Tool-use Reasoning on ToolBench
88API Success RateExpRAG
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ExpRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 88 | 72 | |
| ReMemLLM Backbone=Claude 3.7 Sonnet2025.11 | 87 | 71 | |
| ExpRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 87 | 73 | |
| ReMemLLM Backbone=Gemini 2.5 Flash2025.11 | 85 | 71 | |
| ExpRecentLLM Backbone=Claude 3.7 Sonnet2025.11 | 82 | 66 | |
| ExpRecentLLM Backbone=Gemini 2.5 Flash2025.11 | 78 | 66 | |
| AmemLLM Backbone=Claude 3.7 Sonnet2025.11 | 77 | 63 | |
| SelfRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 77 | 63 | |
| LangMemLLM Backbone=Claude 3.7 Sonnet2025.11 | 77 | 63 | |
| DC-CuLLM Backbone=Claude 3.7 Sonnet2025.11 | 77 | 62 | |
| DC-RSLLM Backbone=Claude 3.7 Sonnet2025.11 | 77 | 62 | |
| BaselineLLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 62 | |
| HistoryLLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 61 | |
| ReActLLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 61 | |
| MemOSLLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 62 | |
| Mem0LLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 61 | |
| AWMLLM Backbone=Claude 3.7 Sonnet2025.11 | 76 | 62 | |
| AmemLLM Backbone=Gemini 2.5 Flash2025.11 | 72 | 60 | |
| SelfRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 72 | 61 | |
| BaselineLLM Backbone=Gemini 2.5 Flash2025.11 | 71 | 61 | |
| MemOSLLM Backbone=Gemini 2.5 Flash2025.11 | 71 | 61 | |
| Mem0LLM Backbone=Gemini 2.5 Flash2025.11 | 71 | 61 | |
| AWMLLM Backbone=Gemini 2.5 Flash2025.11 | 71 | 59 | |
| DC-CuLLM Backbone=Gemini 2.5 Flash2025.11 | 70 | 59 | |
| LangMemLLM Backbone=Gemini 2.5 Flash2025.11 | 68 | 57 | |
| DC-RSLLM Backbone=Gemini 2.5 Flash2025.11 | 68 | 57 | |
| ReActLLM Backbone=Gemini 2.5 Flash2025.11 | 64 | 57 | |
| HistoryLLM Backbone=Gemini 2.5 Flash2025.11 | 62 | 54 |