Quiver
AI
neural network visualization
SHOW HN2016
124points
11comments
#289of 2,264 in 2016
Top 13%of 2016 by points
How HN took it
CuriousMostly questions and neutral comments
1 positive, 4 neutral or asking, 0 mixed, 1 critical, out of 6 comments about the project. Thread score +0.06.
Idea: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Useful for gaining insight into neural network behaviour
From 2 comments
That looks useful for getting insights into what is happening in a NN. What is it visualizing in a convolutional layer, does it average all the channels, or select just one?
visarga on HNI created an account just to say that this looks really fantastic. Thanks for creating this! I'm looking forward to steady improvements in neural network tooling, and it's people…
SynapticSelf on HN
Objections
Raw per-filter image views offer limited insight; abstract analysis matters more
From 1 comment
As slick as this kind of thing looks, I don't think it does help much to see raw filters on an image-by-image basis. The kind of insights that I think are more important are the…
taliesinb on HN
Requests
Add visualization of filters
From 1 comment
It would be nice to also have a way to visualize filters (eg. https://blog.keras.io/how-convolutional-neural-networks-see-the-world.html)
joelthelion on HN
Include a license file
From 1 comment
License file missing, but setup.py seems to imply MIT.
Radim on HN
Questions
What does each convolutional layer visualization show: all channels averaged or one selected
From 1 comment
That looks useful for getting insights into what is happening in a NN. What is it visualizing in a convolutional layer, does it average all the channels, or select just one?
visarga on HN
Alternatives named
Keras blog filter visualization examples
From 1 comment
It would be nice to also have a way to visualize filters (eg. https://blog.keras.io/how-convolutional-neural-networks-see-the-world.html)
joelthelion on HN