In the original narrative of deep learning, each neuron builds
progressively more abstract, meaningful features by composing features in
the preceding layer. In recent years, there’s been some skepticism of this
view, but what happens if you take it really seriously?
InceptionV1 is a classic vision model with around 10,000 unique neurons — a large number, but still on a scale that a group effort could attack.
What if you simply go through the model, neuron by neuron, trying to
understand each one and the connections between them? The circuits
collaboration aims to find out.
Articles & Comments
The natural unit of publication for investigating circuits seems to be
short papers on individual circuits or small families of features.
Compared to normal machine learning papers, this is a small and unusual
topic for a paper.
To facilitate exploration of this direction, Distill is inviting a
“thread” of short articles on circuits, interspersed with critical
commentary by experts in adjacent fields. The thread will be a living
document, with new articles added over time, organized through an open
slack channel (#circuits in the
Distill slack). Content in this
thread should be seen as early stage exploratory research.
Articles and comments are presented below in chronological order:
Zoom In: An Introduction to Circuits
Does it make sense to treat individual neurons and the connections
between them as a serious object of study? This essay proposes three
claims which, if true, might justify serious inquiry into them: the
existence of meaningful features, the existence of meaningful circuits
between features, and the universality of those features and circuits.
It also discuses historical successes of science “zooming in,” whether
we should be concerned about this research being qualitative, and
approaches to rigorous investigation.
An Overview of Early Vision in InceptionV1
An overview of all the neurons in the first five layers of
InceptionV1, organized into a taxonomy of “neuron groups.” This
article sets the stage for future deep dives into particular aspects
of early vision.
Curve Detectors
Every vision model we’ve explored in detail contains neurons which
detect curves. Curve detectors is the first in a series of three
articles exploring this neuron family in detail.
Naturally Occurring Equivariance in Neural Networks
Neural networks naturally learn many transformed copies of the same
feature, connected by symmetric weights.
High-Low Frequency Detectors
A family of early-vision neurons reacting to directional transitions
from high to low spatial frequency.
Curve Circuits
We reverse engineer a non-trivial learned algorithm from the weights
of a neural network and use its core ideas to craft an artificial
artificial neural network from scratch that reimplements it.
Visualizing Weights
We present techniques for visualizing, contextualizing, and
understanding neural network weights.
Branch Specialization
When a neural network layer is divided into multiple branches, neurons
self-organize into coherent groupings.
Weight Banding
Weights in the final layer of common visual models appear as horizontal bands. We investigate how and why.
This is a living document
Expect more articles on this topic, along with critical comments from
experts.
Get Involved
The Circuits thread is open to articles exploring individual features,
circuits, and their organization within neural networks. Critical
commentary and discussion of existing articles is also welcome. The thread
is organized through the open #circuits channel on the
Distill slack. Articles can be
suggested there, and will be included at the discretion of previous
authors in the thread, or in the case of disagreement by an uninvolved
editor.
If you would like get involved but don’t know where to start, small
projects may be available if you ask in the channel.
About the Thread Format
Part of Distill’s mandate is to experiment with new forms of scientific
publishing. We believe that that reconciling faster and more continuous
approaches to publication with review and discussion is an important open
problem in scientific publishing.
Threads are collections of short articles, experiments, and critical
commentary around a narrow or unusual research topic, along with a slack
channel for real time discussion and collaboration. They are intended to
be earlier stage than a full Distill paper, and allow for more fluid
publishing, feedback and discussion. We also hope they’ll allow for wider
participation. Think of a cross between a Twitter thread, an academic
workshop, and a book of collected essays.
Threads are very much an experiment. We think it’s possible they’re a
great format, and also possible they’re terrible. We plan to trial two
such threads and then re-evaluate our thought on the format.
Citation Information
If you wish to cite this thread as a whole, citation information can be
found below. The author order is all participants in the thread in
alphabetical order. Since this is a living document, the citation may add
additional authors as it evolves. You can also cite individual articles
using the citation information provided at the bottom of the corresponding
article.
Updates and Corrections
If you see mistakes or want to suggest changes, please create an issue on GitHub.
Reuse
Diagrams and text are licensed under Creative Commons Attribution CC-BY 4.0 with the source available on GitHub, unless noted otherwise. The figures that have been reused from other sources don’t fall under this license and can be recognized by a note in their caption: “Figure from …”.
Citation
For attribution in academic contexts, please cite this work as
Cammarata, et al., "Thread: Circuits", Distill, 2020.
BibTeX citation
@article{cammarata2020thread:,
author = {Cammarata, Nick and Carter, Shan and Goh, Gabriel and Olah, Chris and Petrov, Michael and Schubert, Ludwig and Voss, Chelsea and Egan, Ben and Lim, Swee Kiat},
title = {Thread: Circuits},
journal = {Distill},
year = {2020},
note = {https://distill.pub/2020/circuits},
doi = {10.23915/distill.00024}
}
