Feed

Trying to decide how to show the selected points. This is a little harsh.

Inspiration

This is great and it's going to take me a while to finish. It shows how a thoughtful interface and framing can change how a process feels (compare this to going through your following list in the normal Twitter interface).

From: https://tokimeki-unfollow.glitch.me


I got the MNIST images rendering in the active learning prototype. White here are the images selected for labeling for the next round.

Inspiration

Rasmus Andersson's favorite technical papers.

From: https://www.dropbox.com/sh/is0sy5350lr4v9j/AADQlhVSQcRw6vCNKQgGWelqa

Inspiration

Lots of neat generative sketches by Kjetil Golid.

From: https://generated.space/


First pass of active learning rounds on MNIST in the prototype. I need to think through what the transitions should be more, but this looks promising.

Inspiration

Oooo I bet there are some interesting things to be done with context.isPointinPath().

From: https://beta.observablehq.com/@mbostock/pixelated-world

Inspiration

Mushy: neural network generated isometric tiles by Everest Pipkin.

From: https://everestpipkin.itch.io/mushy


I'm making a notebook to visualize and figure out what frame and matting options I should choose.


I changed how I'm handling transparency thresholding so the blobs are a lot less muddy now.

From: https://grantcuster.github.io/umap-explorer/


I didn't think this all the way through and ran UMAP on each of these groups separately instead of at the same time.


Starting to mock up the prototype.


I added T-SNE and UMAP with min_dist=0.8 algorithm options to the UMAP explorer. Three.js & tween.js animating the transition of 70,000 points no problem.

From: https://grantcuster.github.io/umap-explorer


Working on getting T-SNE running in the MNIST explorer.


I made a notebook that lets you crop an image.

From: https://beta.observablehq.com/@grantcuster/image-crop


I put up a demo of the UMAP on MNIST interactive visualization I've been working on.

From: https://grantcuster.github.io/umap-explorer/


Cleaning up the layout for the UMAP visualization.


Another hover variation.


You can see how UMAP clusters are structured by things like the orientation of the number if you travel along their axes.


Making a hover display in three.js using a canvas as a texture. Canvas size needs to be a power of two. I'm using Figma to design it and then I can just move the coordinates into code.