Optimization Modeling on IndustryOR (pass@1)
57Accuracy (pass@1)MIRROR
Evaluation results
| Method | Links | |
|---|---|---|
| MIRRORCategory=Ours2026.02 | 57 | |
| COECategory=Agent-based methods2026.02 | 55 | |
| MIRRORCategory=Ours, Model Scale=30B2026.02 | 53 | |
| Deepseek-v3Category=Traditional prompting2026.02 | 51 | |
| OptiMUSCategory=Agent-based methods2026.02 | 51 | |
| ORMindCategory=Agent-based methods2026.02 | 51 | |
| SIRLCategory=Learning-based methods, Model Scale=32B2026.02 | 48 | |
| Backbone modelCategory=Traditional prompting2026.02 | 46 | |
| LLMOPTType=Fine-tuning2026.04 | 46 | |
| qwen3-30BCategory=Traditional prompting, Model Scale=30B2026.02 | 41 | |
| NEDTreeType=Context-based2026.04 | 40 | |
| ORLMCategory=Learning-based methods, Model Scale=8B2026.02 | 38 | |
| ORLM(best)Type=Fine-tuning2026.04 | 38 | |
| ORllmAgentType=Prompt-based2026.04 | 35 | |
| GPT-4oType=Non-reasoning2026.04 | 34 | |
| GPT-o4-miniType=Reasoning2026.04 | 34 | |
| Deepseek-V3Type=Non-reasoning2026.04 | 33 | |
| Deepseek-R1Type=Reasoning2026.04 | 33 | |
| Chain-of-ExpertsType=Prompt-based2026.04 | 33 | |
| OptimusType=Prompt-based2026.04 | 32 | |
| LLMOPTCategory=Learning-based methods, Model Scale=14B2026.02 | 29 | |
| MiniOptCategory=Learning-based methods, Model Scale=14B2026.02 | 27 | |
| OptMATHCategory=Learning-based methods, Model Scale=7B2026.02 | 19 |