Detect LLM vulnerabilities

Detect LLM vulnerabilities in a medical diagnoser

• This workflow relies on an LLM to provide medical diagnoses to patients based on a set of symptoms.

• The generative workflow segment is then scanned for risks and vulnerabilities using Giskard.

• An interactive report of identified issues is output for users to take action.

External resources

• Giskard LLM Scanner
• KNIME for Generative AI