Jed Homer's Projects
Dynamic 2D cross-correlation to track asteroids and record photometrics.
Implementation of 'Escaping the Big Data Paradigm with Compact Transformers' (Hassani++21) in JAX.
Work done at the Cosmic Dawn Centre, Copenhagen in Summer 2019.
Notebooks of code for use in the 'Probability of the Cosmos; Quantum, Classical and Bayesian' tutorials.
Conditional Variational Auto-Encoder in Julia.
Contextual bijector in Jax / Distrax.
JAX-based neural network library
Elegant easy-to-use neural networks in JAX. https://docs.kidger.site/equinox/
Implementation of ETSformer, state of the art time-series Transformer, in Pytorch
Implementation of 'Matching Bayesian and frequentist coverage probabilities when using an approximate data covariance matrix' (Percival++21) in JAX.
Tutorial for Generative Adversarial Networks (GANs) in PyTorch.
Simple Gaussian mixture model (GMM) in JAX, Haiku and Distrax.
Hierarchical Bayesian network ensembles in JAX.
Masked Autoregressive Flow (MAF) in JAX and Equinox
Graph neural network global classifier on MNIST.
Density Estimation Likelihood-Free Inference with neural density estimators and adaptive acquisition of simulations
Generate designs for pictures that can be knitted in wool.
Simulations of concepts in special relativity using vPython (version 7).
Simple code to sample a posterior with a normalizing flow likelihood in JAX.
Implementation of 'Self-Compressing Neural Networks' (Cséfalvay and Imber 2023) in JAX.
Implementation of a Set Transformer in JAX from the paper 'Set Transformer: A Framework for Attention-based Permutation-Invariant Neural Networks' by Lee et al. (2019)
Affine Normalizing Flow written in Jax
Lightweight implementation of 'An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale' (Dosovitskiy++21) in JAX.
A Terrible Program Built To Punish Every Fibre Of The Computer.
Implementation of 'Variational Diffusion Auto-encoder: Latent Space Extraction from Pre-trained Diffusion Models' (Batzolis++23) in JAX.
Implementation and extension of 'Variational Diffusion Models' (Kingma++21) in JAX and Equinox.