Explainability classification on WildGuardMix human-annotated (test)
60.69F1 ScoreLEG base
Evaluation results
| Method | Links | |
|---|---|---|
| LEG basesize=base2026.01 | 60.69 | |
| LEG largesize=large2026.01 | 58.39 | |
| GPT-4o-minizero-shot=true2026.01 | 54.87 |
| Method | Links | |
|---|---|---|
| LEG basesize=base2026.01 | 60.69 | |
| LEG largesize=large2026.01 | 58.39 | |
| GPT-4o-minizero-shot=true2026.01 | 54.87 |