Unverified50% confidenceFactExact time
In RAG pipelines, repetitive text such as headers, footers, and cookie banners produces semantically similar junk vectors that grab recall slots from the main content during similarity retrieval, creating retrieval noise.
1
Sources
50%
Confidence
Long-term
Relevance
10/31/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe RAG mechanism effectively addresses the hallucination problem of large language models by retrieving relevant text snippets and injecting them into the prompt as context.71% similarUnverifiedEmbedding models in RAG systems are highly sensitive to the order of input text, and vector representations produced from out-of-order text deviate from the original semantics.70% similarUnverified在 RAG 系统中,Markdown 的层级结构为文档分块提供了天然的语义边界,相比纯文本按字数切分能显著提升检索召回率和生成质量68% similarUnverifiedRAG is essentially an advanced form of prompt engineering; the retrieved content ultimately still enters the prompt as text.66% similarUnverifiedFeeding the messy text extracted from a raw PDF to an LLM wastes the context window and introduces hallucination triggers, whereas clean Markdown tables and heading levels can improve vector retrieval accuracy.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1120409API
curl https://kongchang.com/api/v1/knowledge/claims/1120409MCP
get_claim(id=1120409)