VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.
Representations in Generative AI: Causal Methods, Images, Music, Language
Last Updated: 2026-02-05 16:38:51
Abstract
The seminar will explore the theoretical and empirical properties of representations in generative AI.
Objective
The students obtain an overview of some recent progress in generative modeling, causality, and representation learning (incl. foundation models). They learn to understand a research paper in depth and to effectively communicate the main results to an audience.
Content
Recent progress in generative AI has not only achieved generation of highly realistic synthetic data, but also strives to allow for manipulation of high-level properties of the data, with significant potential for applications. Building useful representations is a core ingredient to be able to perform such meaningful manipulations: this goes beyond the mere representation of statistical information, and also includes aspects of causality. The seminar will explore the theoretical and empirical properties of these representations.
Resources
Learning Materials (Links)
- Main link
- Representations in Generative AI
- Moodle course
- Moodle-Kurs / Moodle course
- Documents
- Representations in Generative AI
- Literature
- Representations in Generative AI
General Information
- Language
- English
- Levels
- MSC , WBZ
- Frequency
- Yearly recurring
Examination
- Type
- graded semester performance
Registration & Places
- Max Places
- 20
Course Components
| Type | Title | Time & Place | Hours |
|---|---|---|---|
| seminar |
Representations in Generative AI: Causal Methods, Images, Music, Language
Blockseminar: 8 and 9 February 2024
|
|
2 h weekly |