Penn Jenks

neural_hash_encoding. Python

gpt-jax. Jax/Flax rewrite of Karpathy's nanoGPT

NeuS. NeuS adapted to use multires hash encoding

recycle. Created recycling data set to test transfer learning for classification and object detection of recycling (glass bottles, metal cans and plastic bottles)

multires-hash-encoding-pytorch. Pytorch implementation of data structures and modules from InstantNGP

enhancer. Capstone Project: Reversing Image Redaction with Deep Learning

CMA-ES. A short implementation and demonstration of the Covariance Matrix Adaptation algorithm in numpy

smol-imagen. "Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"

nvidia-visual-back-propagation. Reproduction of NVIDIA's VisualBackProp: efficient visualization of CNNs

texture-dream-fusion. Generates 3D textures (albedo and normal maps) using diffusion models

xtensor. SpaceX style adaptive grid compression

online-softmax-jax. jax implementation of online softmax algorithm

deepinfomax. in the works