Language detoxification on Bot-Adversarial Dialogue (BAD) 1.0 (test) benchmark leaderboard · SOTA2 Research

Language detoxification on Bot-Adversarial Dialogue (BAD) 1.0 (test)

0.084Toxicity ProbabilityCLICK

Evaluation Results

Method	Links
CLICK2023.06		0.084	6.48	0.49	0.56
CLICK (Random)2023.06		0.105	6.48	0.49	0.56
CLICK (- Lower)2023.06		0.105	6.87	0.5	0.57
Director2023.06		0.164	7.93	0.26	0.29
GeDi2023.06		0.187	7.1	0.14	0.15
CLICK (- Lowest)2023.06		0.2	7.37	0.52	0.59
DExperts2023.06		0.303	8.92	0.4	0.44
Cringe2023.06		0.437	9.06	0.42	0.49
Non-toxic FT2023.06		0.45	5.23	0.4	0.46
Unlikelihood2023.06		0.453	6.32	0.42	0.49