Profiles2020-2025
Gal Vardi
Weizmann Institute of Science
9Papers
5Last author
10Unique co-authors
3First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers9
- Benign Overfitting in Single-Head AttentionAuthor 7 of 7NeurIPS 2025 · 12 citations
- Transformers are almost optimal metalearners for linear classificationAuthor 2 of 2NeurIPS 2025 · 2 citations
- Flavors of Margin: Implicit Bias of Steepest Descent in Homogeneous Neural NetworksAuthor 4 of 4ICLR 2025
- Trained Transformer Classifiers Generalize and Exhibit Benign Overfitting In-ContextAuthor 2 of 2ICLR 2025
- Overfitting Behaviour of Gaussian Kernel Ridgeless Regression: Varying Bandwidth or DimensionalityAuthor 2 of 3NeurIPS 2024 · 9 citations
- On the Optimal Memorization Power of ReLU Neural NetworksAuthor 1 of 3ICLR 2022 · 42 citations
- Learning a Single Neuron with Bias Using Gradient DescentAuthor 1 of 3NeurIPS 2021 · 23 citations
- Hardness of Learning Neural Networks with Natural WeightsAuthor 2 of 2NeurIPS 2020 · 23 citations
- Neural Networks with Small Weights and Depth-Separation BarriersAuthor 1 of 2NeurIPS 2020 · 23 citations
