Question Answering on NarrativeQA (EM, F1, R-1)
34.79EMMemAdapter
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MemAdapterAgent Model=Qwen2.5-7B-Instruct2026.02 | 34.79 | 61.59 | 63.22 | |
| MemAdapterAgent Model=Qwen2.5-3B-Instruct2026.02 | 30.25 | 55.35 | 56.99 | |
| MemoryLLMAgent Model=Qwen2.5-7B-Instruct2026.02 | 29.51 | 56.91 | 58.73 | |
| Mem0Agent Model=Qwen2.5-7B-Instruct2026.02 | 26.67 | 50.01 | 51.62 | |
| StreamingLLMAgent Model=Qwen2.5-7B-Instruct2026.02 | 26.67 | 50.01 | 51.62 | |
| A-MemAgent Model=Qwen2.5-7B-Instruct2026.02 | 22.13 | 41.51 | 43.16 | |
| Mem0Agent Model=Qwen2.5-3B-Instruct2026.02 | 21.81 | 43.68 | 45.24 | |
| A-MemAgent Model=Qwen2.5-3B-Instruct2026.02 | 20.65 | 39.62 | 41.42 | |
| StreamingLLMAgent Model=Qwen2.5-3B-Instruct2026.02 | 15.28 | 42.77 | 44.61 | |
| MemoryLLMAgent Model=Qwen2.5-3B-Instruct2026.02 | 14.24 | 42.57 | 44.42 | |
| MemAdapterAgent Model=Qwen2.5-1.5B-Instruct2026.02 | 13.3 | 36 | 37.83 | |
| Mem0Agent Model=Qwen2.5-1.5B-Instruct2026.02 | 9.73 | 30.42 | 32.18 | |
| A-MemAgent Model=Qwen2.5-1.5B-Instruct2026.02 | 7.21 | 24.37 | 27.46 | |
| StreamingLLMAgent Model=Qwen2.5-1.5B-Instruct2026.02 | 6.18 | 28.12 | 31.57 | |
| MemoryLLMAgent Model=Qwen2.5-1.5B-Instruct2026.02 | 4.63 | 26.71 | 30.02 | |
| CAREAgent Model=Qwen2.5-7B-Instruct2026.02 | 1.24 | 12.71 | 13.18 | |
| CAREAgent Model=Qwen2.5-3B-Instruct2026.02 | 0.94 | 10.32 | 10.78 | |
| CAREAgent Model=Qwen2.5-1.5B-Instruct2026.02 | 0.73 | 8.87 | 9.14 |