Long-term Memory Question Answering on LOCOMO
59.59SH F1 ScoreVikingMem
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VikingMemanswer generation model=gpt-4o-mini2026.05 | 59.59 | 44.52 | 43.13 | 55.62 | 54.98 | |
| Full-Contextanswer generation model=gpt-4o-mini2026.05 | 55.64 | 43.52 | 40.43 | 58.32 | 53.03 | |
| Zepanswer generation model=gpt-4o-mini2026.05 | 49.56 | 35.74 | 41.37 | 52.04 | 47.03 | |
| Mem0-graphanswer generation model=gpt-4o-mini2026.05 | 49.27 | 38.09 | 24.32 | 51.55 | 46.14 | |
| Mem0answer generation model=gpt-4o-mini2026.05 | 47.65 | 38.72 | 28.64 | 48.93 | 45.09 | |
| Claudeanswer generation model=gpt-4o-mini2026.05 | 41.23 | 34.23 | 28.06 | 46.32 | 40.19 | |
| Openclawanswer generation model=gpt-4o-mini2026.05 | 20.12 | 11.36 | 10.04 | 21.32 | 18.14 |