Permutation Flow Shop Scheduling Problem on PFSP
0.0086Optimality GapFALCON
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FALCONSampling strategy=Adaptive2026.02 | 0.0086 | 1 | |
| FALCONNumber of samples (N)=82026.02 | 0.0089 | 1 | |
| LLMCoSolverNumber of samples (N)=82026.02 | 0.0103 | 1 | |
| FALCONNumber of samples (N)=12026.02 | 0.0176 | 1 | |
| GRPOCategory=Neural CO Solver2026.02 | 0.0212 | 1 | |
| LLMCoSolverNumber of samples (N)=12026.02 | 0.0212 | 1 | |
| SFT OnlyCategory=Neural CO Solver2026.02 | 0.0222 | 1 | |
| SGECategory=LLM-based Optimization2026.02 | 0.0448 | 0.95 | |
| LMEACategory=LLM-based Optimization2026.02 | 0.1431 | 0.98 | |
| GPT-o1Category=Reasoning Model2026.02 | 0.1486 | 0.89 | |
| DeepSeek-R1Category=Reasoning Model2026.02 | 0.1665 | 1 | |
| GPT-o3-miniCategory=Reasoning Model2026.02 | 0.1697 | 0.98 | |
| PHPCategory=LLM-based Optimization2026.02 | 0.1723 | 0.92 | |
| OPROCategory=LLM-based Optimization2026.02 | 0.184 | 0.99 | |
| Claude-3.5-SonnetEvaluation protocol=Zero-shot2026.02 | 0.1842 | 1 | |
| GPT-4oEvaluation protocol=Zero-shot2026.02 | 0.2057 | 0.88 | |
| Claude-3.5-HaikuEvaluation protocol=Zero-shot2026.02 | 0.2058 | 0.73 | |
| DeepSeek-V3Evaluation protocol=Zero-shot2026.02 | 0.2081 | 0.58 | |
| Qwen2.5-72BEvaluation protocol=Zero-shot2026.02 | 0.2113 | 0.98 | |
| GPT-4o-miniEvaluation protocol=Zero-shot2026.02 | 0.2147 | 0.78 | |
| Llama-3.3-70BEvaluation protocol=Zero-shot2026.02 | 0.2197 | 0.98 | |
| EoHBackbone LLM=GPT-5.4, Time (s)=2132, Tokens=1.74M2026.05 | 4.64 | — | |
| CORALBackbone LLM=GPT-5.4, Time (s)=3744, Tokens=1.67M2026.05 | 5.04 | — | |
| HMACEBackbone LLM=GPT-5.4, Time (s)=1472, Tokens=0.72M2026.05 | 5.73 | — |