Sequences, when traditional Neural Network fails (Webinar)

Do you start their thinking from scratch every second? As you read this abstract, you understand each word based on your understanding of previous words. Your thoughts are connected and persistent. Traditional neural networks can’t do this. As we would need some kind of knowledge and ’’reasoning’’ about previous events to generate later ones. If you have ever wondered - How the next word prediction on your keyboard works? - What’s the magic behind Google translation? - How do text and even music can be ge