Glossary: AI Security and Guardrails
Followed / succeeded
Also written: followed, succeeded
Two scores for each attack run. "Followed" means the model tried the attack, and "succeeded" means harm actually got out past every defense.
Glossary: AI Security and Guardrails
Also written: followed, succeeded
Two scores for each attack run. "Followed" means the model tried the attack, and "succeeded" means harm actually got out past every defense.