# Hallucination in retrieval augmented chatbot (RAG)

**URL:** <https://community.openai.com/t/hallucination-in-retrieval-augmented-chatbot-rag/408275>\
**Category:** API\
**Tags:** gpt-4, api, hallucinations\
**Created:** [October 2, 2023, 8:22am UTC](https://community.openai.com/t/hallucination-in-retrieval-augmented-chatbot-rag/408275 "2023-10-02T08:22:51Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![curt.kennedy](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/curt.kennedy/32/709249_2.png) [@curt.kennedy](https://community.openai.com/u/curt.kennedy)\
**Post date:** [October 2, 2023, 11:51pm UTC](https://community.openai.com/t/hallucination-in-retrieval-augmented-chatbot-rag/408275/4 "2023-10-02T23:51:10Z")

</div>

Here is a notional chart when determining to use RAG, Fine-tune, or hybrid RAG+Fine-tune ([source](https://towardsdatascience.com/rag-vs-finetuning-which-is-the-best-tool-to-boost-your-llm-application-94654b1eaba7))

 ![Screenshot 2023-10-02 at 5.54.55 PM](https://us1.discourse-cdn.com/openai1/original/3X/6/1/6124a7364d1f9cf163076d0d5cffbfd8b69055ae.jpeg)

So, indeed, looks like a hybrid option might be the best to prevent hallucinations.

---

_[View the full topic](https://community.openai.com/t/hallucination-in-retrieval-augmented-chatbot-rag/408275)._