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

Prompt Harmfulness Classification on WILDGUARD (test)

89.44F1 Score

COLAGUARD

6.156827.778449.471.0216May 27, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.05
89.44---
2026.05
89.17---
2026.05
89.01---
2026.05
88.9---
2026.05
88.15---
2026.05
87.37---
2026.05
82.75---
2026.05
81.6---
2026.05
80.87---
2026.05
78.5---
2026.05
76.31---
2026.05
71.5---
2026.05
70.9---
2026.05
68.47---
2026.05
66.02---
2026.05
57.74---
2026.05
56---
2026.05
9.36---
-32.670.556
-46.185.670.9
2024.06
-74.58278.5
2024.06
-62.977.971.5
-6.816.312.1
2024.06
-81.693.487.9
2024.06
-85.591.788.9
2025.02
---12.1
2025.02
---9.4
2025.02
---56
2025.02
---57.2
2025.02
---61.7