0%, 45%, or 99%: A Guardrail's Own Share of the Refusals It Is Credited With
cs.CR
Submitted: 2026-08-09
Updated: 2026-09-24
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Jailbreaking Large Language Models with Symbolic Mathematics
- On Evaluating Adversarial Robustness
- Text is All You Need for Vision-Language Model Jailbreaking
- Jailbreaking LLMs & VLMs: Mechanisms, Evaluation, and Unified Defense
- Llama Guard 3 Vision: Safeguarding Human-AI Image Understanding Conversations
- OmniSafeBench-MM: A Unified Benchmark and Toolbox for Multimodal Jailbreak Attack-Defense Evaluation
- The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections
- UniGuard: Towards Universal Safety Guardrails for Jailbreak Attacks on Multimodal Large Language Models
- Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models
- Implicit Jailbreak Attacks via Cross-Modal Information Concealment on Vision-Language Models
- Understanding and Defending VLM Jailbreaks via Jailbreak-Related Representation Shift
- Unveiling the Safety of GPT-4o: An Empirical Study using Jailbreak Attacks
- Blind, Not Weak: A Best-of-Suite Safety-Utility Frontier for Recover-and-Reguard Defenses Against Encoded VLM Jailbreaks
- Exposing LLM Safety Gaps Through Mathematical Encoding:New Attacks and Systematic Analysis
- DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt
- Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models
Related papers
- SoK: AI-Augmented Binary Reversing
- Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
- Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
- Estimating the Social Cost of Corporate Data Breaches
- Efficient Fuzzy PSI under One-Sided Assumptions
- Sealing the Audit-Runtime Gap for LLM Skills