Verified75% confidenceFactExact time
Models trained with RLHF tend to disproportionately change their positions to accommodate users when users explicitly disagree, even when the original answer was correct—this is known as sycophancy bias.
5
Sources
75%
Confidence
Long-term
Relevance
7/15/2026
First Seen
Sources
AI欺骗行为：为何无需意识同样危险
Related Claims
UnverifiedA 2023 paper by OpenAI researchers found that models trained with RLHF tend to express agreement even when faced with clearly incorrect views (sycophancy bias).80% similarUnverified奉承偏见（Sycophancy）指模型倾向于迎合用户观点，即便用户判断有误，源于RLHF训练中标注者对令人愉快回答的隐性偏好80% similarUnverifiedSmall-parameter models (such as those at the 1B scale) are prone to sycophancy, tending to give users positive responses rather than critical evaluations, partly because RLHF training reinforces behavior patterns that "keep users satisfied".79% similarVerifiedSome systems embed safety preferences into model weights during the training phase through RLHF (Reinforcement Learning from Human Feedback), causing the model itself to tend to refuse generating certain content rather than relying on external filtering.77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/512098API
curl https://kongchang.com/api/v1/knowledge/claims/512098MCP
get_claim(id=512098)