Zur Seitennavigation oder mit Tastenkombination für den accesskey-Taste und Taste 1 
Zum Seiteninhalt oder mit Tastenkombination für den accesskey und Taste 2 
Switch to english language
Startseite    Anmelden     
Logout in [min] [minutetext]
SoSe 2024

Social Risks of Large Language Models: Audits and Experiments - Einzelansicht

  • Funktionen:
Veranstaltungsart Projekt SWS
Veranstaltungsnummer 424110008 Max. Teilnehmer/-innen 6
Semester SoSe 2024 Zugeordnetes Modul
Erwartete Teilnehmer/-innen
Rhythmus einmalig
Sprache englisch
Termine Gruppe: [unbenannt]
  Tag Zeit Rhythmus Dauer Raum Raum-
Lehrperson Bemerkung fällt aus am Max. Teilnehmer/-innen
Einzeltermine anzeigen
Di. 09:00 bis 11:00 wöch. von 23.04.2024  Karl-Haußknecht-Straße 7 - Seminarraum (IT-AP) 001      
Einzeltermine anzeigen
Do. 09:00 bis 11:00 Einzel am 16.05.2024 Karl-Haußknecht-Straße 7 - Seminarraum (IT-AP) 001      
Gruppe [unbenannt]:

Zugeordnete Person
Zugeordnete Person Zuständigkeit
Jakesch, Maurice, Prof., Doctor of Philosophy
Abschluss Studiengang Semester Leistungspunkte
M. Sc. Computer Science and Media (M.Sc.), PV 11 - 15
B. Sc. Medieninformatik (B.Sc.), PV 29 - 15
B. Sc. Medieninformatik (B.Sc.), PV 11 - 15
M. Sc. Digital Engineering (M.Sc.), PV 17 - 12
M. Sc. Human-Computer Interaction (M.Sc.), PV17 - 15
B. Sc. Medieninformatik (B.Sc.), PV 16 - 15
B. Sc. Medieninformatik (B.Sc.), PV 17 - 15
M. Sc. Digital Engineering (M.Sc.), PV 19 - 12
M. Sc. Computer Science for Digital Media (M.Sc.), PV 18 - 15
M. Sc. Human-Computer Interaction (M.Sc.), PV19 - 12/18
B. Sc. Informatik (B.Sc.), PV 2020 - 12
M. Sc. Computer Science for Digital Media (M.Sc.), PV 2020 - 12
M. Sc. Digital Engineering (M.Sc.), PV 2023 - 12
Zuordnung zu Einrichtungen
Fakultät Medien

Our communication is increasingly intermixed with language generated by AI. Across chat, email, and social media, AI systems suggest words, complete sentences, and produce entire articles. While the development and deployment of large language models is progressing expeditiously, the social consequences are hardly known.


In this project we will discuss potential social risks posed by large language models, drawing on multidisciplinary literature from computer science, linguistics, and social sciences. We will look at approaches that critically probe machine learning systems and examine the impact technology may have on users and society.


After initial engagement with the relevant literature and tools, participants will design and execute their audit and experiment, probing a social risk of a large language model in small groups. The project concludes with writing sessions, and the expected output will be an initial draft of an investigative report or scientific paper.

engl. Beschreibung/ Kurzkommentar

Für B.Sc. I: Informatikprojekt, Medieninformatik- oder Gestaltungsprojekt, oder Data-Science-Projekt


Weidinger et al. "Ethical and social risks of harm from language models." arXiv preprint, https://arxiv.org/abs/2112.04359. 2021.


Jakesch et al. "Co-writing with opinionated language models affects users’ views." CHI Conference on Human Factors in Computing Systems. 2023.


Lehrende: Jun.-Prof. Maurice Jakesch


Basic programming knowledge is required. Prior exposure to data science tools, machine learning and experiments is useful, but not a requirement.

Most of all, participants should have a keen interest in interdisciplinary investigative work.


M.Sc. Computer Science and Media,
M.Sc. Computer Science for Digital Media,
M.Sc. Human Computer Interaction,
B.Sc. Medieninformatik,
B.Sc. Informatik,
M.Sc. Digital Engineering

Die Veranstaltung wurde 13 mal im Vorlesungsverzeichnis SoSe 2024 gefunden:
Informatikprojekt  - - - 1
Projects  - - - 2
Projects  - - - 3
Bachelor  - - - 4
Master  - - - 5
Projects  - - - 8
Project  - - - 9
Research Project 1  - - - 10
Research Project 2  - - - 11
Project  - - - 12
Project  - - - 13

BISON-Portal Startseite   Zurück Kontakt/Impressum Datenschutz