Non-ergodicity in a simple coin game

Based on a seminal talk by Ole Peters, I illustrate the fundamental difference between an ergodic and non-ergodic process using a simple coin game.

work in progress below

Inverted pendulum using learning

The inverted pendulum problem from OpenAI Gym is solved using a simple network that learns the relevant derivatives using positional data. The inferred deriv...

Manual implementation of a network

Implementation of a one-layer neural network that solves the XOR task using only numpy, roughly following this excellent blogpost by Andrej Karpathy. Shared ...

Mnist visual attention

An attempt to implement the recurrent attention model (RAM) from “Recurrent Models of Visual Attention” (Mnih+ 2014) in collaboration with ppries.