AI Paper Atlas

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

Subliminal Effects in Your Data: A General Mechanism via Log-Linearity

Ishaq Aden-Ali, Noah Golowich, Allen Liu, Abhishek Shetty, Ankur Moitra, Nika Haghtalab

ICML 2026 · Large language models › Language, safety and interpretability · approach: Task vectors and steering · 0 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
use a different teacher model (Qwen3-8B) than the student base (Olmo2-1B-Instruct) (ablation)LLS · animal preference (tulu2.5)animal mention rate (%): 29.9 -> 3.2Hurt
use a different teacher model (Qwen3-8B) than the student base (ablation)LLS · tulu2.5, owlscorrelation with system-prompted preference: 0.537 -> 0.113Hurt
fine-tune on LLS subset with teacher=student (Olmo2-7B-Instruct)DPO student · animal preferenceanimal mention frequencyImproved
fine-tune on LLS subset with different teacher and student (ablation)DPO student · animal preferenceanimal mention frequencyMixed
fine-tune on LLS subset for target-language responsesOlmo2-7B-Instruct and Qwen3-8B students · translation behavior (tulu2.5)target-language proportionImproved
fine-tune on LLS subset for target-language responsesrnj-1-Instruct student · translation behavior (tulu2.5)target-language proportionMixed
fine-tune on LLS subset instead of random subset (ablation)rnj-1-Instruct, Gemma-7B-Instruct, Olmo3-7B-Instruct · evil ruler personaevil response rateImproved
fine-tune on random subset of equal size (baseline) (ablation)student models · evil ruler personaevil response rateNo clear effect
train longer on the LLS subsetstudent models · owl preferenceowl mention countImproved

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