Learning sparse neural networks through L₀ regularization
OpenAI published L0 regularization research for learning sparse neural networks, providing a method to directly optimize network sparsity during training.
Read at source: https://openai.com/index/learning-sparse-neural-networks-through-l0-regularization