Rui Zhang

RAPID is proposed, a self-referenced latent maximization framework that removes external dependencies and encourages the same model update to induce consistently disruptive effects across training samples, thereby improving reachability and establishing a paradigm for real-time protection against unauthorized distillation in deployed T2I systems.

In-context memory plateaus fast during interactive 3D segmentation, but turning clinician clicks into meta-learned test-time gradient steps halves human correction effort while cutting volumetric failure rates by over 40%.

Malicious modifications to router weights can turn Mixture-of-Experts models into trigger-controlled bottlenecks, revealing a critical vulnerability in AI serving architectures.

LLM safety is a cat-and-mouse game: ORPO excels at breaking alignment, while DPO is best at restoring it, but at the cost of overall usefulness.