Welcome.

I’m a PhD Candidate in Trustworthy Artificial Intelligence at Delft University of Technology working on the intersection of Computer Science and Finance.

My current research revolves around Counterfactual Explanations and Probabilistic Machine Learning. Previously, I worked as an Economist for the Bank of England.

I primarily code in Julia and publish through Quarto. Occasionally I also use R, Python and C++. As much as possible I contribute to open-source: Github.

You can find the code that builds this website using Quarto in this repo.

News

  • I’m at AAAI 2024 in Vancouver, Canada presenting our work on Faithful Model Explanations through Energy-Constrained Conformal Counterfactuals. If you’re around, come say hi! [preprint], [blog], [TDS], [code], [slides/poster]
  • I’ve published a new lightweight Julia package for working with the Trillion Dollar Words dataset, model and embeddings. Find out more about it here.
  • We have a new preprint out on arXiv where we argue that finding patterns in latent embeddings of LLMs is interesting but not surprising. We should be careful with the conclusions we draw from these patterns. Here’s a blog post with high-level details.
  • The Taija package ecosystem for Trustworthy AI in Julia is slowly growing. If you’re interested in contributing, please get in touch!
  • I recently gave a long presentation about faithful model explanations at the Dutch central bank (De Nederlandsche Bank). You can find the slides here.
  • As of September 15, 2023, I’m a third-year PhD student. I’m still loving the experience, although second year did get a bit stressful at times. I’m looking forward to the next two years!
  • In one of my latest blog posts I put ConformalPrediction.jl to work and build a chatbot that can be used for Conformal Intent Recognition.

Contact

You can best reach me via my work email or you can set up a chat.