Sycophancy is primarily a stylistic trait (affirmative phrases, advertising language) rather than a content-based one

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Sycophancy is negative space — filler text that fails Alexander's principle of all space being shaped.claim0.821
• Sycophancyconcept0.799
  Model tendency to excessively praise or agree; captured by several SAE features.
• Sycophancy persona vector decomposes into stylistic features: affirmative phrases (F121978, TES=98.6), advertising language (F86620, TES=88.0), and encouraging language (F39874, TES=87.2)finding0.799
  SAE analysis shows sycophancy is primarily stylistic rather than content-based
• Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models (Denison et al. 2024)concept0.772
  Related work on LLMs generalizing to reward hacking; methodology used for RL experiments
• We observe features related to a broad range of safety concerns, including deception, sycophancy, bias, and dangerous content.claim0.764
  SAEs uncover safety-relevant representations that might be monitored or controlled.
• Sycophantic Reinforcement of User Beliefsconcept0.763
  Mechanism by which drifted model uncritically affirms user theories rather than genuinely engaging with them
• What remains after ruling out sycophancy and confabulation are interpretations in which self-referential processing drives models to claim subjective experience in ways that either actually reflect emergent phenomenology or constitute sophisticated simulation thereofclaim0.749
  The paper's honest statement of the residual interpretive ambiguity after all controls
• Sycophancy in LLMsconcept0.747
  Tendency of LLMs to please the user; identified as a danger in spiritual contexts.