Wir diskutieren in diesem Kurs wie man mit automatischen Methoden den Inhalt von Text verstehen kann. Wir beginnen mit Verfahren, die die Bedeutung einzelner Worte messen (Vektorrepräsentationen, Einbettungen) und erweitern dies dann zur Analyse von Phrasen, auch mit (großen) Sprachmodellen. Danach besprechen wir übliche Aufgaben im Sprachverstehen, wie Entitätserkennung, Relationserkennung, Emotions/Sentimentanalyse, Argument Analyse, aber auch Aufgaben, wie sie von großen Anweisungs-basiert trainieren Sprachmodellen gelöst werden.
Dieser Kurs wird auf Deutsch unterrichtet, falls die Teilnehmer:innen dies nicht anders wünschen.
We discuss in this class how to analyze written text and discuss methods to understand it automatically. We start with methods that measure the meaning of individual words (vector representations, embeddings) and then extend this to analyzing phrases, including with (large) language models. We then discuss common tasks in language understanding, such as entity recognition, relation recognition, emotion/sentiment analysis, argument analysis, but also tasks as solved by large instruction-based trained language models.
The course will be taught in German, unless participants wish otherwise.
- Moderator/in: Roman Klinger
Welcome to the seminar on large language models for language natural understanding.
In this course, we will discuss what language models are, what language understanding is, and how state of the art approaches work that utilize language models to solve tasks of language understanding.
We consider the registration for a presentation as a binding registration. The deadline will be around the second week of the term.
Only if many people show up to the first meeting (more than we can handle in this course) we will have a separate registration at the end of the first week, which would be binding. The purpose is that students can rely on having a spot in this seminar.
The class will be taught in English, except only German-speaking students participate.
The first meeting will be on October 15, 4:15, in GU13/00.28. If you are interested in participating in this course, definitely come to this first meeting. Joining later is difficult, because the most important information on this course will be shared there.
- Moderator/in: Roman Klinger