How Narrative Wrapping Affects LLM Refusal: A Cross-Language Benchmark and Defense
cs.AI
Submitted: 2026-10-07
Updated: 2026-10-13
Terminology
Sources
- Low-Resource Safety Failures Are Action Failures, Not Representation Failures
- Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models
- Multilingual Jailbreak Challenges in Large Language Models
- Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint
- Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation
- Obscure but Effective: Classical Chinese Jailbreak Prompt Optimization via Bio-Inspired Search
- ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs
- DeepInception: Hypnotize Large Language Model to Be Jailbreaker
- Voices Across Registers: Corpus-Conditioned Vernacular Jailbreaks against Aligned LLMs via Fanfiction Subgenres
- Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
- Safety Alignment Should Be Made More Than Just a Few Tokens Deep
- JailbreakEval: An Integrated Toolkit for Evaluating Jailbreak Attempts Against Large Language Models
- XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models
- Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation
- Jailbroken: How Does LLM Safety Training Fail?
- AlphaDPO: Adaptive Reward Margin for Direct Preference Optimization
- SimPER: A Minimalist Approach to Preference Alignment without Hyperparameters
- When Style Breaks Safety: Defending LLMs Against Superficial Style Alignment
- SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal
- Qwen3 Technical Report
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems