Multi-agent planning on MAP-THOR 2-agent
66SRLLaMAR
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLaMARLM=GPT-4V, explanation=w/ expl2024.07 | 66 | 91 | 97 | 82 | |
| LLaMARLM=GPT-4V, explanation=w/o expl2024.07 | 62 | 87 | 95 | 82 | |
| LLaMARLM=CogVLM2024.07 | 61 | 89 | 95 | 80 | |
| LLaMARLM=IDEFICS-22024.07 | 57 | 86 | 94 | 78 | |
| LLaMARLM=LLaVA2024.07 | 54 | 84 | 91 | 75 | |
| LLaMARLM=GPT-42024.07 | 51 | 85 | 95 | 83 | |
| ReActLM=GPT-4V2024.07 | 34 | 72 | 92 | 67 | |
| ActLM=GPT-4V2024.07 | 33 | 67 | 91 | 59 | |
| CoELALM=GPT-4V2024.07 | 25 | 46 | 76 | 73 | |
| CoTLM=GPT-4V2024.07 | 14 | 59 | 87 | 62 | |
| SmartLLMLM=GPT-4V2024.07 | 11 | 23 | 91 | 45 |