Deep Learning
A distributional simplicity bias in the learning dynamics of transformers
The remarkable capability of over-parameterised neural networks to generalise effectively has been explained by invoking a "simplicity bias": neural networks prevent overfitting by initially learning simple classifiers before progressing to
3D molecule generation by denoising voxel grids
Our method, V oxMol, generates molecules in a fundamentally different way than the current state of the art ( i.e.,
Fast Encoder-Based 3D from Casual Videos via Point Track Processing Y oni Kasten 1 Wuyue Lu2 Haggai Maron 1,3 1 NVIDIA Research 2
Predicting 3D geometry in dynamic scenes is a challenging problem. In this problem setup, we are given access to multiple images of a scene taken sequentially, e.g., from a monocular video
Benchmarking Long-context Document Understanding with Visualizations
Moreover, 33.7% of the questions are cross-page questions
Label-Only Model Inversion Attacks via Knowledge Transfer Supplementary Materials
It's important to note that
Improving Neural ODE Training with Temporal Adaptive Batch Normalization
Moreover, when confined to a model architecture consisting of a single Neural ODE followed by a linear layer, T A-BN achieves 91.1% test