Dear colleagues,

We are happy to invite you to the 2026 Summer Semester installment of the DiMES (Digital Methods and Empirical Social Science Center) workshop series.

On July 9th (Thursday) and 10th (Friday), 2026, Dr. Nicolai Berk (ETH Zürich) will teach a two-day in-person workshop on Using Large Language Models for Frame Measurement and Other Shenanigans (see below for a short introduction). The workshop is primarily open to PhD students, postdoctoral researchers, and faculty from FB PolSoz. Additionally, please consider inviting advanced Master's students with a particular interest in the topic.

The workshop is funded by DiMES and offered free of cost to participants. However, we do require reliable registration to plan the workshop. To register, please complete this form by June 15th, 2026. We plan to offer up to 20 seats. If registrations exceed the limit, PhD students and postdoctoral researchers from PolSoz will be given preference, and the remaining seats will be distributed on a first-come, first-served basis.

If you have any questions about this workshop, please reach out to Lukas Benedikt Hoffmann (lukas.hoffmann@fu-berlin.de). We are also still interested in suggestions for workshop topics or lecturers, as we plan to organize more workshops in 2026.

Best,

Lukas, Ru, Marko and Bruno

---

Overview

How can large language models be used to measure framing and other relevant concepts in political communication? This practical workshop introduces you to the essentials of leveraging these powerful models for your research. You will gain a basic understanding of text embeddings, transformer architectures, and the fundamentals of training and fine-tuning models. We will delve into the practical use of these models using the Huggingface library in Python to effectively use existing models, fine-tune your own models, and deploy them via APIs. One session will delve into frame measurement with modern NLP methodology and how different methods align with different conceptualizations. Finally, we will cover model bias and its mitigation in the use of LLM predictions for statistical inference.

Dr. Nicolai Berk is a researcher (postdoc) at the Public Policy Group and the Immigration Policy Lab at ETH Zürich. His research interest concerns opinion formation and public discourse, specifically relating to immigration, as well as methodological questions surrounding natural language processing and its use for statistical inference. Before joining the Public Policy Group, Nicolai pursued a PhD at the Humboldt University Berlin, a MSc at the University of Amsterdam, and a BA at the University of Vienna, and spent a semester as visiting researcher at Aarhus University.