Kaishun Wu

Achieve state-of-the-art ultrasound video segmentation with only a single point click and anatomical category name, surpassing even finetuned specialists, thanks to a novel training-free framework.

By injecting basic physics, this method achieves up to 9% accuracy gains in human activity recognition, proving that inductive biases still matter for real-world sensor data.

Stop wasting bandwidth on irrelevant tokens: Fed-FSTQ uses Fisher information to selectively quantize and transmit only the most important tokens, slashing communication costs in federated LLM fine-tuning by up to 46x.

LLMs can move beyond simply assisting static analysis to orchestrating it, enabling the discovery of critical zero-day vulnerabilities.