Text-based embodied task on Alfworld
97.76Success RateQwen2.5-Actor
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-ActorActor=Qwen2.5-Actor, Thinker=Exp-Thinker2026.05 | 97.76 | 98.5 | |
| Qwen2.5-ActorActor=Qwen2.5-Actor, Thinker=No2026.05 | 94.77 | 97.76 | |
| Qwen2.5-72BActor=Qwen2.5-72B, Thinker=No2026.05 | 91.79 | 93.84 | |
| Qwen2.5-7BActor=Qwen2.5-7B, Thinker=Exp-Thinker2026.05 | 90.29 | 94.96 | |
| LLaMA3-ActorActor=LLaMA3-Actor, Thinker=Exp-Thinker2026.05 | 77.61 | 87.75 | |
| LLaMA3-ActorActor=LLaMA3-Actor, Thinker=No2026.05 | 68.66 | 78.92 | |
| Qwen2.5-7BActor=Qwen2.5-7B, Thinker=No2026.05 | 68.65 | 80.03 | |
| Agent-FLANActor=Agent-FLAN, Thinker=No2026.05 | 67.2 | 79.7 | |
| GPT-4oActor=GPT-4o, Thinker=No2026.05 | 66.4 | 79.9 | |
| AgentGymActor=AgentGym, Thinker=No2026.05 | 61.9 | 76.9 | |
| AgentRefineActor=AgentRefine, Thinker=No2026.05 | 44.8 | 63.8 | |
| LLaMA3-8BActor=LLaMA3-8B, Thinker=Exp-Thinker2026.05 | 32.08 | 63.49 | |
| LLaMA3-8BActor=LLaMA3-8B, Thinker=No2026.05 | 7.46 | 32.15 |