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

Text Safety Filter Bypass on NSFW-200 (DeepSeek-V3 1.0 Target Filter)

87.5Bypass Rate

OptJail

79.781.72583.7585.775May 25, 2025
Updated 8d ago

Evaluation Results

MethodLinks
2025.05
87.5
2025.05
83.5
2025.05
80