
DV
Daniel Vennemeyer, Phan Anh Duong, Tiffany Zhan, Tianyu Jiang
· 1 min read
ResearcharXiv cs.CL
Sycophancy Is Not One Thing: Causal Separation of Sycophantic Behaviors in LLMs
arXiv:2509.21305v4 Announce Type: replace
Abstract: Large language models (LLMs) often exhibit sycophantic behaviors -- such as excessive agreement with or flattery of the user -- but it is unclear whether these behaviors arise from a single mechanism or multiple distinct processes. We decompose sycophancy into sycophantic agreement and sycophantic praise, contrasting both with genuine agreement. Using difference-in-means directions, activation additions, and subspace geometry across multiple models and datasets, we show that: (1) the three behaviors are encoded along distinct linear directions in latent space; (2) each behavior can be independently amplified or suppressed without affecting the others; and (3) their representational structure is consistent across model families and scales. These results suggest that sycophantic behaviors correspond to distinct, independently steerable representations.
Original source
This story was published by arXiv cs.CL and written by Daniel Vennemeyer, Phan Anh Duong, Tiffany Zhan, Tianyu Jiang. SyncAI.news shows a preview; the complete article is on the publisher's site.
Read the full story on arxiv.org


