Research map / Large language models / Agents, code and math
Constructing Industrial-Scale Optimization Modeling Benchmark
Zhong Li, Hongliang Lu, Tao Wei, Yuxuan Chen, Wenyu Liu, Yuan Lan, Fan Zhang, Zaiwen Wen
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 changed | Where | Measured | Effect |
|---|---|---|---|
| evaluate on MIPLIB-NL instead of NL4Opt | SIRL-Qwen2.5-32B · NL4Opt vs MIPLIB-NL | Pass@1 accuracy (%): 87.90 -> 1.43 | Hurt |
| evaluate on MIPLIB-NL instead of NL4Opt | GPT-5.1 · NL4Opt vs MIPLIB-NL | Pass@1 accuracy (%): 81.78 -> 39.05 | Hurt |
| evaluate on MIPLIB-NL instead of NL4Opt | bare LLMs average · NL4Opt vs MIPLIB-NL | Pass@1 accuracy (%): 78.78 -> 17.85 | Hurt |
| evaluate on MIPLIB-NL instead of MAMO-E | OptMATH-Qwen2.5-7B · MAMO-E vs MIPLIB-NL | Pass@1 accuracy (%): 97.06 -> 0.00 | Hurt |
| enable thinking mode | Claude-Sonnet-4.5 · MIPLIB-NL | Pass@1 accuracy (%): 24.23 -> 30.00 | Improved |
| enable thinking mode | Claude-Sonnet-4.5 · IndustryOR | Pass@1 accuracy (%): 71.43 -> 76.19 | Improved |
| enable thinking mode | DeepSeek-V3.2 · MIPLIB-NL | Pass@1 accuracy (%): 23.81 -> 22.38 | No clear effect |
| enable thinking mode | Qwen3-Max · MIPLIB-NL | Pass@1 accuracy (%): 21.43 -> 21.43 | No clear effect |
| enable thinking mode | Qwen3-Max · IndustryOR | Pass@1 accuracy (%): 47.62 -> 59.52 | Improved |
| fine-tune on existing toy datasets (SIRL 32B vs 7B scaling) | SIRL-Qwen2.5 · MIPLIB-NL | Pass@1 accuracy (%): 0.48 -> 1.43 | No clear effect |
| increase problem scale from small to very large | bare LLMs · MIPLIB-NL | average Pass@1 | Mixed |
Closest papers by meaning
- MM-OptBench: A Solver-Grounded Benchmark for Multimodal Optimization Modeling (NeurIPS 2026 · similarity 0.89)
- Autoformulation of Mathematical Optimization Models Using LLMs (ICML 2025 · similarity 0.89)
- Solver-Informed RL: Grounding Large Language Models for Authentic Optimization Modeling (NeurIPS 2025 · similarity 0.89)
- PEARL: Solver-in-the-Loop Interactive Optimization Modeling from Natural Language (NeurIPS 2026 · similarity 0.89)
- OptMATH: A Scalable Bidirectional Data Synthesis Framework for Optimization Modeling (ICML 2025 · similarity 0.88)
- Enhancing Decision Making Through the Integration of Large Language Models and Operations Research Optimization (AAAI 2025 · similarity 0.88)
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.
