Embodied AI Planning on BEHAVIOR-1K
100Success RateAgenticCache (Ours+)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| AgenticCache (Ours+)Model Scale=GPT-5, Execution Strategy=AgenticCache (Ours+)2026.04 | 100 | 1.55 | 1.9 | 6.6 | |
| AgenticCache (Ours+)Model Scale=GPT-5-mini, Execution Strategy=AgenticCache (Ours+)2026.04 | 100 | 0.97 | 1.8 | 1.2 | |
| BaselineModel Scale=GPT-5, Execution Strategy=Baseline2026.04 | 97.22 | 3.36 | 3 | 9.3 | |
| ParallelModel Scale=GPT-5, Execution Strategy=Parallel2026.04 | 97.22 | 3 | 4.6 | 15.3 | |
| SpeculativeModel Scale=GPT-5, Execution Strategy=Speculative2026.04 | 94.44 | 3.76 | 4.4 | 10.4 | |
| BaselineModel Scale=GPT-5-mini, Execution Strategy=Baseline2026.04 | 94.44 | 1.81 | 3.2 | 1.8 | |
| SpeculativeModel Scale=GPT-5-mini, Execution Strategy=Speculative2026.04 | 94.44 | 3.05 | 4.5 | 2.1 | |
| ParallelModel Scale=GPT-5-mini, Execution Strategy=Parallel2026.04 | 91.67 | 1.63 | 5 | 3.1 | |
| AgenticCache (Ours+)Model Scale=GPT-5-nano, Execution Strategy=AgenticCache (Ours+)2026.04 | 77.78 | 4.51 | 8.5 | 2 | |
| BaselineModel Scale=GPT-5-nano, Execution Strategy=Baseline2026.04 | 25 | 10.46 | 9.7 | 2.4 | |
| ParallelModel Scale=GPT-5-nano, Execution Strategy=Parallel2026.04 | 25 | 9.27 | 12.8 | 3.2 |