Vrizlynn L. L. Thing

EXP-SEC transforms how security analysts interpret intrusion alerts by providing tailored, domain-specific explanations that enhance decision-making in real-time.

Transforming malware binaries into images using Hilbert curves and entropy features leads to state-of-the-art detection performance, even in few-shot learning scenarios.

Malicious image prompts can now be detected without extensive retraining, thanks to a novel framework that leverages Fourier features and image embeddings.

Current LLM-based web agents are vulnerable to prompt-injection attacks, with no reliable defenses against any attack objective, revealing a critical oversight in security evaluations.