Planning on PlanBench Blocksworld
100Planning CoverageDeepSeek-R1
Evaluation results
| Method | Links | |
|---|---|---|
| DeepSeek-R1Mode=LLM-fb2026.06 | 100 | |
| DeepSeek-R1Mode=LLM-direct2026.06 | 99.1 | |
| Claude-3-5Mode=LLM-fb2026.06 | 83.5 | |
| GPT-4oMode=LLM-fb2026.06 | 69.5 | |
| Claude-3-5Mode=LLM-direct2026.06 | 54.8 | |
| GPT-4oMode=LLM-direct2026.06 | 35.5 |