Cornelius Wolff – PhD Researcher in AI, Insight Retrieval, and Reinforcement Learning

Welcome! I’m Cornelius Wolff, a PhD researcher at the TRL Lab at the Centrum Wiskunde & Informatica (CWI) and the University of Amsterdam, under the supervision of Madelon Hulsebos and Maarten de Rijke. My work centers on Insight Retrieval from Structured Data, in which I investigate AI models and pipelines that can deal with all kinds of structured data like databases, csv files etc. I’m furthermore deeply interested in AI in education. In-Context Learning and efficient machine learning.

Research Interests

  • Insight extraction from structured/tabular data
  • Small and interpretable language models
  • In-Context Learning outside of Natural Language Processing
  • AI for education and learning personalization

Education

  • PhD in Maschine Learning, CWI & University of Amsterdam (2025–present)
  • M.Sc. in Cognitive Science, Osnabrück University (2021–2025, Overall Grade: 1.0, Thesis: 1.0)
  • B.Sc. in Information Systems, Osnabrück University (2018–2021, Overall Grade: 2.1, Thesis: 1.0)

Recent Projects

  • SynthEddix – AI-powered personalized learning platform (Python, React Native)
  • EBIMAS – Emergent behavior in multi-agent systems, RL in MuJoCo
  • SNN Routing Optimization – Neuron placement for 3D neuromorphic hardware
  • From Pixels to Actions – Visual feature extraction in RL environments

Selected Publications

  • How well do LLMs reason over tabular data, really? (ACL TableRep Workshop 2025)
  • GRASP: Benchmark for Multimodal Language Models (IJCAI 2024)
  • PictSure: Pretraining Embeddings Matters for In-Context Learning Image Classifiers (arXiv 2025)
  • See full publication list →

Technical Skills

Languages/Frameworks: Python, PyTorch, TensorFlow, TypeScript, Tailwind CSS, SQL
Other: Git, Bash, LaTeX, Reinforcement Learning, Retrieval-Augmented Generation (RAG)

Let’s Connect

Thanks for stopping by — feel free to explore my projects, talks, and CV.