[image] How to reconstruct the SP algorithm? NuPIC Hi every one I am looking for a code with which I can feed a specific dataset to the SP algorithm and then reconstruct the output of the SP and get the original data. For example, give one of the MNIST digits …

That’s a nice little auto-encoder. It looks like the main differences between your approach and a more traditional auto-encoder are that yours is both shallow and wide. In other words, you are using only one hidden layer, and that layer encodes the latent variables/features. By enforcing k-WTA, you …

I force the values to 0/1 but all features are updated simultaneously so they do affect each other in ways I dont completely understand. I need to include bosting otherwise the features get stuck and all images end up producing the same sdr. [image] CollinsEM: between your approach and a more…

[image] CollinsEM: Are the latent variables scalar values? Or do you force them to 0/1? Are the features at the bottom the result of passing a single latent variable (SDR size of 1) oh, sorry, I misinterpreted what you meant. yea, those are the weights from a single bit, its basically doing …

[image] JarvisGoBrr: well, about that, it still kinda works if I make it shallow and narrow. I guess that would mean that you have more data being encoded in the weight matrix than you need. The words that I would use to describe it are that the columns (or rows) of your weight matrix are enc…

[image] CollinsEM: Since you are forcing your latent variables to 0/1, it would seem that the filters are adjusting to allow you to stack them in such as way as to obtain brighter pixels in regions where they overlap positive weights and darker pixels where they overlap a mix of signed weights, …

[image] JarvisGoBrr: The algorithm goes as follows How can you guarantee the sparsity? How fast is your algorithm?

sparsity is enforced by K-WTA with boosting just like spatial pooler and it runs at around 2000 iterations per second for 1000/30 sparsity on my machine.

@JarvisGoBrr thanks. I am interested in how your encoder can be realized with HTM neurons or SP instead of using floating weights

@thanh-binh.to starting with step 2: run the SDR backwards through the matrix to get a reconstruction estimate of the image. here-s a multiplication matrix HTM works with binary synapses attached to dendrites attached to cells grouped in columns, you would have a bit of a problem figuring it o…

[image] JarvisGoBrr: yea, thats what I think is happening but the particular shapes they choose are always a little funny and seem to have some gravitating effect where they force others to be similar instead of different, If I could get rid of that I think the encoder would have better performa…

Lossy mnist encoder

Engineering

thanh-binh.to October 11, 2022, 7:02pm 7

How can you guarantee the sparsity?
How fast is your algorithm?

Topic		Replies	Views
How to reconstruct the SP algorithm? NuPIC	14	1214	September 1, 2022
Encoder to skip SP! Implementations	5	744	March 3, 2021
Encoder for convert image to binary matrix Applications encoders , image-recognition	17	1807	August 7, 2020
The output of the Encoder can be a dense representation? Applications	7	576	February 2, 2020
Encoder and Spatial Pooler Confusion Getting Started	17	1027	April 5, 2019

Lossy mnist encoder

Related topics