ORLoopBench: Solver-in-the-Loop Benchmarks for Self-Correction and Behavioral Rationality in Operations Research","domain":"cs.LG","source":"arxiv","authors":["Ruicheng Ao","David Simchi-Levi","Xinshang Wang"],"reviews":0,"subject":"cs.LG","summary":"Operations Research practitioners debug infeasible models through an iterative process: inspecting Irreducible Infeasible Subsystems ( IIS), identifying constraint conflicts, and repairing formulations until feasibility is restored. Existing LLM benchmarks mostly treat OR as one-shot translation from problem descriptions to solver code, omitting this diagnostic loop. We formalize infeasible-model repair as a solver-in-the-loop Markov Decision Process in which each action triggers solver re-execution and IIS recomputation, yielding deterministic, verifiable feedback. We introduce ORLoopBench,…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-01-28","reactions":{}}