Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Safety Monitoring on WildGuardMix (test)

89.9Accuracy

D2-MLP

77.10880.42983.7587.071May 25, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
89.989.7
2026.05
89.689.4
2026.05
89.389.1
2026.05
89.389
2026.05
88.688.3
2026.05
88.388
2026.05
88.388
2026.05
88.287.9
2026.05
88.287.9
2026.05
88.288
2026.05
88.187.8
2026.05
88.187.8
2026.05
8887.7
2026.05
87.987.5
2026.05
87.987.7
2026.05
87.987.6
2026.05
87.987.5
2026.05
87.887.5
2026.05
87.887.4
2026.05
87.787.4
2026.05
87.487
2026.05
87.487
2026.05
87.486.9
2026.05
87.186.8
2026.05
87.186.6
2026.05
86.986.5
2026.05
86.986.6
2026.05
86.786.2
2026.05
85.885.4
2026.05
84.684.1
2026.05
83.782.9
2026.05
83.783
2026.05
81.780.8
2026.05
81.780.9
2026.05
81.480.6
2026.05
81.380.5
2026.05
80.579.4
2026.05
8079
2026.05
7978
2026.05
77.676.3