4 ms·
Sparse Distributed Memory
- bobosha 8y agoKanerva's approach has been championed by Jeff Hawkins' Numenta and Redwood Neuroscience Institute. Numenta uses the terminology Sparse Distributed Representations (SDR). While the theoretical basis seems sound, a practical implementation comparable to the SOTA is lacking.
- bra-ket 8y agoHa, I wrote most of this article. Also check out https://en.wikipedia.org/wiki/Holographic_associative_memory https://en.wikipedia.org/wiki/Holographic_associative_memory but it needs some work
- schrodingersket 8y agoLinks aside, it seems there's another out there with a kindred username! Also, great write-up. That article is awesome!
- espeed 8y agoDouble Ha! I just posted that too, in a comment last week... Discussion: Algebra and the Lambda Calculus (1993) https://news.ycombinator.com/item?id=18296384 https://news.ycombinator.com/item?id=18296384 Is holographic associative memory one of your areas of research too?
- bra-ket 8y agoit's the closest we can get to modeling something like phase precession or phase resetting in grid cells http://www.scholarpedia.org/article/Grid_cells http://www.scholarpedia.org/article/Grid_cells
- espeed 8y agoAny good papers/code examples that discuss data model design and implementation details?
- evgen 8y agoI don't keep around many items from my collection of my mid-90s AI books, but for some reason Kanerva's Sparse Distributed Memory book always managed to make the cut whenever I go in to cull things. There is something about the idea that seems appealing and I keep waiting for some excuse to see if any of the ideas can be melded with more recent approaches to NN memory...