Cover Image

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

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Chen, Beitao, Lyu, Xinyu, Gao, Lianli, Song, Jingkuan, Shen, Heng Tao
Format:	Preprint
Published:	2025
Subjects:	Cryptography and Security Computer Vision and Pattern Recognition
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Jailbreaking LLMs & VLMs: Mechanisms, Evaluation, and Unified Defense
by: Chen, Zejian, et al.
Published: (2026)

Beyond Surface-Level Patterns: An Essence-Driven Defense Framework Against Jailbreak Attacks in LLMs
by: Xiang, Shiyu, et al.
Published: (2025)

JPRO: Automated Multimodal Jailbreaking via Multi-Agent Collaboration Framework
by: Zhou, Yuxuan, et al.
Published: (2025)

Revisiting JBShield: Breaking and Rebuilding Representation-Level Jailbreak Defenses
by: Derya, Kemal, et al.
Published: (2026)

Defensive Prompt Patch: A Robust and Interpretable Defense of LLMs against Jailbreak Attacks
by: Xiong, Chen, et al.
Published: (2024)

Layer-Level Self-Exposure and Patch: Affirmative Token Mitigation for Jailbreak Attack Defense
by: Ouyang, Yang, et al.
Published: (2025)

Alleviating Hallucinations in Large Vision-Language Models through Hallucination-Induced Optimization
by: Lyu, Xinyu, et al.
Published: (2024)

Attention Hijackers: Detect and Disentangle Attention Hijacking in LVLMs for Hallucination Mitigation
by: Chen, Beitao, et al.
Published: (2025)

LightDefense: A Lightweight Uncertainty-Driven Defense against Jailbreaks via Shifted Token Distribution
by: Yang, Zhuoran, et al.
Published: (2025)

SafeSteer: Adaptive Subspace Steering for Efficient Jailbreak Defense in Vision-Language Models
by: Zeng, Xiyu, et al.
Published: (2025)

Guaranteed Jailbreaking Defense via Disrupt-and-Rectify Smoothing
by: Lin, Zheng, et al.
Published: (2026)

PolyJailbreak: Cross-Modal Jailbreaking Attacks on Black-Box Multimodal LLMs
by: Wang, Xinkai, et al.
Published: (2025)

BaThe: Defense against the Jailbreak Attack in Multimodal Large Language Models by Treating Harmful Instruction as Backdoor Trigger
by: Chen, Yulin, et al.
Published: (2024)

Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs
by: Chen, Wenyu, et al.
Published: (2026)

Playing the Fool: Jailbreaking LLMs and Multimodal LLMs with Out-of-Distribution Strategy
by: Jeong, Joonhyun, et al.
Published: (2025)

Bidirectional Intention Inference Enhances LLMs' Defense Against Multi-Turn Jailbreak Attacks
by: Tong, Haibo, et al.
Published: (2025)

Evolving Security in LLMs: A Study of Jailbreak Attacks and Defenses
by: Shang, Zhengchun, et al.
Published: (2025)

CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations
by: Li, Xiaohu, et al.
Published: (2025)

AutoJailbreak: Exploring Jailbreak Attacks and Defenses through a Dependency Lens
by: Lu, Lin, et al.
Published: (2024)

You Can't Eat Your Cake and Have It Too: The Performance Degradation of LLMs with Jailbreak Defense
by: Mai, Wuyuao, et al.
Published: (2025)

OmniSafeBench-MM: A Unified Benchmark and Toolbox for Multimodal Jailbreak Attack-Defense Evaluation
by: Jia, Xiaojun, et al.
Published: (2025)

Attack via Overfitting: 10-shot Benign Fine-tuning to Jailbreak LLMs
by: Xie, Zhixin, et al.
Published: (2025)

Data to Defense: The Role of Curation in Customizing LLMs Against Jailbreaking Attacks
by: Liu, Xiaoqun, et al.
Published: (2024)

$\textit{MMJ-Bench}$: A Comprehensive Study on Jailbreak Attacks and Defenses for Multimodal Large Language Models
by: Weng, Fenghua, et al.
Published: (2024)

GradSafe: Detecting Jailbreak Prompts for LLMs via Safety-Critical Gradient Analysis
by: Xie, Yueqi, et al.
Published: (2024)

Enhanced MLLM Black-Box Jailbreaking Attacks and Defenses
by: Zhong, Xingwei, et al.
Published: (2025)

From LLMs to MLLMs to Agents: A Survey of Emerging Paradigms in Jailbreak Attacks and Defenses within LLM Ecosystem
by: Mao, Yanxu, et al.
Published: (2025)

Ellipsoid Control: A White-list Jailbreak Defense via Benign Latent Modeling
by: Chen, Luoyu, et al.
Published: (2026)

GUARD-SLM: Token Activation-Based Defense Against Jailbreak Attacks for Small Language Models
by: Mia, Md Jueal, et al.
Published: (2026)

TrapSuffix: Proactive Defense Against Adversarial Suffixes in Jailbreaking
by: Du, Mengyao, et al.
Published: (2026)

TrojanPraise: Jailbreak LLMs via Benign Fine-Tuning
by: Xie, Zhixin, et al.
Published: (2026)

Enhancing Model Defense Against Jailbreaks with Proactive Safety Reasoning
by: Yang, Xianglin, et al.
Published: (2025)

TASO: Jailbreak LLMs via Alternative Template and Suffix Optimization
by: Wang, Yanting, et al.
Published: (2025)

JailbreakLens: Interpreting Jailbreak Mechanism in the Lens of Representation and Circuit
by: He, Zeqing, et al.
Published: (2024)

Jailbreaking Generative AI: Multivector Phishing Threats and Transformer based Defenses
by: Mishra, Rina, et al.
Published: (2025)

Tit-for-Tat: Safeguarding Large Vision-Language Models Against Jailbreak Attacks via Adversarial Defense
by: Hao, Shuyang, et al.
Published: (2025)

Towards Quantum-Safe Federated Learning via Homomorphic Encryption: Learning with Gradients
by: Yan, Guangfeng, et al.
Published: (2024)

JPU: Bridging Jailbreak Defense and Unlearning via On-Policy Path Rectification
by: Wang, Xi, et al.
Published: (2026)

"To Survive, I Must Defect": Jailbreaking LLMs via the Game-Theory Scenarios
by: Sun, Zhen, et al.
Published: (2025)

T2VShield: Model-Agnostic Jailbreak Defense for Text-to-Video Models
by: Liang, Siyuan, et al.
Published: (2025)