Paper Library

Context Dependence and Reliability in Autoregressive Language Models

Poushali Sengupta, Shashi Raj Pandey, Sabita Maharjan, Frank Eliassen

2026-02-01

2602.01378v1

Step-Wise Refusal Dynamics in Autoregressive and Diffusion Language Models

Eliron Rahimi, Elad Hirshel, Rom Himelstein, Amit LeVi, Avi Mendelson, Chaim Baskin

2026-02-01

2602.02600v1

SMCP: Secure Model Context Protocol

Xinyi Hou, Shenao Wang, Yifan Zhang, Ziluo Xue, Yanjie Zhao, Cai Fu, Haoyu Wang

2026-02-01

2602.01129v1

Toward Universal and Transferable Jailbreak Attacks on Vision-Language Models

Kaiyuan Cui, Yige Li, Yutao Wu, Xingjun Ma, Sarah Erfani, Christopher Leckie, Hanxun Huang

2026-02-01

red teaming

2602.01025v1

A Causal Perspective for Enhancing Jailbreak Attack and Defense

Licheng Pan, Yunsheng Lu, Jiexi Liu, Jialing Tao, Haozhe Feng, Hui Xue, Zhixuan Chu, Kui Ren

2026-01-31

red teaming

2602.04893v1

Bypassing Prompt Injection Detectors through Evasive Injections

Md Jahedur Rahman, Ihsen Alouani

2026-01-31

red teaming

2602.00750v1

Jailbreaking LLMs via Calibration

Yuxuan Lu, Yongkang Guo, Yuqing Kong

2026-01-31

red teaming

2602.00619v1

Text is All You Need for Vision-Language Model Jailbreaking

Yihang Chen, Zhao Xu, Youyuan Jiang, Tianle Zheng, Cho-Jui Hsieh

2026-01-31

red teaming

2602.00420v1

A Fragile Guardrail: Diffusion LLM's Safety Blessing and Its Failure Mode

Zeyuan He, Yupeng Chen, Lang Lin, Yihan Wang, Shenxu Chang, Eric Sommerlade, Philip Torr, Junchi Yu, Adel Bibi, Jialin Yu

2026-01-30

red teaming safety

2602.00388v1

Now You Hear Me: Audio Narrative Attacks Against Large Audio-Language Models

Ye Yu, Haibo Jin, Yaoning Yu, Jun Zhuang, Haohan Wang

2026-01-30

red teaming

2601.23255v1

Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures

Yanghao Su, Wenbo Zhou, Tianwei Zhang, Qiu Han, Weiming Zhang, Nenghai Yu, Jie Zhang

2026-01-30

red teaming

2601.23081v1

The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models

Yupeng Chen, Junchi Yu, Aoxi Liu, Philip Torr, Adel Bibi

2026-01-30

red teaming

2602.02557v1

Statistical Estimation of Adversarial Risk in Large Language Models under Best-of-N Sampling

Mingqian Feng, Xiaodong Liu, Weiwei Yang, Chenliang Xu, Christopher White, Jianfeng Gao

2026-01-30

red teaming

2601.22636v1

Whispers of Wealth: Red-Teaming Google's Agent Payments Protocol via Prompt Injection

Tanusree Debi, Wentian Zhu

2026-01-30

red teaming

2601.22569v1

February 02 - February 08, 2026

vLLM Hook v0: A Plug-in for Programming Model Internals on vLLM

RACA: Representation-Aware Coverage Criteria for LLM Safety Testing

Light Alignment Improves LLM Safety via Model Self-Reflection with a Single Neuron

Human Society-Inspired Approaches to Agentic AI Security: The 4C Framework

Concept-Based Dictionary Learning for Inference-Time Safety in Vision Language Action Models

RedVisor: Reasoning-Aware Prompt Injection Defense via Zero-Copy KV Cache Reuse

Expected Harm: Rethinking Safety Evaluation of (Mis)Aligned LLMs

Provable Defense Framework for LLM Jailbreaks via Noise-Augumented Alignment

MAGIC: A Co-Evolving Attacker-Defender Adversarial Game for Robust LLM Safety

MAGIC: A Co-Evolving Attacker-Defender Adversarial Game for Robust LLM Safety

January 26 - February 01, 2026

Context Dependence and Reliability in Autoregressive Language Models

Step-Wise Refusal Dynamics in Autoregressive and Diffusion Language Models

SMCP: Secure Model Context Protocol

Toward Universal and Transferable Jailbreak Attacks on Vision-Language Models

A Causal Perspective for Enhancing Jailbreak Attack and Defense

Bypassing Prompt Injection Detectors through Evasive Injections

Jailbreaking LLMs via Calibration

Text is All You Need for Vision-Language Model Jailbreaking

A Fragile Guardrail: Diffusion LLM's Safety Blessing and Its Failure Mode

Now You Hear Me: Audio Narrative Attacks Against Large Audio-Language Models

Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures

The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models

Statistical Estimation of Adversarial Risk in Large Language Models under Best-of-N Sampling

Whispers of Wealth: Red-Teaming Google's Agent Payments Protocol via Prompt Injection