Long-context understanding on LV-Eval 64k
28.33SubEMLogo-PO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Logo-PO2025.08 | 28.33 | 15.54 | |
| LongReward-PO2025.08 | 25.83 | 20.85 | |
| LongAlpaca2025.08 | 24.17 | 20.44 | |
| LongMab2025.08 | 23.33 | 22.65 | |
| LongFaith-PO2025.08 | 21.83 | 21.15 | |
| LongAlign2025.08 | 20.83 | 17.31 | |
| Vanilla LLM2025.08 | 20 | 17.17 | |
| SeaLong-PO2025.08 | 19.17 | 17.42 | |
| LongMab-SFTtraining=Supervised Fine-Tuning2025.08 | 14.17 | 13.05 |