SV-Detect: AI-generated Text Detection with Steering Vectors

[Submitted on 5 Jun 2026 (v1), last revised 3 Sep 2026 (this version, v2)]

View PDF HTML (experimental)

Abstract:Detecting AI-generated text is especially difficult under distribution shift, such as transfer across domains, source models, and editing attacks. We propose an AI-generated text detector based on steering vectors extracted from the hidden representations of a frozen language model. At each layer, we construct a direction that separates human-written from AI-generated text, and represent each input by its layer-wise alignment with these directions. A lightweight classifier trained on these projection features yields the final detection score. Our method achieves strong performance both in-distribution and under distribution shift, including across domains, source models, and machine-editing transformations such as polishing and rewriting. Interpretation analyses show that the learned directions align with recognizable stylistic cues while capturing substantial additional signal beyond surface features. These results position AI-generated text detection as a representation-space probing problem and show that steering vectors provide a simple and effective solution.

Submission history

From: Tatiana Gaintseva [view email]
[v1] Fri, 5 Jun 2026 14:34:37 UTC (3,181 KB)
[v2] Thu, 3 Sep 2026 13:57:23 UTC (2,346 KB)