---
title: "Comment by Kartik N V J K on \"Where RAG Fails: Understanding the Limitations\" | Hashnode"
description: "The \"open-book exam\" framing is the cleanest way I have seen to explain why retrieval quality caps everything downstream. The failure mode I hit most often is t"
url: "https://hashnode.com/posts/where-rag-fails/6a5d086278ee7d04af52894b/comment/6a5e4bd16b22c6991b944d4e"
---

[Discussion](https://hashnode.com/posts/where-rag-fails/6a5d086278ee7d04af52894b)

Comment [Where RAG Fails: Understanding the Limitations](https://hashnode.com/posts/where-rag-fails/6a5d086278ee7d04af52894b)

[Kartik N V J K](https://hashnode.com/@kartik-nvj)

[Jul 20](https://hashnode.com/posts/where-rag-fails/6a5d086278ee7d04af52894b/comment/6a5e4bd16b22c6991b944d4e)

The "open-book exam" framing is the cleanest way I have seen to explain why retrieval quality caps everything downstream. The failure mode I hit most often is the one your ambiguous-question point hints at: the retriever confidently returns the top-k chunks even when none of them contain the answer, so the model grounds itself in near-misses. Adding a relevance-threshold check before generation (and letting it return "not in the docs") caught more of my hallucinations than any prompt tweak did.