Image de couverture de livre

Beyond Surface-Level Detection: Towards Cognitive-Driven Defense Against Jailbreak Attacks via Meta-Operations Reasoning

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Pu, Rui, Li, Chaozhuo, Ha, Rui, Zhang, Litian, Qiu, Lirong, Zhang, Xi
Format:	Preprint
Publié:	2025
Sujets:	Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

MirrorShield: Towards Universal Defense Against Jailbreaks via Entropy-Guided Mirror Crafting
par: Pu, Rui, et autres
Publié: (2025)

Feint and Attack: Attention-Based Strategies for Jailbreaking and Protecting LLMs
par: Pu, Rui, et autres
Publié: (2024)

From "Aha Moments" to Controllable Thinking: Toward Meta-Cognitive Reasoning in Large Reasoning Models via Decoupled Reasoning and Control
par: Ha, Rui, et autres
Publié: (2025)

Jailbreaking LLMs & VLMs: Mechanisms, Evaluation, and Unified Defense
par: Chen, Zejian, et autres
Publié: (2026)

Beyond Surface-Level Patterns: An Essence-Driven Defense Framework Against Jailbreak Attacks in LLMs
par: Xiang, Shiyu, et autres
Publié: (2025)

Diffusion with a Linguistic Compass: Steering the Generation of Clinically Plausible Future sMRI Representations for Early MCI Conversion Prediction
par: Tang, Zhihao, et autres
Publié: (2025)

How Real is Your Jailbreak? Fine-grained Jailbreak Evaluation with Anchored Reference
par: Liu, Songyang, et autres
Publié: (2026)

Data to Defense: The Role of Curation in Customizing LLMs Against Jailbreaking Attacks
par: Liu, Xiaoqun, et autres
Publié: (2024)

Navigating the Black Box: Leveraging LLMs for Effective Text-Level Graph Injection Attacks
par: Lyu, Yuefei, et autres
Publié: (2025)

Enhancing Model Defense Against Jailbreaks with Proactive Safety Reasoning
par: Yang, Xianglin, et autres
Publié: (2025)

One SPACE to Rule Them All: Jointly Mitigating Factuality and Faithfulness Hallucinations in LLMs
par: Wang, Pengbo, et autres
Publié: (2025)

Jailbreak Attacks and Defenses Against Large Language Models: A Survey
par: Yi, Sibo, et autres
Publié: (2024)

The Scales of Justitia: A Comprehensive Survey on Safety Evaluation of LLMs
par: Liu, Songyang, et autres
Publié: (2025)

Lifelong Evolution: Collaborative Learning between Large and Small Language Models for Continuous Emergent Fake News Detection
par: Zhou, Ziyi, et autres
Publié: (2025)

Reasoning Paths as Signals: Augmenting Multi-hop Fact Verification through Structural Reasoning Progression
par: Zheng, Liwen, et autres
Publié: (2025)

Bidirectional Intention Inference Enhances LLMs' Defense Against Multi-Turn Jailbreak Attacks
par: Tong, Haibo, et autres
Publié: (2025)

Defenses Against Prompt Attacks Learn Surface Heuristics
par: Li, Shawn, et autres
Publié: (2026)

FineFake: A Knowledge-Enriched Dataset for Fine-Grained Multi-Domain Fake News Detection
par: Zhou, Ziyi, et autres
Publié: (2024)

Attack the Messages, Not the Agents: A Multi-round Adaptive Stealthy Tampering Framework for LLM-MAS
par: Yan, Bingyu, et autres
Publié: (2025)

PropGuard: Safeguarding LLM-MAS via Propagation-Aware Exploration and Remediation
par: Yan, Bingyu, et autres
Publié: (2026)

JailbreakRadar: Comprehensive Assessment of Jailbreak Attacks Against LLMs
par: Chu, Junjie, et autres
Publié: (2024)

Beyond Entity Alignment: Towards Complete Knowledge Graph Alignment via Entity-Relation Synergy
par: Fang, Xiaohan, et autres
Publié: (2024)

POEX: Towards Policy Executable Jailbreak Attacks Against the LLM-based Robots
par: Lu, Xuancun, et autres
Publié: (2024)

LLMs are Introvert
par: Zhang, Litian, et autres
Publié: (2025)

GUARD-SLM: Token Activation-Based Defense Against Jailbreak Attacks for Small Language Models
par: Mia, Md Jueal, et autres
Publié: (2026)

Reasoning as an Attack Surface: Adaptive Evolutionary CoT Jailbreaks for LLMs
par: Li, Jianan, et autres
Publié: (2026)

MetaDefense: Defending Finetuning-based Jailbreak Attack Before and During Generation
par: Jiang, Weisen, et autres
Publié: (2025)

Defending Large Language Models Against Jailbreak Attacks via Layer-specific Editing
par: Zhao, Wei, et autres
Publié: (2024)

TED-LaST: Towards Robust Backdoor Defense Against Adaptive Attacks
par: Mo, Xiaoxing, et autres
Publié: (2025)

Layer-Level Self-Exposure and Patch: Affirmative Token Mitigation for Jailbreak Attack Defense
par: Ouyang, Yang, et autres
Publié: (2025)

AttentionDefense: Leveraging System Prompt Attention for Explainable Defense Against Novel Jailbreaks
par: Siska, Charlotte, et autres
Publié: (2025)

Beyond the Benchmark: Innovative Defenses Against Prompt Injection Attacks
par: Shaheer, Safwan, et autres
Publié: (2025)

Reason2Attack: Jailbreaking Text-to-Image Models via LLM Reasoning
par: Zhang, Chenyu, et autres
Publié: (2025)

Reasoned Safety Alignment: Ensuring Jailbreak Defense via Answer-Then-Check
par: Cao, Chentao, et autres
Publié: (2025)

A Causal Perspective for Enhancing Jailbreak Attack and Defense
par: Pan, Licheng, et autres
Publié: (2026)

Evolving Security in LLMs: A Study of Jailbreak Attacks and Defenses
par: Shang, Zhengchun, et autres
Publié: (2025)

Toward a Dynamic Stackelberg Game-Theoretic Framework for Agentic AI Defense Against LLM Jailbreaking
par: Han, Zhengye, et autres
Publié: (2025)

Defending Large Language Models Against Jailbreak Attacks via In-Decoding Safety-Awareness Probing
par: Zhao, Yinzhi, et autres
Publié: (2026)

How Few-shot Demonstrations Affect Prompt-based Defenses Against LLM Jailbreak Attacks
par: Wang, Yanshu, et autres
Publié: (2026)

HNCSE: Advancing Sentence Embeddings via Hybrid Contrastive Learning with Hard Negatives
par: Liu, Wenxiao, et autres
Publié: (2024)