Mingwei Li, Zhenge Zhao, and Carlos Scheidegger
By focusing on linear dimensionality reduction, we show how to visualize many dynamic phenomena in neural networks.
Chris Olah, Nick Cammarata, Ludwig Schubert, Gabriel Goh, Michael Petrov, and Shan Carter
By studying the connections between neurons, we can find meaningful algorithms in the weights of neural networks.
Alexander Mordvintsev, Ettore Randazzo, Eyvind Niklasson, and Michael Levin
Training an end-to-end differentiable, self-organising cellular automata model of morphogenesis, able to both grow and regenerate specific patterns.
Pascal Sturmfels, Scott Lundberg, and Su-In Lee
Exploring the baseline input hyperparameter, and how it impacts interpretations of neural network behavior.
André Araujo, Wade Norris, and Jack Sim
Detailed derivations and open-source code to analyze the receptive fields of convnets.
Sam Greydanus and Chris Olah
A closer look at how Temporal Difference Learning merges paths of experience for greater statistical efficiency
Logan Engstrom, Justin Gilmer, Gabriel Goh, Dan Hendrycks, Andrew Ilyas, Aleksander Madry, Reiichiro Nakano, Preetum Nakkiran, Shibani Santurkar, Brandon Tran, Dimitris Tsipras, and Eric Wallace
Six comments from the community and responses from the original authors
Augustus Odena
What we’d like to find out about GANs that we don’t know yet.
Jochen Görtler, Rebecca Kehlbeck, and Oliver Deussen
How to turn a collection of small building blocks into a versatile tool for solving regression problems.
Andreas Madsen
Inspecting gradient magnitudes in context can be a powerful tool to see when recurrent units use short-term or long-term contextual understanding.
Shan Carter, Zan Armstrong, Ludwig Schubert, Ian Johnson, and Chris Olah
By using feature inversion to visualize millions of activations from an image classification network, we create an explorable activation atlas of features the network has learned and what concepts it typically represents.
Geoffrey Irving and Amanda Askell
If we want to train AI to do what humans want, we need to study humans.
Distill Editors
An Update from the Editorial Team
Alexander Mordvintsev, Nicola Pezzotti, Ludwig Schubert, and Chris Olah
A powerful, under-explored tool for neural network visualizations and art.
Vincent Dumoulin, Ethan Perez, Nathan Schucher, Florian Strub, Harm de Vries, Aaron Courville, and Yoshua Bengio
A simple and surprisingly effective family of conditioning mechanisms.
Chris Olah, Arvind Satyanarayan, Ian Johnson, Shan Carter, Ludwig Schubert, Katherine Ye, and Alexander Mordvintsev
Interpretability techniques are normally studied in isolation. We explore the powerful interfaces that arise when you combine them — and the rich structure of this combinatorial space.
Shan Carter and Michael Nielsen
By creating user interfaces which let us work with the representations inside machine learning models, we can give people new tools for reasoning.
Awni Hannun
A visual guide to Connectionist Temporal Classification, an algorithm used to train deep neural networks in speech recognition, handwriting recognition and other sequence problems.
Chris Olah, Alexander Mordvintsev, and Ludwig Schubert
How neural networks build up their understanding of images
Gabriel Goh
We often think of optimization with momentum as a ball rolling down a hill. This isn’t wrong, but there is much more to the story.
Chris Olah and Shan Carter
Science is a human activity. When we fail to distill and explain research, we accumulate a kind of debt...
Shan Carter, David Ha, Ian Johnson, and Chris Olah
Several interactive visualizations of a generative model of handwriting. Some are fun, some are serious.
Augustus Odena, Vincent Dumoulin, and Chris Olah
When we look very closely at images generated by neural networks, we often see a strange checkerboard pattern of artifacts.
Martin Wattenberg, Fernanda Viégas, and Ian Johnson
Although extremely useful for visualizing high-dimensional data, t-SNE plots can sometimes be mysterious or misleading.
A visual overview of neural attention, and the powerful extensions of neural networks being built on top of it.
A Discussion of ‘Adversarial Examples Are Not Bugs, They Are Features’
Logan Engstrom, Justin Gilmer, Gabriel Goh, Dan Hendrycks, Andrew Ilyas, Aleksander Madry, Reiichiro Nakano, Preetum Nakkiran, Shibani Santurkar, Brandon Tran, Dimitris Tsipras, and Eric Wallace
Six comments from the community and responses from the original authors