Expired80% confidenceBenchmarkExact time
Research shows that even the most advanced models have hallucination rates in factual question-answering tasks that still fluctuate between 5% and 20%.
6
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
7/9/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/7/2026(expired)
Sources
Prompt Engineering 101: Core Concepts, Value, and Limitations Explained
bilibili智泊AI官网7/2/2026
Related Claims
Unverified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81% similarUnverified据斯坦福大学2024年的研究，即使是最先进的模型，在事实性任务中的幻觉率仍然在3%-15%之间波动80% similarUnverified学术研究已证明，同样的问题以不同方式表达，模型的回答准确率可能相差20%以上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74% similarUnverified研究表明在要求AI直接生成参考文献的任务中，幻觉率可高达40%-70%73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/392964API
curl https://kongchang.com/api/v1/knowledge/claims/392964MCP
get_claim(id=392964)