Multi-turn Agentic Task on PDDL
83Success RateReMem
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReMemLLM Backbone=Claude 3.7 Sonnet2025.11 | 83 | 95 | |
| ReActLLM Backbone=Claude 3.7 Sonnet2025.11 | 75 | 91 | |
| EVOLIB2026.05 | 72.8 | 86.2 | |
| ExpRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 72 | 89 | |
| DCType=test-time learning (TTL)2026.05 | 65 | 83.6 | |
| HistoryLLM Backbone=Claude 3.7 Sonnet2025.11 | 65 | 85 | |
| SelfRAGLLM Backbone=Claude 3.7 Sonnet2025.11 | 65 | 84 | |
| Mem0LLM Backbone=Claude 3.7 Sonnet2025.11 | 65 | 84 | |
| BaseType=base sampling2026.05 | 62.8 | 79.1 | |
| AmemLLM Backbone=Claude 3.7 Sonnet2025.11 | 62 | 84 | |
| DC-CuLLM Backbone=Claude 3.7 Sonnet2025.11 | 62 | 84 | |
| DC-RSLLM Backbone=Claude 3.7 Sonnet2025.11 | 62 | 84 | |
| AWMLLM Backbone=Claude 3.7 Sonnet2025.11 | 60 | 82 | |
| ExpRecentLLM Backbone=Claude 3.7 Sonnet2025.11 | 53 | 76 | |
| ExpRAGType=test-time learning (TTL)2026.05 | 39.4 | 56.7 | |
| ReActLLM Backbone=Gemini 2.5 Flash2025.11 | 22 | 33 | |
| ReMemLLM Backbone=Gemini 2.5 Flash2025.11 | 22 | 33 | |
| ExpRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 17 | 27 | |
| BaselineLLM Backbone=Claude 3.7 Sonnet2025.11 | 17 | 39 | |
| ExpRecentLLM Backbone=Gemini 2.5 Flash2025.11 | 13 | 22 | |
| BaselineLLM Backbone=Gemini 2.5 Flash2025.11 | 12 | 20 | |
| AmemLLM Backbone=Gemini 2.5 Flash2025.11 | 10 | 16 | |
| Mem0LLM Backbone=Gemini 2.5 Flash2025.11 | 10 | 19 | |
| HistoryLLM Backbone=Gemini 2.5 Flash2025.11 | 8 | 15 | |
| SelfRAGLLM Backbone=Gemini 2.5 Flash2025.11 | 8 | 16 | |
| DC-CuLLM Backbone=Gemini 2.5 Flash2025.11 | 8 | 17 | |
| AWMLLM Backbone=Gemini 2.5 Flash2025.11 | 8 | 16 | |
| DC-RSLLM Backbone=Gemini 2.5 Flash2025.11 | 7 | 15 |