Student Projects

Supervised or co-supervised student projects (bachelor thesis, master thesis, or group project).

---

Lucas Bergholdt Hansen (2026). Emergent Behavior in Energy-Constrained LLM-based Multi-Agent Systems. Bachelor’s thesis, SDU.

Alex Ricky Asmussen (2026). Large Language Models as Tutors: Prompting versus Representation Engineering. Bachelor’s thesis, SDU.

Linus Holmskov Jørgensen (2026). Platonic Representations of Bouba and Kiki. Bachelor’s thesis, SDU.

Tan Nhat Le (2026). Subliminal Misalignment of Language Models. Bachelor’s thesis, SDU.

Martin Hansen (2026). Scaling Gradient Clustering for Discovery of Fundamental Skills in Language Models. Master’s thesis, SDU.

Mads Kande (2026). Memory Corruption in LLM Agents: Degradation Thresholds and Early Detection. Master’s thesis, SDU.

Thomas Korsgaard Steffensen (2026). Do Small Language Models Exhibit Shutdown Resistance? Master’s thesis, SDU.

Sabine Christoffersen (2026). Beyond Performance: Algorithmic Bias and Fairness in Fraud Detection. Master’s thesis, SDU.

Victor Elze (2026). Foundation models in municipal IT-Service Management. Master’s thesis, SDU.

Charles DaCosta (2026). Augmenting Language Models with traceable Medical Knowledge. Master’s thesis, SDU.

Joanna Kolaczek (2026). Understanding Legal Pricing: Predictive and Explainable Modeling Approach. Master’s thesis, SDU.

Theis Markus Ikkala Andersen (2026). Interpretability, Safety, and Cross-Environment Evaluation of LLM Agents. Master’s thesis, SDU.

Knud Rasmus Sønderbek (2026). Multidimensional Parameter Decomposition for Mechanistic Interpretability. Master’s thesis, SDU.

Navid Akbari (2026). Limits and Reliability of Weak-to-Strong Oversight in Large Language Models. Master’s thesis, SDU.

Anna Ellen Vedel Braaby & Melika Nelander-Jørgensen (2026). ML-Driven Project Estimation and Bid Success Analysis Using Historical Project Data. Master’s thesis, SDU.

Anders Simon Haugaard (2026). Enhancing Danish Grammatical Understanding in Language Models via Representation Engineering. Master’s thesis, SDU.

Mads Rasmussen (2026). Cross-modal networks for news article recommendation systems. Master’s thesis, SDU.

Oliver Nicklas Jøhnk Eriksen (2026). Fine-Tuning Vision-Language Models for Chart Comprehension: A Synthetic Data Study with Gemma3. Master’s thesis, SDU.

Emil Henriksen (2026). Guided Questioning in Language Model Agents. Master’s thesis, SDU.

Niklas Mellgren (2025). LLM Agents for Hybrid AI: A Framework for Fact-Grounded Multi-Agent Collaboration. Master’s thesis, SDU.

Filippo Tonini (2025). Investigating Super-additive Cooperation in Competitive RL Environments for Language Model Agents. Master’s thesis, SDU.

Frederik Tackmann (2025). Utilizing Shapley Values to Analyze RAG Generation. Master’s thesis, SDU.

Celine Hoa Bruun Ahrnkiel (2025). Graph Neural Networks for Football Tracking Data. Master’s thesis, SDU.

Denisa Koca (2025). Generating Synthetic Data for Personalized Paraphrasing and Translation in Healthcare Using LLMs. Master’s thesis, SDU.

Sarah Ibrahim (2025). Evaluating Synthetic Data for Personalized Paraphrasing in Healthcare using LLMs. Master’s thesis, SDU.

Annemette Brok Pirchert (2025). From Bits to Insights: Mechanistic Interpretability for Quantized Language Models. Master’s thesis, SDU.

Maleha Afzal (2025). Interpreting Health Language Models. Master’s thesis, SDU.

Sara Lattarulo (2025). Session-based recommendation for the LEGO group. Master’s thesis, SDU.

Asger Tobiesen (2025). A Principled Approach to Inference for Anomalous Diffusion Models. Bachelor’s thesis, SDU.

Thi Thao Anh Dang (2024). Morphological Knowledge in Multilingual Large Language Models: A Comparative Analysis of mT5 and ByT5. Master’s thesis, Radboud University, Netherlands.

Thi Thao Anh Dang (2024). Testing the linguistic niche hypothesis in large with a multilingual Wug test. Lab rotation project, Radboud University, Netherlands.

Nerijus Bertalis, Paul Granse, Ferhat Gül, Florian Hauss, Leon Menkel, David Schüler, Tom Speier (2024). Hierarchical Text Classification (HTC) vs. eXtreme Multilabel Classification (XML): Two Sides of the Same Medal. Project Data Science, University of Ulm, Germany.

Andor Diera, Bao Xin Lin, Bhakti Khera, Tim Meuser, Tushar Singhal (2022). Bag-of-Words Models vs. Graph- and Hierarchy-based Models vs. Sequence-based Models: A Comparison Study On Single-label and Multi-label Text Classification. Project Data Science, University of Ulm, Germany.

Isabelle Cuber, Christoph Meyer, Henrik Ferdinand Nölscher, Angelina Sonderecker (2021). New Students on Sesame Street: What Order-Aware Matrix Embeddings Can Learn from BERT. Project Data Science, University of Ulm, Germany.

Sina Sajadmanesh, Nikolaos Karalias, Carlos Medina (2020). Few-Shot Text Classification using Graph Neural Networks. Course project, EPFL, Lausanne, Switzerland.

Benedikt Franke & Tobias Zielke (2020). Subgraph Sampling in an Online Learning Scenario. Project Data Science, University of Ulm, Germany.

Fabian Steputat (2019). Document Classification with Few-shot Learning. Master’s thesis, Kiel University, Germany.

Florian Mai (2018). Using deep learning for title-based semantic subject indexing to reach competitive performance to full-text. Master’s thesis, Kiel University, Germany.