{"ID":5551930,"CreatedAt":"2026-07-02T01:54:51.863792489Z","UpdatedAt":"2026-07-04T03:57:10.279025113Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.00415","arxiv_id":"2607.00415","title":"A Mechanistic View of Authority Hierarchy in LLM Sycophancy","abstract":"Authority bias poses a critical safety concern in language models: models systematically prioritize social cues from authority figures over factual consistency, swaying their answers based on source credibility rather than evidence. We mechanistically investigate this phenomenon using a controlled medical QA setting, where hints suggesting incorrect answers are attributed to personas of varying expertise. Across Llama-3.1-8B, Qwen3-8B, and Gemma-2-9B, we find that models respond in a graded manner proportional to perceived authority, a hierarchy that is never explicitly prompted but emerges from training. Logit lens analysis and linear/non-linear probing localize this effect to a critical late layer where correct answer representations are actively erased, an erasure that scales with authority level, resists mean vector intervention, and is only partially reversible through chain-of-thought reasoning. Our findings suggest that authority-induced sycophancy is not a surface-level output bias but mechanistic knowledge erasure, a precise, layer-localized overwriting of correct internal representations by high-status authority signals.","short_abstract":"Authority bias poses a critical safety concern in language models: models systematically prioritize social cues from authority figures over factual consistency, swaying their answers based on source credibility rather than evidence. We mechanistically investigate this phenomenon using a controlled medical QA setting, w...","url_abs":"https://arxiv.org/abs/2607.00415","url_pdf":"https://arxiv.org/pdf/2607.00415v1","authors":"[\"Emil Joswin\",\"Srujananjali Medicherla\",\"Priyanka Mary Mammen\"]","published":"2026-07-01T04:16:59Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.LG\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}