News for hackers. Stuff that matters.
Re:A New Approach to Computation Reimagines Artificial Intel... by intalentive on Thursday April 13, 2023 @12:52PM (#35558389)
>Weiss figured out how to represent a complex image as a single hyperdimensional vector that contains information about all the objects in the image, including their properties, such as colors, positions and sizes.

How is this different from an N-dimensional embedding produced by, say, a convolutional neural network?

Re:A New Approach to Computation Reimagines Artificial Intel... by esafak on Thursday April 13, 2023 @02:16PM (#35559736)
> The team then trained a neural network to examine an image and generate a bipolar hypervector — an element can be +1 or −1 — that’s as close as possible to some superposition of hypervectors in the dictionary...

It seems the former uses binary embeddings, which in ML is used more frequently for similarity search. They also seem to care about orthogonality. Beyond that we'd have to read the paper to know how the embeddings are derived. If you're motivated, here they are: http://www.rctn.org/bruno/papers/

Re:A New Approach to Computation Reimagines Artificial Intel... by HybridCurve on Thursday April 13, 2023 @06:33PM (#35563055)
I am reasonably certain this is the paper the article is discussing:

https://redwood.berkeley.edu/wp-content/uploads/2018/01/rahi...