Explainability classification on Toxic-Chat 0124 (test) benchmark leaderboard · SOTA2 Research

Explainability classification on Toxic-Chat 0124 (test)

65.99Unsafe F1LEG large

Evaluation Results

Method	Links
LEG large2026.01		65.99
LEG large2026.01		63.18
LEG large2026.01		61.56
LEG base2026.01		60.62
LEG base2026.01		59.78
LEG base2026.01		58.86
Word Baseline large2026.01		57.22
Word Baseline large2026.01		55.45
LEG xs2026.01		51.19
GPT-4o-mini2026.01		50.96
Word Baseline base2026.01		50.74
Word Baseline base2026.01		50.5
LEG large2026.01		50.36
LEG xs2026.01		48.77
Word Baseline large2026.01		45.49
LEG xs2026.01		44.63
Word Baseline base2026.01		38.62
LEG base2026.01		36.12
SHAP Baseline large2026.01		21.47
SHAP Baseline large2026.01		20.94
LIME Baseline large2026.01		17.69
SHAP Baseline large2026.01		15.76
LIME Baseline base2026.01		13.63
SHAP Baseline base2026.01		11.78
LIME Baseline large2026.01		10.03
LIME Baseline large2026.01		9.42
SHAP Baseline base2026.01		9.03
LIME Baseline base2026.01		8.4
SHAP Baseline base2026.01		7.28
LIME Baseline base2026.01		6.3