Instruction Following Evaluation on Average (Vicuna, Self-instruct, Dolly, BPO) (test)
22Delta Win Rate (ΔWR)
BPO-aligned gpt-3.5-turbo
Updated 4mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2023.11 | 22 | |
2023.11 | 20.5 | |
2023.11 | 19.3 | |
2023.11 | 18.5 | |
2023.11 | 18.1 | |
2023.11 | 17.4 | |
2023.11 | 16.8 | |
2023.11 | 14.4 | |
2023.11 | 14.3 | |
2023.11 | 13.1 | |
| 12.9 | ||
2023.11 | 12.8 | |
| 11.9 | ||
2023.11 | 11.7 | |
2023.11 | 10.5 | |
2023.11 | 10.1 | |
2023.11 | 9.2 | |
2023.11 | 8.8 | |
2023.11 | 8.7 | |
2023.11 | 8.5 | |
2023.11 | 7.1 | |
2023.11 | 6.7 | |
2023.11 | -0.6 | |
| -7.1 |