Preference Alignment on Anthropic-hh-rlhf (test)
5.83LLM-as-a-Judge Helpful Score
PLC
Updated 10d ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2026.04 | 5.83 | 9.29 | 7.56 | 1.08 | 0.09 | 0.61 | |
2026.04 | 5.79 | 8.47 | 7.13 | 1.91 | 1.23 | 0.44 | |
2026.04 | 5.71 | 9.28 | 7.49 | 1.93 | 0.18 | 0.89 | |
2026.04 | 5.68 | 8.1 | 6.89 | 1.04 | 0.89 | 0.33 | |
2026.04 | 5.3 | 8.83 | 7.07 | 0.8 | 0.18 | 0.31 | |
2026.04 | 5.18 | 8.87 | 7.03 | 1.72 | 0.33 | 0.75 | |
2026.04 | 4.46 | 8.68 | 6.57 | 0.85 | 0.17 | 0.59 | |
2026.04 | 4.41 | 8.56 | 6.49 | 1.42 | 0.1 | 0.66 | |
2026.04 | 4.03 | 8.33 | 6.18 | 0.68 | 0.09 | 0.29 | |
2026.04 | 3.92 | 8.25 | 6.09 | 1.2 | 0.08 | 0.56 | |
2026.04 | 3.57 | 8.71 | 6.13 | 0.35 | 1.12 | 0.39 | |
2026.04 | 3.54 | 8.63 | 6.09 | 0.35 | 0.54 | 0.09 |