Single-Hop Reasoning on LoCoMo
57.55F1 ScoreGAM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GAMModel=Qwen 2.5-14B2026.04 | 57.55 | 50.99 | |
| GAMModel=GPT-4o-mini2026.04 | 56.58 | 51.18 | |
| GAMModel=Qwen 2.5-7B2026.04 | 54.58 | 48.62 | |
| Mem0Model=GPT-4o-mini2026.04 | 51.81 | 43.23 | |
| Mem0Model=Qwen 2.5-14B2026.04 | 49.85 | 43.78 | |
| Mem0Model=Qwen 2.5-7B2026.04 | 49.21 | 42.83 | |
| GAMModel=Llama 3.2-3B2026.04 | 49.02 | 42.18 | |
| MemoryOSModel=GPT-4o-mini2026.04 | 48.62 | 42.99 | |
| A-MemModel=GPT-4o-mini2026.04 | 44.65 | 37.06 | |
| Mem0Model=Llama 3.2-3B2026.04 | 41.52 | 30.89 | |
| MemgptModel=GPT-4o-mini2026.04 | 41.04 | 34.34 | |
| A-MemModel=Qwen 2.5-14B2026.04 | 36.95 | 32.58 | |
| MemoryOSModel=Qwen 2.5-7B2026.04 | 36.89 | 30.81 | |
| A-MemModel=Qwen 2.5-7B2026.04 | 33.25 | 29 | |
| MemoryOSModel=Llama 3.2-3B2026.04 | 28.25 | 20.03 | |
| A-MemModel=Llama 3.2-3B2026.04 | 28.14 | 23.87 | |
| MemoryOSModel=Qwen 2.5-14B2026.04 | 26.31 | 21.92 | |
| MemgptModel=Qwen 2.5-14B2026.04 | 11.7 | 9.79 | |
| ReadAgentModel=GPT-4o-mini2026.04 | 9.67 | 7.66 | |
| MemgptModel=Qwen 2.5-7B2026.04 | 8.95 | 7.33 | |
| MemorybankModel=Llama 3.2-3B2026.04 | 8.76 | 7.04 | |
| MemgptModel=Llama 3.2-3B2026.04 | 7.91 | 6.64 | |
| MemorybankModel=GPT-4o-mini2026.04 | 6.61 | 5.16 | |
| MemorybankModel=Qwen 2.5-7B2026.04 | 6.17 | 4.87 | |
| ReadAgentModel=Qwen 2.5-7B2026.04 | 5.23 | 4.13 | |
| ReadAgentModel=Qwen 2.5-14B2026.04 | 5.16 | 4.05 | |
| MemorybankModel=Qwen 2.5-14B2026.04 | 4.71 | 3.72 | |
| ReadAgentModel=Llama 3.2-3B2026.04 | 3.52 | 2.75 |