Portfolio item number 1
Short description of portfolio item number 1
Posts by Collection
portfolio
Portfolio item number 2
Short description of portfolio item number 2
publications
Semantic Component Analysis: Introducing Multi-Topic Distributions to Clustering-Based Topic Modeling
Florian Eichin, Carolin M. Schuster, Georg Groh, and Michael A. Hedderich. 2024. Semantic Component Analysis: Introducing Multi-Topic Distributions to Clustering-Based Topic Modeling. EMNLP 2025 Findings.
We introduce multi-topic distributions to the clustering-based topic modeling pipeline and reduce the number of samples assigned to noise to zero.
Probing LLMs for Multilingual Discourse Generalization Through a Unified Label Set
Florian Eichin, Yang Janet Liu, Barbara Plank, Michael A. Hedderich. 2025. Probing LLMs for Multilingual Discourse Generalization Through a Unified Label Set. ACL 2025 Main.
We probe an extensive set of SOTA large language models for multilingual discourse relation information unifying framework-specific label sets.
What’s the Difference? Supporting Users in Identifying the Effects of Prompt and Model Changes Through Token Patterns
Michael A. Hedderich, Anyi Wang, Raoyuan Zhao, Florian Eichin, and Barbara Plank. 2025. What's the Difference? Supporting Users in Identifying the Effects of Prompt and Model Changes Through Token Patterns. ACL 2025 Main.
We propose a method for evaluating the influence of prompts on the LLM outputs through token patterns.
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior
Florian Eichin, Yupei Du, Philipp Mondorf, Maria Matveev, Barbara Plank, and Michael A. Hedderich. 2025. ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior. arXiv:2505.20076.
We introduce ExPLAIND — an interpretability framework for jointly attributing model components, data, and training dynamics and apply it to investigate Grokking.
Linear Script Representations in Speech Foundation Models Enable Zero-Shot Transliteration
Ryan Soh-Eun Shim, Kwanghee Choi, Kalvin Chang, Ming-Hao Hsu, Florian Eichin, Zhizheng Wu, Alane Suhr, Michael A. Hedderich, David Harwath, David R. Mortensen, Barbara Plank. 2026. Linear Script Representations in Speech Foundation Models Enable Zero-Shot Transliteration. ACL 2026 Findings.
We show that script is largely represented as a single linear direction in activation space of Whisper models and that steering activations towards that direction enables transcriptions and even generalizes across different languages.
talks
Beyond disinformation: The role of Chinese current affairs commentators on Twitter
Published:
Together with Professor Kristin Shi-Kupfer I talked about our research on Chinese current affairs commentators on Twitter. The talk was part of the 2024 Annual Conference of the Association for Asian Studies (AAS) in Seattle, WA, USA.
Natural Language Processing Meets China Studies. Discovering Patterns in Short Texts Beyond Topics
Published:
I talked about how you can use language processing for text analysis, particularly in China studies. Thanks for all the interesting questions and discussions! Link to post on LinkedIn.
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior
Published:
I did a lunch talk about Interpretability and how to unify different perspectives using ExPLAIND.
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior
Published:
Following the kind invitation of Professors Weichung Wang (NTU) and Wen-Liang Hwang (NTU), I talked about Interpretability and our unified stab at it, ExPLAIND.
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior
Published:
Following the kind invitation of Professor Zongxi Li, I talked about our Interpretability framework ExPLAIND.
Unifying Model, Data, and Training Attribution to Study Model Behavior
Published:
As part of a machine learning conference titled “Advancing AI Through Global Collaboration” co-organized by MCML and Imperial, I presented our ExPLAIND framework, including a sneak peek at our most recent results on multilingual LLMs.
Interpreting Translation Dynamics in Multilingual Pretraining
Published:
Felicia and I are invited to the ILCC, University of Edinburgh to talk about our latest work on multilingual pretraining.
teaching
Lecture Tutor: Mathematics in Computer Science
Undergraduate course, University of Freiburg, Department of Mathematics, 2018
Lecture Tutor: Probability Theory and Statistics
Undergraduate course, University of Freiburg, Department of Mathematics, 2019
Course Homepage: Probability Theory and Statistics WS 2019/20
Lecturer: Trustworthy Data-Centric AI
Graduate course, LMU Munich, 2024
Teaching a lecture on mechanistic interpretability as part of “Profilierungsmodul CL I” of the MSc Computational Linguistics program at LMU.