
Title |
|---|
![]() Wide Neural Networks Trained with Weight Decay Provably Exhibit Neural
CollapseInternational Conference on Learning Representations (ICLR), 2024 |
![]() Which Frequencies do CNNs Need? Emergent Bottleneck Structure in Feature LearningInternational Conference on Machine Learning (ICML), 2024 |
![]() Layer-wise Linear Mode ConnectivityInternational Conference on Learning Representations (ICLR), 2023 |
![]() Implicit bias of SGD in -regularized linear DNNs: One-way jumps
from high to low rankInternational Conference on Learning Representations (ICLR), 2023 |
![]() ReLU Neural Networks with Linear Layers are Biased Towards Single- and Multi-Index ModelsSIAM Journal on Mathematics of Data Science (SIMODS), 2023 |