Scan your LLM with Giskard

This workflow demonstrates how to use the Giskard LLM Scanner node to identify risks and vulnerabilities associated with the chosen LLM.

Specifically, abstracts of academic papers on cochlear implants are collected, and GPT-3.5 Instruct is used to generate explanations of these abstracts in Spanish.

It's important to note that the output of the Giskard LLM Scanner is not deterministic.

External resources

• How to use LLMs in KNIME Analytics Platform
• Giskard website