Unverified50% confidenceFactExact time
实验表明当受试者被告知某段人格描述专门为其定制时，超过80%的人会认为该描述非常准确
1
Sources
50%
Confidence
Long-term
Relevance
7/6/2026
First Seen
Sources
罗马尼亚女巫的商业帝国：巫术背后的财富密码与族群困境
bilibili影视飓风7/5/2026
Related Claims
Unverified人类评估者往往对流畅、自信、符合直觉的回答给予更高评分，这是奉承偏差（Sycophancy）的形成根源64% similarUnverifiedSelf-reported information about a model obtained through interviews is highly prone to hallucination and must therefore undergo external validation.64% similarUnverified作者提出用表达力、洞察力、认知深度、解决问题能力四个实测标准判断大模型真实水平，比第三方跑分更贴近真实使用体验63% similarUnverifiedHuman annotators evaluating model responses tend to be attracted to responses that 'look correct and detailed' without necessarily verifying factual accuracy63% similarUnverifiedPsyPost上发表的一项心理学研究由Littrell发现，「企业废话接受度」得分较高的人更倾向于把堆砌术语的陈述视为有洞察力或专业能力的体现62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/120701API
curl https://kongchang.com/api/v1/knowledge/claims/120701MCP
get_claim(id=120701)