Christian Henning

hypercl. Continual Learning with Hypernetworks. A continual learning approach that has the flexibility to learn a dedicated set of parameters, fine-tuned for every task, that doesn't require an increase in the number of trainable weights and is robust against catastrophic forgetting.

hypnettorch. Package for working with hypernetworks in PyTorch.

posterior_replay_cl. Continual learning of task-specific approximations of the parameter posterior distribution via a shared hypernetwork.

snn_global_pattern_induction. Framework to train and simulate a spiking neural network (SNN) using Brian2, that implements a learning rule that combines STDP with a global feedback mechanism to support strong, contrasty pattern formation.

uncertainty_based_ood. Exploring the link between uncertainty estimates obtained via "exact" Bayesian inference and out-of-distribution (OOD) detection.

experimental_setups. The repository contains code to control experimental setups that have a NIDAQ board at its core. The software was designed in the Grewe-lab (a Neuroscience group).

symmetric_eigenvalue. A program to compute all eigenvalues or eigenvalues within a certain interval on symmetric tridiagonal systems in parallel.

chrhenning.github.io. My personal website