Long-context Memory Retrieval and Reasoning on LongMemEval 128K
47.26F1 ScoreMemSifter
Evaluation Results
| Method | Links | |
|---|---|---|
| MemSifterBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 47.26 | |
| NemoriBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 46.09 | |
| ReasonRankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 44.26 | |
| GemmaEmbBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 43.04 | |
| Qwen3-30BBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 42.2 | |
| BGE-M3Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 41.75 | |
| A-MEMBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 40.46 | |
| HippoRAGBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 40.13 | |
| RearankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 39.55 | |
| Mem0Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 38.56 | |
| MemSifterBackbone LLM=DeepSeek V3.22026.03 | 35.38 | |
| DS-V3.2Backbone LLM=DeepSeek V3.22026.03 | 35.15 | |
| NemoriBackbone LLM=DeepSeek V3.22026.03 | 34.39 | |
| ReasonRankBackbone LLM=DeepSeek V3.22026.03 | 34.31 | |
| Mem0Backbone LLM=DeepSeek V3.22026.03 | 33.93 | |
| HippoRAGBackbone LLM=DeepSeek V3.22026.03 | 33.64 | |
| RearankBackbone LLM=DeepSeek V3.22026.03 | 33.53 | |
| GemmaEmbBackbone LLM=DeepSeek V3.22026.03 | 32.62 | |
| A-MEMBackbone LLM=DeepSeek V3.22026.03 | 31.96 | |
| BGE-M3Backbone LLM=DeepSeek V3.22026.03 | 31.62 |