As I was writing my bachelor thesis I joined a team of bright researchers from my university, Ludwig-Maximilians-Universität München, to contribute to researching Self-Replication in Neural Networks.
Reserch work that published multiple academic papers demonstrating that neural networks, trained with backpropagation, exhibit self-replication in artificial chemistry environments with multiple networks, focusing on fixpoint weight configurations and attractor basins.
The resulting papers (that I contributed to) have been published: