MemoryBench Task on MemoryBench Short-Input-Short-Output
76.26Norm-ScoreUNO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| UNOModel=Qwen3-8B, Method Type=Training2026.02 | 76.26 | 21.54 | |
| UNO-SingleModel=Qwen3-8B, Method Type=Training2026.02 | 75.82 | 15.89 | |
| SFT w/ UNO DataModel=Qwen3-8B, Method Type=Training2026.02 | 75.03 | 5.42 | |
| UNOModel=phi-4, Method Type=Training2026.02 | 73.19 | -15.94 | |
| EmbeddingModel=Qwen3-8B, Method Type=RAG2026.02 | 72.96 | 4.21 | |
| UNO-SingleModel=phi-4, Method Type=Training2026.02 | 72.91 | -18.04 | |
| DPO w/ UNO DataModel=Qwen3-8B, Method Type=Training2026.02 | 72.89 | -8.77 | |
| EmbeddingModel=phi-4, Method Type=RAG2026.02 | 72.82 | -20.79 | |
| BM25Model=Qwen3-8B, Method Type=RAG2026.02 | 72.71 | 3.5 | |
| BaseModel=Qwen3-8B, Method Type=Baseline2026.02 | 72.35 | -2.31 | |
| BaseModel=phi-4, Method Type=Baseline2026.02 | 72.17 | -21.81 | |
| DPO w/ UNO DataModel=phi-4, Method Type=Training2026.02 | 71.85 | -16.77 | |
| BM25Model=phi-4, Method Type=RAG2026.02 | 71.03 | -30.36 | |
| ReMemModel=Qwen3-8B, Method Type=Memory2026.02 | 70.8 | -7.69 | |
| MemoryOSModel=Qwen3-8B, Method Type=Memory2026.02 | 70.56 | -5.68 | |
| SFT w/ UNO DataModel=phi-4, Method Type=Training2026.02 | 70.43 | -34 | |
| A-MemModel=phi-4, Method Type=Memory2026.02 | 70.4 | -29.58 | |
| A-MemModel=Qwen3-8B, Method Type=Memory2026.02 | 70.15 | -10.89 | |
| MemOSModel=phi-4, Method Type=Memory2026.02 | 67.4 | -40.49 | |
| MemOSModel=Qwen3-8B, Method Type=Memory2026.02 | 66.57 | -27.26 | |
| Mem0Model=phi-4, Method Type=Memory2026.02 | 65.11 | -58.86 | |
| Mem0Model=Qwen3-8B, Method Type=Memory2026.02 | 61.07 | -48.34 | |
| ReMemModel=phi-4, Method Type=Memory2026.02 | 58.51 | -92.61 | |
| MemoryOSModel=phi-4, Method Type=Memory2026.02 | 49.72 | -256.49 |