Learning Robust Representations via Multi-View Information Bottleneck | Read Paper on Bytez