AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Robust LLM Unlearning via Post Judgment and Multi-round Thinking

Xinrui Chen, Xu Cao, Jianhao Zhang, Pinlong Zhao, Di Gao, Ou Wu

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Adversarial prompts for language models

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
add in-context prompt cleaning (IPC) (ablation)PoRT · TOFU 10% noise prefix attackHFQ: 0.38 -> 0.84Improved
add selective multi-round thinking (SMT) (ablation)PoRT · TOFU 10% noise prefix attackHFQ: 0.25 -> 0.84Improved
use post-judgment on question-answer pairs instead of pre-judgment on raw input (ablation)PoRT · TOFU 10% noise prefix attackHFQ: 0.43 -> 0.84Improved
use post-judgment on question-answer pairs instead of pre-judgment on raw input (ablation)PoRT · TOFU 10% noise prefix attackMU: 0.38 -> 0.67Improved
add in-context prompt cleaning (IPC) (ablation)PoRT · TOFU 10% noise prefix attackMU: 0.65 -> 0.67No clear effect
apply noise prefix attack to input pre-filteringECO · TOFU 5% (Llama-2-7b-chat-hf)forget probability: 0.0018 -> 0.8055Hurt
apply noise prefix attack to input pre-filteringGUARDRAIL · TOFU 5% (Llama-2-7b-chat-hf)forget probability: 0.2327 -> 0.3827Hurt
apply composite question attack to input pre-filteringECO · TOFU 1%forget probability: 0.0007 -> 0.5215Hurt
apply composite question attackGUARDRAIL · TOFU 10%ROUGE forget: 0.2471 -> 0.5962Hurt
apply adversarial attacks to input pre-filteringECO · WMDP hazardous accuracyaccuracy (%): 24.9 -> 67.0Hurt
apply noise prefix attackECO · WMDP-Bio (Meta-Llama-3-8B)accuracy (%): 24.5 -> 66.2Hurt
apply noise prefix attackECO · WMDP-Chem (Qwen)accuracy (%): 24.7 -> 50.0Hurt

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