Introducing my work / code for sparse binary hypervectors...

[Zhonghao (Kevin) Yang]

Hey, VSA community:

This is my first post after joining this community for a while...

I have been exploring VSA/hypervector computing for a long while, starting as a personal hobby to better understand human cognition and neuroscience. Jeff Hawkins' On Intelligence and Pentti Kanerva's Sparse Distributed Memory were both eye-openers for me, and I never looked back.

Recently I connected with people from Berkeley (Prof. Bruno Olshausen, Pentti Kanerva and Fritz Sommer, etc), and decided to open-source the Python module for my binary sparse hypervectors ( https://github.com/yangzh/hv ) so that the VSA community can evaluate it and possibly build interesting applications on top of it.

In summary, the module features:

1. sparse binary hypervector and its innovative bind / bundle operators;

2. a highly optimized computing / storage engine for such vectors;

3. Ergonomic APIs to help people easily experiment with Python/Notebooks.

4. Permissive license (BSD);

Again, the code, which includes plenty of documentation and examples, can be found here:

https://github.com/yangzh/hv

I'm also open and eager to hear the community's feedback and interesting directions. Thanks.

--
Zhonghao (Kevin) Yang,  (yan...@gmail.com)