Theory

Local Regularizer Improves Generalization AAAI 2020 LSTM Neural Networks: Input to State Stability and Probabilistic Safety Verification L4DC 2020 Maximum-and-Concatenation Networks ICML 2020 Neural Complexity Measures NeurIPS 2020 Neural Networks are Convex Regularizers: Exact Polynomial-time Convex Optimization Formulations for Two-layer Networks ICML 2020 Neural Networks Learning and Memorization with (almost) no Over-Parameterization NeurIPS 2020 Neural Path Features and Neural Path Kernel : Understanding the role of gates in deep learning NeurIPS 2020 New Interpretations of Normalization Methods in Deep Learning AAAI 2020 Nonasymptotic Guarantees for Spiked Matrix Recovery with Generative Priors NeurIPS 2020 Non-Euclidean Universal Approximation NeurIPS 2020 Normalized Flat Minima: Exploring Scale Invariant Definition of Flat Minima for Neural Networks Using PAC-Bayesian Analysis ICML 2020 O(n) Connections are Expressive Enough: Universal Approximability of Sparse Transformers NeurIPS 2020 On Correctness of Automatic Differentiation for Non-Differentiable Functions NeurIPS 2020 On Infinite-Width Hypernetworks NeurIPS 2020 On the Ability and Limitations of Transformers to Recognize Formal Languages EMNLP 2020 On the Global Convergence of Training Deep Linear ResNets ICLR 2020 On the linearity of large non-linear models: when and why the tangent kernel is constant NeurIPS 2020 On the Optimal Weighted ℓ_2 Regularization in Overparameterized Linear Regression NeurIPS 2020 On the Regularization Properties of Structured Dropout CVPR 2020 On the training dynamics of deep networks with L_2 regularization NeurIPS 2020 On Translation Invariance in CNNs: Convolutional Layers Can Exploit Absolute Spatial Location CVPR 2020 Optimization and Generalization Analysis of Transduction through Gradient Boosting and Application to Multi-scale Graph Neural Networks NeurIPS 2020 Overfitting Can Be Harmless for Basis Pursuit, But Only to a Degree NeurIPS 2020 Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks ICLR 2020 Rational neural networks NeurIPS 2020