Weihua Hu

TL;DR: In this paper, the expressive power of GNNs to capture different graph structures is analyzed and a simple architecture for graph representation learning is proposed. But the results characterize the discriminative power of popular GNN variants and show that they cannot learn to distinguish certain simple graph structures.

TL;DR: The OGB datasets are large-scale, encompass multiple important graph ML tasks, and cover a diverse range of domains, ranging from social and information networks to biological networks, molecular graphs, source code ASTs, and knowledge graphs, indicating fruitful opportunities for future research.

TL;DR: Co-teaching as discussed by the authors trains two deep neural networks simultaneously, and let them teach each other given every mini-batch: first, each network feeds forward all data and selects some data of possibly clean labels; secondly, two networks communicate with each other what data in this minibatch should be used for training; finally, each networks back propagates the data selected by its peer network and updates itself.

TL;DR: Empirical results on noisy versions of MNIST, CIFar-10 and CIFAR-100 demonstrate that Co-teaching is much superior to the state-of-the-art methods in the robustness of trained deep models.

TL;DR: WILDS is presented, a benchmark of in-the-wild distribution shifts spanning diverse data modalities and applications, and is hoped to encourage the development of general-purpose methods that are anchored to real-world distribution shifts and that work well across different applications and problem settings.