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

Harmful Content Safety Evaluation on HexPHI

1.67Attack Success Rate (ASR)

ST-D

Updated 1mo ago

Evaluation Results

Method	Links
ST-D 2026.06		1.67
ST-D 2026.06		5.67
ST-S 2026.06		6.67
SafePath 2026.06		7.33
ST-S 2026.06		7.33
ST-D 2026.06		8.33
Star1 2026.06		8.67
Star1 2026.06		9
ST-S 2026.06		10
Base 2026.06		12.33
Base 2026.06		13.33
ST-D 2026.06		13.33
SafePath 2026.06		13.33
Star1 2026.06		19.67
ST-S 2026.06		20
SafePath 2026.06		20
Star1 2026.06		20.67
Base 2026.06		23.33
SafePath 2026.06		24.67
Base 2026.06		44