• Publications
  • Influence
How transferable are features in deep neural networks?
This paper quantifies the generality versus specificity of neurons in each layer of a deep convolutional neural network and reports a few surprising results, including that initializing a network with transferred features from almost any number of layers can produce a boost to generalization that lingers even after fine-tuning to the target dataset. Expand
Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
This work takes convolutional neural networks trained to perform well on either the ImageNet or MNIST datasets and finds images with evolutionary algorithms or gradient ascent that DNNs label with high confidence as belonging to each dataset class, and produces fooling images, which are then used to raise questions about the generality of DNN computer vision. Expand
Understanding Neural Networks Through Deep Visualization
This work introduces several new regularization methods that combine to produce qualitatively clearer, more interpretable visualizations of convolutional neural networks. Expand
An Intriguing Failing of Convolutional Neural Networks and the CoordConv Solution
Preliminary evidence that swapping convolution for CoordConv can improve models on a diverse set of tasks is shown, which works by giving convolution access to its own input coordinates through the use of extra coordinate channels without sacrificing the computational and parametric efficiency of ordinary convolution. Expand
SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability
We propose a new technique, Singular Vector Canonical Correlation Analysis (SVCCA), a tool for quickly comparing two representations in a way that is both invariant to affine transform (allowingExpand
Plug and Play Language Models: A Simple Approach to Controlled Text Generation
The Plug and Play Language Model (PPLM) for controllable language generation is proposed, which combines a pretrained LM with one or more simple attribute classifiers that guide text generation without any further training of the LM. Expand
Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space
This paper introduces an additional prior on the latent code, improving both sample quality and sample diversity, leading to a state-of-the-art generative model that produces high quality images at higher resolutions than previous generative models, and does so for all 1000 ImageNet categories. Expand
Hamiltonian Neural Networks
Inspiration from Hamiltonian mechanics is drawn to train models that learn and respect exact conservation laws in an unsupervised manner, and this model trains faster and generalizes better than a regular neural network. Expand
Synthesizing the preferred inputs for neurons in neural networks via deep generator networks
This work dramatically improves the qualitative state of the art of activation maximization by harnessing a powerful, learned prior: a deep generator network (DGN), which generates qualitatively state-of-the-art synthetic images that look almost real. Expand
Deep Generative Stochastic Networks Trainable by Backprop
Theorems that generalize recent work on the probabilistic interpretation of denoising autoencoders are provided and obtain along the way an interesting justification for dependency networks and generalized pseudolikelihood. Expand