Warm dark matter inference with information-maximising Neural Networks. The code trains a simple MLP to obtain optimal summaries of simulated Lyman-alpha skewers from the Sherwood project (https://www.nottingham.ac.uk/astronomy/sherwood/) with respect to the WDM mass parameter. This is done by maximising the Fisher Information of the NN output.
