Personalized Retrieval and QA on PersonaBench (Noise 0.3)
29.89F1 ScoreMemCoE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MemCoEBackbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 29.89 | 52.02 | |
| A-MemBackbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 28.56 | 42.64 | |
| RAGBackbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 28.16 | 36.68 | |
| Mem0Backbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 19.75 | 38.23 | |
| MemAgentBackbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 19.36 | 45 | |
| Long ContextBackbone=Qwen2.5-7B-Instruct2026.05 | 19.1 | 26.9 | |
| LightMemBackbone=Qwen2.5-7B-Instruct, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 18.74 | 41.21 | |
| Mem-αBackbone=Qwen3-4B, Retrieval model=all-MiniLM-L6-v2, Top-K retrieval=102026.05 | 17.02 | 44.19 |