From Neural Activations to Concepts: A Survey on Explaining Concepts in Neural Networks