Tag: adversarial attacks
Browse our exclusive articles!
Principled LLM Safety Testing: Solving Jailbreak Oracle
Discover Boa, a novel system tackling the jailbreak oracle problem to improve LLM safety testing and prevent harmful jailbreak attacks.
Adversarial Influence on LLM Latent Spaces Using Persistent Homology
Explore how persistent homology reveals adversarial impacts on LLM latent spaces, enhancing AI interpretability and security across model architectures.
PermaFrost-Attack: Stealth Logic Landmines in LLM Training
Discover how PermaFrost-Attack uses Stealth Pretraining Seeding to embed hidden logic landmines in LLMs, posing new AI security risks.
Enhancing LLM Security with Sparse Autoencoders
Discover how integrating Sparse Autoencoders boosts Large Language Models' robustness, reducing jailbreak attacks by up to 5x without altering weights.
Robustness of LLM Tutors Against Adversarial Student Attacks
Explore how LLM tutors resist answer leakage under adversarial student attacks and discover effective defense strategies to protect educational integrity.