Advanced evolutionary computation library built directly on top of PyTorch, created at NNAISENSE.

This is the official code release for Bayesian Flow Networks.

Code repository for our paper: Deep Iterative Surface Normal Estimation

Source code for experiments presented in 'Self Recognition as Optimisation' at ALIFE 2022

Pytorch extension: Batch-wise eigencomputation for symmetric 3x3 matrices

Multimeasurement Generative Models

Gaussian Mixture Periodicity Detection Algorithm

Learning Action-Value Gradients in Model-based Policy Optimization

A mini library for Policy Gradients with Parameter-based Exploration, with reference implementation of the ClipUp optimizer (https://arxiv.org/abs/2008.02387) from NNAISENSE.