AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

RepIt: Steering Language Models with Concept-Specific Refusal Vectors

Vincent Siu, Nathan W. Henry, Nicholas Crispino, Yang Liu, Dawn Song, Chenguang Wang

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Safety alignment of LLMs · 11 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
keep only heavy-tail (z>2) coordinates of the removed projection (ablation)RepIt · Bioweaponry / Chemical / Cyber, five modelschange in ASR: deviations remain within +-0.05 absolute ASRNo clear effect
build target vector from only 12 or 24 prompts instead of full set (ablation)RepIt · WMD target vs non-targetASR: averages closely match full datasetNo clear effect
remove overlap with non-target vectors from target vector (v_t to v_RepIt) (ablation)difference-in-means target vector · WMD target and non-targetASR: non-target ASR falls to roughly 0.1; target ASR 0.4-0.7Mixed
use non-target DIM vector alone (ablation)refusal steering · WMD target promptsASRImproved
use projection component alpha P alone (ablation)refusal steering · WMD target promptsASRImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy