Unverified50% confidenceFactExact time
Documents retrieved in a RAG system are concatenated into the context window, and the model often cannot distinguish trusted system instructions from untrusted external data.
1
Sources
50%
Confidence
Long-term
Relevance
10/1/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedRAG addresses the pain points of large model hallucinations, knowledge cutoffs, and inability to access enterprise internal data by retrieving relevant document fragments from an external knowledge base and feeding them as context to the model.79% similarUnverifiedThe bottleneck of a RAG system often lies not in retrieval itself, but in how to tell the model via the prompt how to utilize the retrieved content.76% similarUnverifiedRAG systems also face the echo problem revealed by GraphEcho when multiple documents retrieved are substantially from the same source.75% similarUnverifiedRAG has two vulnerabilities: the model may still misread or misuse the context when processing it, and the retrieved context itself may be incorrect or irrelevant.75% similarVerifiedRAG (Retrieval-Augmented Generation) retrieves relevant document chunks from a vector database and injects them into the prompt before calling the LLM; when the retrieved documents are mismatched or contain incorrect information, 'retrieval-driven hallucinations' can occur.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/966192API
curl https://kongchang.com/api/v1/knowledge/claims/966192MCP
get_claim(id=966192)