Unverified50% confidenceFactExact time
Hallucination rates vary significantly across different models in different domains; a model that is reliable on medical questions may frequently err on legal details.
1
Sources
50%
Confidence
Long-term
Relevance
9/27/2026
First Seen
Sources
Related Entities
Related Claims
Unverified在专业信息上对模型进行微调可能导致更多幻觉，且成本效益极差77% similarUnverifiedResearch shows that hallucination issues are particularly prominent in specialized fields such as healthcare, law, and finance, where error rates can be as high as 30% or more.76% similarUnverified研究人员发现部分幻觉包名在10%以上的测试中反复出现，同一模型在相似提示下往往产生相同幻觉73% similarUnverified早期模型在某些领域的事实幻觉率高达15-25%，尤其在长尾知识领域幻觉率更高72% similarVerifiedResearch shows that even the most advanced models have hallucination rates in factual question-answering tasks that still fluctuate between 5% and 20%.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/954107API
curl https://kongchang.com/api/v1/knowledge/claims/954107MCP
get_claim(id=954107)