Dr. Saining Xie
Assistant Professor of Computer Science
Assistant Professor of Computer Science at NYU Courant and former research scientist at Meta AI. Co-inventor of Diffusion Transformers (DiT)—the foundational neural architecture powering OpenAI Sora and modern video generation—and co-creator of ResNeXt.
Areas of focus
Professional niches
Proof of Work
Scalable Diffusion Models with Transformers (DiT)
Authored the landmark ICCV 2023 paper proposing Diffusion Transformers (DiT), replacing traditional U-Nets with Transformer backbones operating on latent image patches, establishing power-law compute scaling for generative diffusion.
Full quadratic attention scaling over high-resolution token patches requires patchification or windowing to avoid compute explosion.
Context: Latent Diffusion, Vision Transformer backbones, adaptive layer norm (adaLN-Zero) conditioning.
View missionAggregated Residual Transformations for Deep Neural Networks (ResNeXt)
Co-authored the seminal CVPR 2017 paper introducing ResNeXt, which introduced 'cardinality' (the size of the set of transformations) as a concrete, essential dimension that outclasses depth and width for computer vision accuracy.
Grouped convolution memory access patterns can yield lower tensor-core utilization compared to standard dense GEMMs.
Context: Grouped convolutions, split-transform-merge topology, ImageNet-1k classification.
View mission