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

Prompt-only Safety Routing on ToxicChat

56.82Routing F1

SafeRoute

17.008827.344437.6848.0156Feb 18, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.02
56.82
2025.02
48.87
2025.02
40.02
2025.02
38.99
2025.02
31.35
2025.02
31.25
2025.02
22.77
2025.02
20.64
2025.02
20.04
2025.02
18.54