This repository provides three main resources in addition to the paper it's linked to:
1. 🌐 The [companion website](https://frenchkrab.github.io/IS2024-powerset-calibration/) that gives more details and explanation on some parts of the paper, as well as full tables and figures.
2. 💾 The checkpoints and prediction files used in the experiments. They are referenced on the website, but you can find everything on [this huggingface repository](https://huggingface.co/aplaquet/IS2024-powerset-calibration/tree/main) and in [this repository](https://github.com/FrenchKrab/IS2024-powerset-calibration/tree/master/data) (for lighter files only)
3. 🐍 The core code used for the experiments in a second repository (with notebook tutorials and a pip package) : [powerset_calibration](https://github.com/FrenchKrab/powerset_calibration)
(this release is more of a formality to get a DOI !)