AI Paper Atlas

Research map / Large language models / Agents, code and math

ORLoopBench: Solver-in-the-Loop Benchmarks for Self-Correction and Behavioral Rationality in Operations Research

Ruicheng Ao, David Simchi-Levi, Xinshang Wang

ICML 2026 · Large language models › Agents, code and math · approach: Code generation benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train Qwen3-8B with solver-verified GRPO on top of SFT/curriculumQwen3-8B · OR-DEBUG-BENCH LP repairRR@5: 95.3% vs 92.4% (best API)Improved
add SFT on teacher trajectories to base model (ablation)Qwen3-8B · OR-DEBUG-BENCH pilot validationRR@5: 51.2% -> 93.1%Improved
add GRPO to SFT baseline (ablation)Qwen3-8B SFT · OR-DEBUG-BENCH validation (200)RR@5 / DA: RR@5 91.5% -> 92.0%; DA 68.0% -> 66.0%Mixed
add curriculum stage before GRPO/SFT (ablation)Qwen3-8B SFT · OR-DEBUG-BENCH validation (200)RR@5: 91.5% -> 95.0%Improved
add process reward model (PRM) supervision (ablation)Qwen3-8B SFT · OR-DEBUG-BENCH validation (200)DA: 68.0% -> 72.7%Improved
combine curriculum with GRPO (ablation)Qwen3-8B SFT · OR-DEBUG-BENCH validation (200)RR@5 / DA: RR@5 91.5% -> 95.3%; DA 68.0% -> 62.4%Mixed
remove faithfulness penalty from GRPO reward (ablation)Qwen3-8B GRPO · OR-DEBUG-BENCHdiagnostic qualityHurt
constraint-level repair instead of whole-model code regenerationMILP repair · OR-DEBUG-BENCH MILP semantic driftcorrect objective preserved: GPT-5.4 regeneration: 90% OPTIMAL but 28.2% correct objectiveImproved
zero-shot transfer of LP-trained model to MILP, then add MILP-specific trainingQwen3-8B · OR-DEBUG-BENCH MILP repairRR@5: 78.8% -> 87.1%Improved
add three-stage curriculum (Extreme, Boundary, Full) for biasQwen3-8B · OR-BIAS-BENCH OODOOD bias: 20.0% -> 10.4%Improved
prompting only (zero-shot CoT or three-shot ICL) instead of training (ablation)Qwen3-8B · OR-DEBUG-BENCHRR@5: 23.0% (CoT), 38.7% (3-shot ICL)Hurt
train on OR formulation (ORLM/SIRL/LLMOPT) instead of repairOR-domain models · OR-DEBUG-BENCH LP repairRR@5: 71.0%/73.8%/78.4% vs 95.3%Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy