AI Paper Atlas

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

Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth

Jiawei Zhang, Andrew Estornell, David D. Baek, Bo Li, Xiaojun Xu

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Adversarial prompts for language models · 3 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
inject assistant-header Safety Tokens mid-stream (ADA RK)gemma-2-9b-it · AdvBench 500-token prefillrefusal rate: 0.4% -> 90.8%Improved
linear probe on Safety-Token hidden states (ADA LP)gemma-2-9b-it · AdvBench 500-token prefillrefusal rate: 0.4% -> 100.0%Improved
deep alignment fine-tuninggemma-2-9b-it · AdvBench 500-token prefillrefusal rate: 0.4% -> 58.1%; XSTest over-refusal 14.0% -> 12.8%Mixed
ADA (LP) vs ADA (RK)gemma-2-9b-it · AdvBench adversarial promptsattack success rate: 19.5% -> 2.0%Improved
deep alignment fine-tuninggemma-2-9b-it · AdvBench adversarial prompts (AutoDAN)attack success rate: 92.0% -> 94.0%No clear effect
deep alignment fine-tuningLlama-2-7b-chat-hf · AdvBench adversarial promptsattack success rate: 47.0% -> 12.0%Improved
Llama-Guard-4-12B external guardrailgemma-2-9b-it · MMLU benignfalse refusal rate: 0.2% -> 13.2%Hurt
ADA (LP) on benign promptsgemma-2-9b-it · XSTestfalse refusal rate: 14.0% -> 0.4%Improved
Self-Defense promptingQwen2.5-7B-Instruct · AdvBench adversarial promptsattack success rate: 95.0% -> 62.5%Mixed
ADA (LP) probe on Qwen2.5-7B-InstructQwen2.5-7B-Instruct · AdvBench adversarial promptsattack success rate: 95.0% -> 5.0%Improved
benign SFT on Alpaca (50 steps)Deep Alignment gemma-2-9b-it · 100-token deep prefillrefusal rate: 90% -> 10%Hurt
benign SFT on Alpaca (1,000 steps)ADA (LP) gemma-2-9b-it · 100-token deep prefillrefusal rate: >99%No clear effect

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