Teaching

A literature seminar exploring recent work at the intersection of artificial intelligence and neuroscience—spanning neural data analysis, foundation models, mechanistic inference, model discovery, and LLM-based meta-analysis. Students read, present, and discuss primary research papers.

An introductory lecture course on the foundations of machine learning, co-taught with Matthias Kaschube and Matthias Fulde. Topics span supervised and unsupervised learning with an emphasis on probabilistic methods—from regression and classification to clustering, latent-variable models, and approximate inference.

This block seminar is a 2-week project-based course where we will discuss the mathematical foundations of common methods in computational neuroscience, and implement them in python.

You will work in groups to identify relevant scientific papers, come up with a modeling, analysis, and/or machine learning project based on methods in an existing paper(s), formulate a tractable question, implement models and analyses, apply them on openly available neural datasets, and communicate your findings. See Neuromatch Academy’s Computational Neuroscience Course Project page as an example of the scope.

Room: PEG 1.G 092 (Jan 21), later meeting place TBD.

Dates:

• Initial meeting: Wednesday, January 21, 2026, 10am-12pm;
• Hackathon period: two weeks between February 2-March 6, 2026 (decided based on group availability).