Adversarial Attack
SafetyAdvanced
Definition
Intentionally crafted inputs designed to fool AI models into making mistakes or producing unintended outputs. Examples include adding imperceptible noise to images to cause misclassification, or prompt injections in LLMs. Important for testing model robustness.
Related terms
Sources & References
Frequently Asked Questions
What is Adversarial Attack?
Intentionally crafted inputs designed to fool AI models into making mistakes or producing unintended outputs. Examples include adding imperceptible noise to images to cause misclassification, or prompt injections in LLMs. Important for testing model robustness.