MemoryBench Task on MemoryBench Short-Input-Long-Output 1.0
77.09Norm-ScoreUNO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| UNOModel=Qwen3-8B, Method Type=Training2026.02 | 77.09 | 7.16 | |
| UNO-SingleModel=Qwen3-8B, Method Type=Training2026.02 | 76.36 | 2.99 | |
| MemoryOSModel=Qwen3-8B, Method Type=Memory2026.02 | 74.62 | -8.96 | |
| BM25Model=Qwen3-8B, Method Type=RAG2026.02 | 74.54 | -8.83 | |
| BaseModel=Qwen3-8B, Method Type=Baseline2026.02 | 74.43 | -9.52 | |
| EmbeddingModel=Qwen3-8B, Method Type=RAG2026.02 | 73.22 | -17.04 | |
| MemOSModel=Qwen3-8B, Method Type=Memory2026.02 | 72.29 | -23.07 | |
| ReMemModel=Qwen3-8B, Method Type=Memory2026.02 | 71.95 | -25.55 | |
| UNOModel=phi-4, Method Type=Training2026.02 | 70.66 | -36.69 | |
| DPO w/ UNO DataModel=Qwen3-8B, Method Type=Training2026.02 | 70.39 | -34.68 | |
| A-MemModel=Qwen3-8B, Method Type=Memory2026.02 | 70.36 | -35.02 | |
| SFT w/ UNO DataModel=Qwen3-8B, Method Type=Training2026.02 | 69.95 | -37.85 | |
| UNO-SingleModel=phi-4, Method Type=Training2026.02 | 69.61 | -43.08 | |
| BM25Model=phi-4, Method Type=RAG2026.02 | 69.33 | -44.72 | |
| Mem0Model=phi-4, Method Type=Memory2026.02 | 68.81 | -48.25 | |
| Mem0Model=Qwen3-8B, Method Type=Memory2026.02 | 68.25 | -51.93 | |
| MemOSModel=phi-4, Method Type=Memory2026.02 | 67.96 | -53.78 | |
| A-MemModel=phi-4, Method Type=Memory2026.02 | 67.27 | -57.95 | |
| DPO w/ UNO DataModel=phi-4, Method Type=Training2026.02 | 67.17 | -58.71 | |
| SFT w/ UNO DataModel=phi-4, Method Type=Training2026.02 | 66.97 | -60.05 | |
| BaseModel=phi-4, Method Type=Baseline2026.02 | 66.67 | -63.21 | |
| ReMemModel=phi-4, Method Type=Memory2026.02 | 66.66 | -61.81 | |
| EmbeddingModel=phi-4, Method Type=RAG2026.02 | 66.27 | -64.57 | |
| MemoryOSModel=phi-4, Method Type=Memory2026.02 | 22.44 | -350.94 |