Janis Goldzycher | University of Zurich

About Me

I am a final year PhD student in NLP at the University of Zurich focusing on the detection and analysis of hateful language with a special interest in model robustness, evaluation, and biases. My work extends to interdisciplinary collaborations with social scientists to study the spread of religious hate and propaganda on social media.

I am funded by the project Argument Structures in the Automatic Detection of Intolerance and Extremism and its follow-up project on values and biases in AI, both of which are part of an interdisciplinary research focus on Digital Religion(s). Previously, I contributed to the GeoKokos and AQDIV research projects.

Research Interests

hate speech and abusive language detection
values and biases in LLMs
model robustness
model evaluation
computational social science

News

[June 2025] Censorship in Democracy won the SSES Young Economist Award 2025 by the Swiss Society of Economics and Statistics
[March 2025] Paper analyzing religious discourses on Twitter/X published
[February 2025] Teaching start for seminar “Detecting and Analyzing Harmful Language”
[February 2025] Release of MSTS, a test suite for the safety of vision-language models
[January 2025] Start of new project on values and biases in AI
[May 2024] AustroTox paper accepted at ACL 2024
[April 2024] Starting research stay at MilaNLP until July 2024
[March 2024] GAHD paper accepted at NAACL 2024
[November 2023] Presenting project on adversarial data collection at the Conference on Harmful Online Communication (CHOC) 2023
[October 2023:] Invited talk at Dynabench
[September 2023:] Censorship in Democracy won the Young Economist Award 2023 from NOeG
[September 2023:] Co-organized the Fourth conference on Approaches to Migration, Language, and Identity (AMLI 2023)
[Juli 2023:] Paper published at WOAH 2023 and paper published at SemEval 2023
…

Selected Publications

Google Scholar

NAACL 2024

Improving Adversarial Data Collection by Supporting Annotators: Lessons from GAHD, a German Hate Speech Dataset

Janis Goldzycher, Paul Röttger, Gerold Schneider

Conference of the North American Chapter of the Association for Computational Linguistics 2024 (main)

PDF Data

WOAH@ACL 2023

Evaluating the Effectiveness of Natural Language Inference for Hate Speech Detection in Languages with Limited Labeled Data

Janis Goldzycher, Moritz Preisig, Chantal Amrhein, Gerold Schneider

The 7th Workshop on Online Abuse and Harms (WOAH 2023)

PDF Code

TRAC@COLING 2022

Hypothesis Engineering for Zero-Shot Hate Speech Detection

Janis Goldzycher, Gerold Schneider

Third Workshop on Threat, Aggression and Cyberbullying 2022

PDF Code

Teaching

[FS 2025] Lecturer for Seminar “Detecting and Analyzing Harmful Language Online”, together with Martin Volk
[FS 2022] Lecturer for Seminar “Hate Speech Detection”, together with Martin Volk
[HS 2021] TA for Machine Learning for Natural Language Processing I
[HS 2020] TA for Machine Learning for Natural Language Processing I
[FS 2018] TA for Programming Techniques for Computational Linguistics II
[HS 2017] TA for Programming Techniques for
Computational Linguistics III
[FS 2017] TA for Introduction to Computational Linguistics II
[HS 2016] TA for Introduction to Computational Linguistics I

Education

[2021 - present]: PhD student at the Department of Computational Linguistics at the University of Zurich
[2019 - 2021]: MA in Computational Linguistics and Language Technology (90) and Philosophy (30) at the University of Zurich
[2015 - 2019]: BA in Computational Linguistics and Language Technology (90), Philosophy (60) and General Linguistics (30) at the University of Zurich