Long-context Memory Retrieval and Reasoning on PersonaMem 128K
23.75F1 ScoreMemSifter
Evaluation Results
| Method | Links | |
|---|---|---|
| MemSifterBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 23.75 | |
| RearankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.95 | |
| BGE-M3Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.73 | |
| NemoriBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.69 | |
| Mem0Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.51 | |
| GemmaEmbBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.36 | |
| HippoRAGBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 21.31 | |
| MemSifterBackbone LLM=DeepSeek V3.22026.03 | 21.14 | |
| Qwen3-30BBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.82 | |
| A-MEMBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.78 | |
| ReasonRankBackbone LLM=DeepSeek V3.22026.03 | 20.74 | |
| Mem0Backbone LLM=DeepSeek V3.22026.03 | 20.63 | |
| ReasonRankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.18 | |
| DS-V3.2Backbone LLM=DeepSeek V3.22026.03 | 20.15 | |
| NemoriBackbone LLM=DeepSeek V3.22026.03 | 19.89 | |
| HippoRAGBackbone LLM=DeepSeek V3.22026.03 | 19.24 | |
| BGE-M3Backbone LLM=DeepSeek V3.22026.03 | 18.83 | |
| RearankBackbone LLM=DeepSeek V3.22026.03 | 18.41 | |
| GemmaEmbBackbone LLM=DeepSeek V3.22026.03 | 18.22 | |
| A-MEMBackbone LLM=DeepSeek V3.22026.03 | 18.14 |