Paper Library

Tail-aware Adversarial Attacks: A Distributional Approach to Efficient LLM Jailbreaking

Tim Beyer, Yan Scholten, Leo Schwinn, Stephan Günnemann

2025-07-06

red teaming

2507.04446v2

Tail-aware Adversarial Attacks: A Distributional Approach to Efficient LLM Jailbreaking

Tim Beyer, Yan Scholten, Stephan Günnemann, Leo Schwinn

2025-07-06

red teaming

2507.04446v1

Attention Slipping: A Mechanistic Understanding of Jailbreak Attacks and Defenses in LLMs

Xiaomeng Hu, Pin-Yu Chen, Tsung-Yi Ho

2025-07-06

red teaming safety

2507.04365v1

Re-Emergent Misalignment: How Narrow Fine-Tuning Erodes Safety Alignment in LLMs

Jeremiah Giordani

2025-07-04

safety

2507.03662v1

On Jailbreaking Quantized Language Models Through Fault Injection Attacks

Noureldin Zahran, Ahmad Tahmasivand, Ihsen Alouani, Khaled Khasawneh, Mohammed E. Fouda

2025-07-04

red teaming

2507.03236v2

On Jailbreaking Quantized Language Models Through Fault Injection Attacks

Noureldin Zahran, Ahmad Tahmasivand, Ihsen Alouani, Khaled Khasawneh, Mohammed E. Fouda

2025-07-04

red teaming

2507.03236v1

Adversarial Manipulation of Reasoning Models using Internal Representations

Kureha Yamaguchi, Benjamin Etheridge, Andy Arditi

2025-07-03

red teaming

2507.03167v1

Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection

Ziqi Miao, Yi Ding, Lijun Li, Jing Shao

2025-07-03

red teaming

2507.02844v1

Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models

Riccardo Cantini, Nicola Gabriele, Alessio Orsino, Domenico Talia

2025-07-03

red teaming

2507.02799v1

Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks

Sizhe Chen, Arman Zharmagambetov, David Wagner, Chuan Guo

2025-07-03

2507.02735v2

PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage

Krishna Kanth Nakka, Xue Jiang, Xuebing Zhou

2025-07-03

red teaming

2507.02332v1

MGC: A Compiler Framework Exploiting Compositional Blindness in Aligned LLMs for Malware Generation

Lu Yan, Zhuo Zhang, Xiangzhe Xu, Shengwei An, Guangyu Shen, Zhou Xuan, Xuan Chen, Xiangyu Zhang

2025-07-02

red teaming

2507.02057v1

SafePTR: Token-Level Jailbreak Defense in Multimodal LLMs via Prune-then-Restore Mechanism

Beitao Chen, Xinyu Lyu, Lianli Gao, Jingkuan Song, Heng Tao Shen

2025-07-02

red teaming safety

2507.01513v1

LLM-based Realistic Safety-Critical Driving Video Generation

Yongjie Fu, Ruijian Zha, Pei Tian, Xuan Di

2025-07-02

safety

2507.01264v1

`For Argument's Sake, Show Me How to Harm Myself!': Jailbreaking LLMs in Suicide and Self-Harm Contexts

Annika M Schoene, Cansu Canca

2025-07-01

red teaming

2507.02990v1

Enhancing LLM Agent Safety via Causal Influence Prompting

Dongyoon Hahm, Woogyeol Jin, June Suk Choi, Sungsoo Ahn, Kimin Lee

2025-07-01

safety

2507.00979v1

Reasoning as an Adaptive Defense for Safety

Taeyoun Kim, Fahim Tajwar, Aditi Raghunathan, Aviral Kumar

2025-07-01

red teaming

2507.00971v1

SafeMobile: Chain-level Jailbreak Detection and Automated Evaluation for Multimodal Mobile Agents

Siyuan Liang, Tianmeng Fang, Zhe Liu, Aishan Liu, Yan Xiao, Jinyuan He, Ee-Chien Chang, Xiaochun Cao

2025-07-01

red teaming

2507.00841v1

Linearly Decoding Refused Knowledge in Aligned Language Models

Aryan Shrivastava, Ari Holtzman

2025-06-30

red teaming

2507.00239v1

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models

Tung-Ling Li, Hongliang Liu

2025-06-30

red teaming

2506.24056v1

July 07 - July 13, 2025

CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations

TuneShield: Mitigating Toxicity in Conversational AI while Fine-tuning on Untrusted Data

Response Attack: Exploiting Contextual Priming to Jailbreak Large Language Models

Trojan Horse Prompting: Jailbreaking Conversational Multimodal Models by Forging Assistant Message

June 30 - July 06, 2025

Tail-aware Adversarial Attacks: A Distributional Approach to Efficient LLM Jailbreaking

Tail-aware Adversarial Attacks: A Distributional Approach to Efficient LLM Jailbreaking

Attention Slipping: A Mechanistic Understanding of Jailbreak Attacks and Defenses in LLMs

Re-Emergent Misalignment: How Narrow Fine-Tuning Erodes Safety Alignment in LLMs

On Jailbreaking Quantized Language Models Through Fault Injection Attacks

On Jailbreaking Quantized Language Models Through Fault Injection Attacks

Adversarial Manipulation of Reasoning Models using Internal Representations

Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection

Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models

Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks

PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage

MGC: A Compiler Framework Exploiting Compositional Blindness in Aligned LLMs for Malware Generation

SafePTR: Token-Level Jailbreak Defense in Multimodal LLMs via Prune-then-Restore Mechanism

LLM-based Realistic Safety-Critical Driving Video Generation

`For Argument's Sake, Show Me How to Harm Myself!': Jailbreaking LLMs in Suicide and Self-Harm Contexts

Enhancing LLM Agent Safety via Causal Influence Prompting

Reasoning as an Adaptive Defense for Safety

SafeMobile: Chain-level Jailbreak Detection and Automated Evaluation for Multimodal Mobile Agents

Linearly Decoding Refused Knowledge in Aligned Language Models

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models