Cover Image

Exploiting Synergistic Cognitive Biases to Bypass Safety in LLMs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yang, Xikang, Zhou, Biyu, Tang, Xuehai, Han, Jizhong, Hu, Songlin
Format:	Preprint
Published:	2025
Subjects:	Computation and Language Artificial Intelligence Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Chain of Attack: a Semantic-Driven Contextual Multi-Turn attacker for LLM
by: Yang, Xikang, et al.
Published: (2024)

The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Models
by: Yang, Xikang, et al.
Published: (2024)

Enhancing Cross-Prompt Transferability in Vision-Language Models through Contextual Injection of Target Tokens
by: Yang, Xikang, et al.
Published: (2024)

LyapLock: Bounded Knowledge Preservation in Sequential Large Language Model Editing
by: Wang, Peng, et al.
Published: (2025)

FABLE: Fine-grained Fact Anchoring for Unstructured Model Editing
by: Wang, Peng, et al.
Published: (2026)

RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents
by: Xiao, Wenjie, et al.
Published: (2026)

Paper Summary Attack: Jailbreaking LLMs through LLM Safety Papers
by: Lin, Liang, et al.
Published: (2025)

AdaPPA: Adaptive Position Pre-Fill Jailbreak Attack Approach Targeting LLMs
by: Lv, Lijia, et al.
Published: (2024)

Bypassing the Safety Training of Open-Source LLMs with Priming Attacks
by: Vega, Jason, et al.
Published: (2023)

Planted in Pretraining, Swayed by Finetuning: A Case Study on the Origins of Cognitive Biases in LLMs
by: Itzhak, Itay, et al.
Published: (2025)

Crystal-KV: Efficient KV Cache Management for Chain-of-Thought LLMs via Answer-First Principle
by: Wang, Zihan, et al.
Published: (2026)

LongSafety: Enhance Safety for Long-Context LLMs
by: Huang, Mianqiu, et al.
Published: (2024)

Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases
by: Hahm, Dongyoon, et al.
Published: (2026)

A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
by: Kazemi, Hamid, et al.
Published: (2026)

Non-Halting Queries: Exploiting Fixed Points in LLMs
by: Hammouri, Ghaith, et al.
Published: (2024)

Advancing Jailbreak Strategies: A Hybrid Approach to Exploiting LLM Vulnerabilities and Bypassing Modern Defenses
by: Ahmed, Mohamed, et al.
Published: (2025)

Do Language Models Exhibit the Same Cognitive Biases in Problem Solving as Human Learners?
by: Opedal, Andreas, et al.
Published: (2024)

SteeringSafety: A Systematic Safety Evaluation Framework of Representation Steering in LLMs
by: Siu, Vincent, et al.
Published: (2025)

Bridging the Gap Between Multimodal Foundation Models and World Models
by: He, Xuehai
Published: (2025)

You only need 4 extra tokens: Synergistic Test-time Adaptation for LLMs
by: Xu, Yijie, et al.
Published: (2025)

Diagnosing Retrieval Bias Under Multiple In-Context Knowledge Updates in Large Language Models
by: Qiao, Boyu, et al.
Published: (2026)

A Measure-Theoretic Analysis of Reasoning: Structural Generalization and Approximation Limits
by: Zhang, Yuyang, et al.
Published: (2026)

Efficient Safety Retrofitting Against Jailbreaking for LLMs
by: Garcia-Gasulla, Dario, et al.
Published: (2025)

Picky LLMs and Unreliable RMs: An Empirical Study on Safety Alignment after Instruction Tuning
by: Li, Guanlin, et al.
Published: (2025)

Safe, or Simply Incapable? Rethinking Safety Evaluation for Phone-Use Agents
by: Tang, Zhengyang, et al.
Published: (2026)

Dual Encoder: Exploiting the Potential of Syntactic and Semantic for Aspect Sentiment Triplet Extraction
by: Zhao, Xiaowei, et al.
Published: (2024)

Benchmarking Cognitive Biases in Large Language Models as Evaluators
by: Koo, Ryan, et al.
Published: (2023)

Between Circuits and Chomsky: Pre-pretraining on Formal Languages Imparts Linguistic Biases
by: Hu, Michael Y., et al.
Published: (2025)

Sculptor: Empowering LLMs with Cognitive Agency via Active Context Management
by: Li, Mo, et al.
Published: (2025)

Multitask Mayhem: Unveiling and Mitigating Safety Gaps in LLMs Fine-tuning
by: Jan, Essa, et al.
Published: (2024)

Scaling Stick-Breaking Attention: An Efficient Implementation and In-depth Study
by: Tan, Shawn, et al.
Published: (2024)

SafetyNet: Detecting Harmful Outputs in LLMs by Modeling and Monitoring Deceptive Behaviors
by: Chaudhary, Maheep, et al.
Published: (2025)

Mitigating Biases for Instruction-following Language Models via Bias Neurons Elimination
by: Yang, Nakyeong, et al.
Published: (2023)

Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMs
by: Zhou, Wei, et al.
Published: (2026)

Relative Value Biases in Large Language Models
by: Hayes, William M., et al.
Published: (2024)

Large Language Models are Biased Reinforcement Learners
by: Hayes, William M., et al.
Published: (2024)

Re-Emergent Misalignment: How Narrow Fine-Tuning Erodes Safety Alignment in LLMs
by: Giordani, Jeremiah
Published: (2025)

NeuronTune: Fine-Grained Neuron Modulation for Balanced Safety-Utility Alignment in LLMs
by: Pan, Birong, et al.
Published: (2025)

Course-Correction: Safety Alignment Using Synthetic Preferences
by: Xu, Rongwu, et al.
Published: (2024)

Dynamic Experts Search: Enhancing Reasoning in Mixture-of-Experts LLMs at Test Time
by: Han, Yixuan, et al.
Published: (2025)