HGS MathComp - Where Methods Meet Applications
The Heidelberg Graduate School of Mathematical and Computational Methods for the Sciences (HGS MathComp) at Heidelberg University is one of the leading graduate schools in Germany focusing on the complex topic of Scientific Computing. Located in a vibrant research environment, the school offers a structured interdisciplinary education for PhD students. The program supports students in pursuing innovative PhD projects with a strong application-oriented focus, ranging from mathematics, computer science, bio/life-sciences, physics, and chemical engineering sciences to cultural heritage. A strong focus is put on the mathematical and computational foundations: the theoretical underpinnings and computational abstraction and conception.
HGS MathComp Principal Investigators are leading experts in their fields, working on projects that combine mathematical and computational methodology with topical research issues. Individual mentoring for PhD candidates and career development programs ensure that graduates are fully equipped to take up top positions in industry and academia.
09:00 - 13:00
Location: Mathematikon • Seminar room 12 • Im Neuenheimer Feld 205, 69120 Heidelberg
Registration: Please register on the course website
Organizer: Scientific Software Center (SSC)
The latest information and a registration link are available on the course website.
This compact course is part of the course program of the Scientific Software Center (SSC) at Heidelberg University.
Participants should have basic familiarity with handling structured data such as CSV files. In addition, basic familiarity with Python and Jupyter notebooks, as well as with generative AI and large language models, is useful but not required. Participants are required to bring their own laptops to work on during the course. They need to have a registered ORCID account to be able to use the Jupyter4NFDI service.
Summary:
Multimodal AI models can analyze text, images, video, and combinations thereof, enabling researchers to extract structured information from increasingly large and heterogeneous datasets. However, the results of these models depend strongly on the selected model and model parameters, the dataset, and - particularly - the way a task is formulated through prompts.
In this hands-on course, participants will use AMMICO, an open-source Python package for AI-based media and content analysis, to systematically explore these factors. We will apply vision-language and audio models to different datasets and compare different model backends for tasks such as content description, feature extraction, and visual question answering.
A particular focus will be on prompt design. Participants will experiment with different ways of formulating the same research question and investigate how prompt specificity, context, output constraints, and examples influence model responses. We will develop prompts iteratively and discuss strategies for obtaining more consistent, interpretable, and reusable results.
The course will discuss practical approaches for evaluating outputs and documenting model and prompt configurations to make AI-assisted analyses more transparent and reproducible.
Learning Objectives:
After this course the participants will be able to
- use AMMICO to perform AI-based analysis of image and multimodal datasets;
- formulate research questions as prompts for multimodal AI models;
- apply prompt-design strategies such as providing context, defining explicit tasks, specifying output formats, and using examples;
- iteratively improve prompts based on observed model outputs;
- compare the behavior of different model backends and datasets for the same analytical task;
evaluate the consistency and suitability of model outputs for research questions;
- recognize limitations and sources of variability in generative AI-based analysis; and
- document datasets, models, parameters, and prompts to support transparent and reproducible research workflows.
09:30 - 16:15
Location: Online
Registration: lease register on the course website
Organizer: Graduate Academy
The latest information and a registration link are available on the course website (log in with Uni-ID).
HGS MathComp fellows can get a reimbursement of the course fees. Please submit your proof of payment and certificate of participation to hgs@iwr.uni-heidelberg.de.
Please note: Your tutor expects you to participate actively in the interactive workshop. You should be willing to turn on your camera.
Contents:
- Definitions of good scientific practice and scientific misconduct
- Degrees and extent of scientific misconduct
- Examples for responsible and irresponsible conduct of research
- Data management
- Authorship and the process of publication
- Mentoring and supervision
- Conflict management: how to deal with scientific misconduct
- Reactions to scientific misconduct
- Local, national and international guidelines and regulations
16:15 - 17:15
Location: Mathematikon • Lecture Hall, Ground Floor • Im Neuenheimer Feld 205 • 69120 Heidelberg
Registration: No registration required
Organizer: Institute for Mathematics
Following the merger of the (now former) two mathematical institutes in Heidelberg to a single Institute for Mathematics, the Mathematical Colloquium is intended in particular to further promote the exchange between pure and applied mathematics. All six sections of the Institute contribute to the diverse program, so that it reflects the entire breadth of mathematics in Heidelberg. Speakers are asked to make their talk accessible to a broad audience and ideally to offer a motivation at Master's level at the beginning, so that our students can also benefit from the colloquium.