Embodied Household Tasks on ALFWorld Seen
90.7Average Accumulated RewardQ-Evolve
Evaluation Results
| Method | Links | |
|---|---|---|
| Q-EvolveBackbone=Llama-2-7B-Chat2026.06 | 90.7 | |
| QLASSBackbone=Llama-2-7B-Chat2026.06 | 77.9 | |
| ETOBackbone=Llama-2-7B-Chat2026.06 | 68.6 | |
| RFTBackbone=Llama-2-7B-Chat2026.06 | 62.9 | |
| Best-of-NBackbone=Llama-2-7B-Chat, N=62026.06 | 62.1 | |
| SFTBackbone=Llama-2-7B-Chat2026.06 | 60 | |
| ReflexionPrompting=ReAct2026.06 | 45.7 | |
| GPT-4Prompting=ReAct2026.06 | 42.9 | |
| PPOBackbone=Llama-2-7B-Chat2026.06 | 22.1 | |
| GPT-3.5-TurboPrompting=ReAct2026.06 | 7.9 | |
| Base AgentBackbone=Llama-2-7B-Chat2026.06 | 0 |