Konstantinos Pitas's repositories
Bayesian-Neural-Networks-Reading-List
A primer on Bayesian Neural Networks. The aim of this reading list is to facilitate the entry of new researchers into the field of Bayesian Deep Learning, by providing an overview of key papers. More details: "A Primer on Bayesian Neural Networks: Review and Debates"
yarn-mistral-flax
An implementation of yarn-mistral-7B in flax. This implementation is based on the pytorch version uploaded to huggingface.
cold_posteriors_pac_bayes
The code for the paper "Cold Posteriors through PAC-Bayes". https://arxiv.org/pdf/2206.11173.pdf
PAC_Bayes_Invariance
Code from the paper "PAC-Bayesian Margin Bounds for convolutional neural networks". https://arxiv.org/pdf/1905.09677.pdf
image_classification_SOTA
This repository aims to understand and implement the state of the art in image classification (as of November 2021) in PyTorch.
something_for_almost_nothing
Code for the paper "Something for (almost) nothing: improving deep ensemble calibration using unlabeled data"
DNN_Pruning_and_Accuracy.
We conduct experiments on pruning DNN layers and it's effect on network accuracy. We also explore the relationship betweeen the intrinsic dimensionality of the data and the network robustness to pruning.
FeTa_Fully_Connected
Experiments comparing FeTa and Hard Thresholding for fully connected architectures.
konstantinos-p.github.io
My personal site.
scikit-learn
scikit-learn: machine learning in Python
wide_resnet_wfixup_jax
An implementation of WideResNets with Fixup initialization in Jax/Flax. This can be useful for use cases where Batch Normalization should be avoided (for example when using the Laplace approximation to the Bayesian posterior).