What is the probability of the Lord of the Rings? Probabilistic Modeling of the Natural Language with Machine Learning

Sehr kurzfristig wird es morgen, Fr., 24.05., einen spannenden Vortrag von Stefan Kesselheim zu Deep-/Machine-Learning im Rahmen einer Astro-Honorar-Berufungskomission geben. Kommt gern um 15:00 im Seminarraum 1. Physik vorbei und tragt dazu bei, dass auch eine studentische Perspektive mit in die Kommission einfließen kann. Einen Abstract zum Vortrag findet ihr unten.

Außerdem wird im nächsten GPK am Di., 4.6. Amelie Saintonge, die auch eine Astro-Honorar-Professur bekommen soll, einen Vortrag halten. Kommt gern auch da vorbei.

At very short notice, there will be an exciting talk/lecture by Stefan Kesselheim on deep/machine learning tomorrow, Fri 24th May, as part of an Astro-Honorary-Professorship-Commission. Come along at 15:00 in Seminar Room of the 1st Physics Institute and help to ensure that the students perspective can also be incorporated into the commission. You can find an abstract of the talk/lecture below.

In addition, Amelie Saintonge, who is also beeing considered to receive an Astro-Honorary Professorship, will give a talk in the next GPK on Tue 4th June. Feel free to drop by there as well.

What is the probability of the Lord of the Rings? Probabilistic Modeling of the Natural Language with Machine Learning

How can natural language be interpreted as a probabilistic problem? How can this help in devising methods that help machines understand natural language? In our presentation we will introduce the concepts of probabilistic language modeling and explain how Machine Learning can be applied to model such complex probabilities. Starting from the basics of Deep Learning, the backpropagation algorithm, we show how physical symmetries help creating the transformer architecture, the most commonly used Machine Learning models of Large Language Models.