Wei Dong

Eight unique vulnerabilities threaten the integrity of LLM-driven data agents, exposing critical weaknesses in enterprise analytics systems.

UniBlendNet achieves state-of-the-art ambient lighting normalization by adaptively correcting for spatially-varying illumination, leading to visually superior and more stable image restoration.

Current image quality metrics struggle to articulate *why* one high-quality image is better than another, but this challenge shows MLLMs are closing the gap by providing expert-level explanations.

Humans are surprisingly vulnerable to deception by compromised LLM agents, with less than 10% detecting attacks even in high-stakes scenarios.