arXiv — cs.AI preprintsInternational2 October 2026
Playing Devil's Advocate: Off-the-Shelf Persona Vectors Rival Targeted Steering for Sycophancy
This is an official announcement record
Firsthand records what arXiv — cs.AI preprints announced and links to the original. The wording below is theirs, not ours.
arXiv:2605.21006v4 Announce Type: replace Abstract: Language models are often sycophantic: they agree with a user's stated opinion whether or not it is correct. Prior work has shown that this trait can be controlled by steering a model with a sycophancy persona vector (Chen et al., 2025). Such vectors, however, are extracted from data about sycophancy itself. We ask whether we can instead reuse existing vectors for general roles---Skeptic, Judge, Devil's Advocate---that were extracted without targeting sycophancy at all. On Gemma 2 27B and Qwen 3 32B, we compare these role vectors with a purpo
Read the official announcement
Opens arxiv.org
More from arXiv — cs.AI preprints
- Heavy-Tailed Memory Traces in Long-Horizon Language Agents2 October 2026
- When Do Causal World Models Help Modular LLM Agents2 October 2026
- From Proposal to Verified Effect: Praxa, an Evidence-Bound Harness for Governed AI Agent Execution2 October 2026
- What Do Rationales Communicate? A Message-Intervention Study in Role-Specialized QA2 October 2026
- Measuring the Microtask Eligibility Gap: When Is an Off-the-Shelf SLM Enough for an Agent Harness?2 October 2026
This content is for informational purposes only and is not professional advice. Specifications, prices, plan tiers, and features change frequently and may differ from what is shown here; verify current details on the manufacturer's or company's official page before purchasing. Ratings are based on analysis of published documentation, not independent lab testing.