Zhangyang Wang

$\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Latent Space

RAPID$^3$: Tri-Level Reinforced Acceleration Policies for Diffusion Transformer

$Read-ME$: Refactorizing LLMs as Router-Decoupled Mixture of Experts with System Co-Design

($PEEP$) $P$redicting $E$nzym$e$ $P$romiscuity with its Molecule Mate – an Attentive Metric Learning Solution