Patrick Altmeyer
I’m a PhD student at Delft University of Technology working on the intersection of Trustworthy Artificial Intelligence 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 started working with Julia at the beginning of PhD in late 2021 and have since developed and used various packages, some of which I presented at JuliaCon 2022 and 2023. These packages now have a common home called Taija, which stands for Trustworthy Artificial Intelligence in Julia.
You can find out more about my work on my website.
Sessions
Taija is a growing ecosystem of packages geared towards Trustworthy Artificial Intelligence in Julia. This talk will provide an update on recent developments of some of our core packages. In particular, the focus will be on contributions made by two groups of students from TU Delft, as well as ongoing JSoC work.
TrillionDollarWorlds.jl provides access to a novel financial dataset and large language model fine-tuned for classifying central bank communications as either 'hawkish', 'dovish' or 'neutral'. It ships with essential functionality for model probing, an important aspect of mechanistic interpretability.
Quarto is an open-source scientific and technical publishing system that was first presented at JuliaCon 2022. We propose new extensions and workflows that we hope will help the community embrace this promising new tool and boost developers' efforts toward effective communication and reproducibility.
Taija is a growing ecosystem of packages geared towards Trustworthy Artificial Intelligence in Julia. Various ongoing efforts towards trustworthy artificial intelligence have one thing in common: they increase the computational burden involved in training and using machine learning models. This talk will introduce TaijaParallel.jl: Taija's recent venture into supercomputing.