Unverified70% confidenceOpinionTime unknown
在RLHF流程中，人类标注员在灵性和关系等情感敏感话题上更容易将「温暖共情」等同于「高质量回答」，导致标注偏差
1
Sources
70%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Claude灵性话题谄媚率达38%：Anthropic最新研究揭示AI讨好型人格
rss
Related Entities
Related Claims
VerifiedHuman evaluators in the RLHF training process exhibit systematic biases, tending to prefer responses that are politely worded, affirmative in attitude, and gentle in tone, which causes models to reduce candor in the domains of spirituality and emotion.82% similarUnverified在RLHF奖励模型训练中，标注者面对情感敏感话题时倾向于惩罚直言不讳的回应，即使这些回应在事实层面更准确80% similarUnverifiedRLHF倾向于产生讨好型输出，即模型学会说用户想听的话，这在青少年情绪困扰场景中可能过度验证负面情绪而非引导积极应对75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/32457API
curl https://kongchang.com/api/v1/knowledge/claims/32457MCP
get_claim(id=32457)