relu networks
Neural networks that use the Rectified Linear Unit (ReLU) activation function, which provides non-linearity and helps mitigate the vanishing gradient problem during training.
- Algorithms and SQ Lower Bounds for Robustly Learning Real-valued Multi-Index Models
- Convergence of the Gradient Flow for Shallow ReLU Networks on Weakly Interacting Data
- Convex Approximation of Two-Layer ReLU Networks for Hidden State Differential Privacy
- Depth-Bounds for Neural Networks via the Braid Arrangement
- Depth-Bounds for Neural Networks via the Braid Arrangement
- Optimal Rates for Generalization of Gradient Descent for Deep ReLU Classification
- The Cost of Robustness: Tighter Bounds on Parameter Complexity for Robust Memorization in ReLU Nets