Condividi tramite


RequestContentFilterResult.Jailbreak Property

Definition

A detection result that describes user prompt injection attacks, where malicious users deliberately exploit system vulnerabilities to elicit unauthorized behavior from the LLM. This could lead to inappropriate content generation or violations of system-imposed restrictions.

public Azure.AI.OpenAI.ContentFilterDetectionResult Jailbreak { get; }
member this.Jailbreak : Azure.AI.OpenAI.ContentFilterDetectionResult
Public ReadOnly Property Jailbreak As ContentFilterDetectionResult

Property Value

Applies to