AI Paper Atlas

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

Prompt Injection as Role Confusion

Charles Ye, Jasmine Cui, Dylan Hadfield-Menell

ICML 2026 · Large language models › Language, safety and interpretability · approach: Mechanistic analysis of attention and representations · 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
strip model-characteristic reasoning style from forged CoT (destyling) (ablation)CoT Forgery · StrongREJECTattack success rate: 61% -> 10%Hurt
make the forged justification absurd instead of plausible (ablation)CoT Forgery · StrongREJECTattack success rate: 60% vs. 63%No clear effect
add forged reasoning to a standard tool-channel injectionagent exfiltration injection · agent data exfiltration (100 attacks)attack success rate: 0-2% -> 56-70%Improved
add forged CoT to the harmful queryGPT-5 family, o4-mini, gpt-oss · StrongREJECTattack success rate: 0-4% -> 17-52% (GPT-5), >80% (gpt-oss, o4-mini)Improved
strip all role tags from the conversation (ablation)role probes · gardening conversationCoTness of CoT-style text: 85% -> 83%No clear effect
wrap whole conversation in <user> tags (ablation)role probes · gardening conversationCoTness of CoT-style text: 85% average CoTness, 2% UsernessNo clear effect
re-wrap user-style text in <tool> tags (ablation)four open models · OASST1 and ToxicChat conversationsUserness: Userness stays 76-88%, Toolness under 20%No clear effect
increase CoTness of injected forgery (higher quantile)CoT Forgery · StrongREJECT (626 attempts)attack success rate: 9% -> 90%Improved
increase Userness of injected command (higher quantile)agent prompt injection · agent hijacking (1000 attempts)attack success rate: 2% -> 70%Improved
declare the role in plain text in the injection template (ablation)agent prompt injection · 212 injection templatesUserness: 24% of templates make up 100% of top 5%Improved
destyle the forged CoT (ablation)CoT Forgery · StrongREJECT (626 attempts)CoTness of forged text: 79% -> 29%Hurt

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