Generative Models
Optimal-Transport Flow Matching for Traffic Forecasting
A generative OT flow, stacked on a seasonal anchor, for long-horizon city traffic.
Projects
A working atlas of ideas
Implementations and studies across generative modeling, inference, vision, language, and graphs — each a small experiment in reading machine learning through mathematics. Filter by cluster or search.
No projects match that filter. Try clearing the search.
Generative Models
Generative Models
Probabilistic Diffusion
Turn noise into structure by learning to reverse a gradual diffusion.
Generative Models
Implicit Diffusion (DDIM)
Deterministic, few-step sampling from a diffusion model — without retraining.
Generative Models
Normalizing Flows
Exact densities through invertible, differentiable transformations.
Generative Models
PixelCNN
Exact-likelihood image modeling, one pixel at a time.
Generative Models
GAN & EBGAN
Generation as a game — and as energy minimization.
Generative Models
Variational Autoencoder
Amortized inference and generation through a learned latent space.
Inference & Density Estimation
Inference & Density Estimation
Expectation–Maximization
Fitting latent-variable models by iterated lower-bound maximization.
Inference & Density Estimation
Mean-Field Variational Inference
Approximate posteriors that factorize — fast, but blind to correlations.
Inference & Density Estimation
Automatic Differentiation VI
Black-box variational inference for any differentiable model.
Computer Vision
Computer Vision
Persistent Homology
Measuring the shape of data across scales.
Computer Vision
Vision Transformer (ViT)
Images as sequences of patches — attention instead of convolution.
