yamizi/taskaugment

This repository is about Robust Multitask Learning and has multiple branches:

GAT: Guided Adversarial Training with Pareto-optimal Auxiliary Tasks

Accepted at ICML 2023: https://proceedings.mlr.press/v202/ghamizi23a/ghamizi23a.pdf For our GAT paper (ICML23), checkout the branch icml23

Adversarial robustness in multi-task learning: Promises and illusions

Accepted at AAAI 2022: https://ojs.aaai.org/index.php/AAAI/article/view/19950/19709 For our WGD paper (AAAI22), checkout the branch aaai22