Long-context Memory Retrieval and Reasoning on WebWalker 128K
27.44F1 ScoreMemSifter
Evaluation Results
| Method | Links | |
|---|---|---|
| MemSifterBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 27.44 | |
| A-MEMBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 26.13 | |
| MemSifterBackbone LLM=DeepSeek V3.22026.03 | 26.11 | |
| NemoriBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 25.8 | |
| ReasonRankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 25.72 | |
| Qwen3-30BBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 25.25 | |
| HippoRAGBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 25.16 | |
| BGE-M3Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 25.01 | |
| GemmaEmbBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 24.51 | |
| NemoriBackbone LLM=DeepSeek V3.22026.03 | 24.39 | |
| Mem0Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 24.24 | |
| HippoRAGBackbone LLM=DeepSeek V3.22026.03 | 24.19 | |
| RearankBackbone LLM=DeepSeek V3.22026.03 | 24.19 | |
| RearankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 24.14 | |
| Mem0Backbone LLM=DeepSeek V3.22026.03 | 23.46 | |
| ReasonRankBackbone LLM=DeepSeek V3.22026.03 | 22.79 | |
| BGE-M3Backbone LLM=DeepSeek V3.22026.03 | 22.61 | |
| DS-V3.2Backbone LLM=DeepSeek V3.22026.03 | 22.49 | |
| A-MEMBackbone LLM=DeepSeek V3.22026.03 | 21.88 | |
| GemmaEmbBackbone LLM=DeepSeek V3.22026.03 | 21.69 |