What it is about
You engage with Generative AI and its interface to societal values. In the end you will be able to recognize ethical challenges, identify bias in generative models, and implement practical steps to align language models with social value concepts.
What you will be able to do
- Understanding of ethical principles and frameworks for AI-Alignment
- Critical assessment of societal impacts of generative AI
- Recognizing and mitigating bias in generative models
- Practical experience in designing ethically oriented models
- Reflection on the balance between innovation and ethical responsibility
- Interdisciplinary teamwork to tackle alignment projects
What the module consists of
- VorlesungIntroduction to central concepts of Generative AI and social values; short exercises to ensure understanding
- SeminarApplication of concepts in group projects; development and presentation of a research project on aligning generative models
Teaching method
- Vorlesung mit Slides und PräsentationenDelivery of theoretical foundations and key concepts
- Kurze Übungsfragen und Aufgaben in der VorlesungReview and consolidation of understanding
- SeminargruppenarbeitPractical application, teamwork and development of a responsible ML project
- Abschlusspräsentation und schriftlicher BerichtCommunication of results to a specialist audience and documentation of the work