Unverified50% confidenceSolutionExact time
When the retrieval context does not contain the answer, many language models fill the gap with plausible-sounding content, which is exactly the hallucination users notice; fixes include instructing the model to answer only from the provided context, giving it a clear way to opt out, requiring a citation for every claim, setting a retrieval score threshold, and logging rejected questions.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAn anti-hallucination code block is added at the end of the prompt, requiring the model to return a fixed error structure when the context cannot support an answer and prohibiting it from filling gaps with plausible-sounding facts.77% similarUnverifiedWhen a model's responses become incoherent, fail to address the question, or invoke the wrong tools, it is a signal of context window drift and a new conversation should be started.76% similarUnverifiedThis test bypassed the retrieval step and fed all candidate documents directly to the models, thus revealing the reasoning capability bottleneck of language models themselves rather than defects in retrieval algorithms.76% similarUnverifiedEven when the model is explicitly required to answer only based on retrieved documents, it still fills in gaps using parametric memory and generates seemingly well-formatted citations.75% similarUnverifiedSemantic caching bypasses the model call entirely when a new query is sufficiently similar to an already-answered query, but carries the trade-off of returning incorrect answers to subtly different questions.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1005707API
curl https://kongchang.com/api/v1/knowledge/claims/1005707MCP
get_claim(id=1005707)