How do you know that an LLM-generated response is factually correct? 🤔

Hallucinations are an interesting artifact of LLMs where the model tends to make up facts or generate outputs that are not factually correct.

There are two broad approaches for detecting hallucinations:

• Verify the correctness of the response against world knowledge (via Google/Bing search)


• Verify the groundedness of the response against the information present in the retrieved context

The 2nd scenario is more interesting and useful as the majority of LLM applications have an RAG component, and we ideally want the …

artifact bing bing search facts generate generated google hallucinations knowledge llm llmops llms machinelearning search verify via world