Multi-turn embodied reasoning on AlfWorld
92Success RateReMem
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReMemLLM Backbone=Claude 3.7 Sonnet2025.11 | 92 | 96 | |
| ExpRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 74 | 89 | |
| ReMemLLM Backbone=Gemini 2.5 Flash2025.11 | 66 | 81 | |
| ExpRecentLLM Backbone=Claude 3.7 Sonnet2025.11 | 66 | 83 | |
| ExpRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 59 | 79 | |
| SelfRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 52 | 75 | |
| ReActLLM Backbone=Claude 3.7 Sonnet2025.11 | 51 | 75 | |
| Mem0LLM Backbone=Claude 3.7 Sonnet2025.11 | 51 | 74 | |
| HistoryLLM Backbone=Claude 3.7 Sonnet2025.11 | 50 | 73 | |
| DC-CuLLM Backbone=Claude 3.7 Sonnet2025.11 | 50 | 74 | |
| DC-RSLLM Backbone=Claude 3.7 Sonnet2025.11 | 50 | 74 | |
| AWMLLM Backbone=Claude 3.7 Sonnet2025.11 | 49 | 73 | |
| AmemLLM Backbone=Claude 3.7 Sonnet2025.11 | 48 | 73 | |
| ExpRecentLLM Backbone=Gemini 2.5 Flash2025.11 | 37 | 65 | |
| HistoryLLM Backbone=Gemini 2.5 Flash2025.11 | 28 | 60 | |
| Mem0LLM Backbone=Gemini 2.5 Flash2025.11 | 27 | 61 | |
| DC-RSLLM Backbone=Gemini 2.5 Flash2025.11 | 27 | 60 | |
| AWMLLM Backbone=Gemini 2.5 Flash2025.11 | 26 | 59 | |
| AmemLLM Backbone=Gemini 2.5 Flash2025.11 | 25 | 59 | |
| SelfRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 25 | 59 | |
| DC-CuLLM Backbone=Gemini 2.5 Flash2025.11 | 25 | 59 | |
| ReActLLM Backbone=Gemini 2.5 Flash2025.11 | 24 | 56 | |
| BaselineLLM Backbone=Claude 3.7 Sonnet2025.11 | 18 | 49 | |
| BaselineLLM Backbone=Gemini 2.5 Flash2025.11 | 12 | 34 |