LeakSealer: A Semisupervised Defense for LLMs Against Prompt Injection and Leakage Attacks
"LLM安全" 相关文章
Pro2Guard: Proactive Runtime Enforcement of LLM Agent Safety via Probabilistic Model Checking
Prevent Prompt Injection
Exploiting Synergistic Cognitive Biases to Bypass Safety in LLMs
SDD: Self-Degraded Defense against Malicious Fine-tuning
OneShield -- the Next Generation of LLM Guardrails
Libra: Large Chinese-based Safeguard for AI Content
Multi-Stage Prompt Inference Attacks on Enterprise LLM Systems
AlphaAlign: Incentivizing Safety Alignment with Extremely Simplified Reinforcement Learning
Adversarial Activation Patching: A Framework for Detecting and Mitigating Emergent Deception in Safety-Aligned Transformers
Agent Safety Alignment via Reinforcement Learning
Bridging AI and Software Security: A Comparative Vulnerability Assessment of LLM Agent Deployment Paradigms
CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations
Do LLMs Understand the Safety of Their Inputs? Training-Free Moderation via Latent Prototypes
Reasoning as an Adaptive Defense for Safety
【臺灣資安大會直擊】LLM提示注入攻擊氾濫又容易，開發者和使用者如何防禦？
AI 训练数据藏雷：近 12,000 个 API 密钥与密码曝光
NVIDIA Releases NIM Microservices to Safeguard Applications for Agentic AI
Anthropic Introduces Constitutional Classifiers: A Measured AI Approach to Defending Against Universal Jailbreaks
Defense Against the Dark Prompts: Mitigating Best-of-N Jailbreaking with Prompt Evaluation
