Operations Research Problem Solving on NL4OPT (test)
80.4Correct RateBPP-Search Variant
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BPP-Search VariantPolicy Model=GPT-4o2024.11 | 80.4 | 15 | |
| Beam Search VariantPolicy Model=GPT-4o2024.11 | 78.3 | 15 | |
| SCPolicy Model=GPT-4o2024.11 | 71.3 | 4 | |
| Greedy Search VariantPolicy Model=GPT-4o2024.11 | 71.3 | 9 | |
| ToT-Randomly-ChosenPolicy Model=GPT-4o2024.11 | 62.9 | 39 | |
| ToT-RethinkPolicy Model=GPT-4o2024.11 | 62.2 | 40 | |
| CoTPolicy Model=GPT-4o2024.11 | 56.6 | 1 |