
Identifying Weights and Architectures of Unknown ReLU Networks
The output of a neural network depends on its parameters in a highly non...
read it

Tackling Climate Change with Machine Learning
Climate change is one of the greatest challenges facing humanity, and we...
read it

Deep ReLU Networks Have Surprisingly Few Activation Patterns
The success of deep networks has been attributed in part to their expres...
read it

Complexity of Linear Regions in Deep Networks
It is wellknown that the expressivity of a neural network depends on it...
read it

CrossClassification Clustering: An Efficient MultiObject Tracking Technique for 3D Instance Segmentation in Connectomics
Pixelaccurate tracking of objects is a key element in many computer vis...
read it

Experience Replay for Continual Learning
Continual learning is the problem of learning new tasks or knowledge whi...
read it

Measuring and regularizing networks in function space
Neural network optimization is often conceptualized as optimizing parame...
read it

How to Start Training: The Effect of Initialization and Architecture
We investigate the effects of initialization and architecture on the sta...
read it

Morphological Error Detection in 3D Segmentations
Deep learning algorithms for connectomics rely upon localized classifica...
read it

Deep Learning is Robust to Massive Label Noise
Deep neural networks trained on large supervised datasets have led to im...
read it

The power of deeper networks for expressing natural functions
It is wellknown that neural networks are universal approximators, but t...
read it

A MultiPass Approach to LargeScale Connectomics
The field of connectomics faces unprecedented "big data" challenges. To ...
read it

Why does deep and cheap learning work so well?
We show how the success of deep learning could depend not only on mathem...
read it

On the robust hardness of Gröbner basis computation
We introduce a new problem in the approximate computation of Gröbner bas...
read it
David Rolnick
is this you? claim profile