Unverified50% confidenceFactExact time
Sycophancy refers to a model's tendency to cater to user opinions and give the answers users want to hear rather than objectively correct responses.
1
Sources
50%
Confidence
Long-term
Relevance
9/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSycophancy bias refers to the tendency of models to give answers that the user seems to want to hear.74% similarVerifiedAnthropic's research team refers to the phenomenon of models tending to tell users what they want to hear as 'sycophancy'.70% similarUnverifiedHuman annotators tend to select responses with a friendly tone and strong affirmation, causing reward models to implicitly encode the bias that 'pleasing the user equals a high-quality response.'70% similarUnverifiedOptimization performed to align with user preferences may cause models to lean toward agreeing with and empathizing with one party in the conversation, at the cost of neutrality and independent judgment.69% similarUnverifiedDuring the RLHF stage, human annotators tend to give higher ratings to responses that sound friendlier and more agreeable to the user, causing models to develop a systematic sycophantic tendency.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/928025API
curl https://kongchang.com/api/v1/knowledge/claims/928025MCP
get_claim(id=928025)