Long-context Memory Retrieval and Reasoning on PerM 128K V2
24.49F1 ScoreQwen3-30B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-30BBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 24.49 | |
| NemoriBackbone LLM=DeepSeek V3.22026.03 | 23.72 | |
| MemSifterBackbone LLM=DeepSeek V3.22026.03 | 23.57 | |
| A-MEMBackbone LLM=DeepSeek V3.22026.03 | 23.04 | |
| GemmaEmbBackbone LLM=DeepSeek V3.22026.03 | 22.95 | |
| HippoRAGBackbone LLM=DeepSeek V3.22026.03 | 22.94 | |
| MemSifterBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.81 | |
| ReasonRankBackbone LLM=DeepSeek V3.22026.03 | 22.27 | |
| RearankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 22.17 | |
| Mem0Backbone LLM=DeepSeek V3.22026.03 | 21.63 | |
| DS-V3.2Backbone LLM=DeepSeek V3.22026.03 | 21.55 | |
| BGE-M3Backbone LLM=DeepSeek V3.22026.03 | 21.45 | |
| Mem0Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 21.11 | |
| A-MEMBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 21.07 | |
| HippoRAGBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.62 | |
| BGE-M3Backbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.21 | |
| GemmaEmbBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 20.12 | |
| RearankBackbone LLM=DeepSeek V3.22026.03 | 19.76 | |
| ReasonRankBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 19.05 | |
| NemoriBackbone LLM=Qwen3-30B-A3B-Ins2026.03 | 18.62 |