Skip to content

Add implementation of paper "Efficient Vector Symbolic Architectures from Histogram Recovery" #185

@Harshahi

Description

@Harshahi

Hello,

I am a graduate student working with Dr. Netanel Raviv at Washington University in St. Louis. We would like to contribute an implementation of the methods from his histogram recovery paper to the TorchHD package.

Based on my review of the codebase, this will require:

1.An encoding function to map data to RS-hadamard codes
2.A new classifier based on the paper's methodology
3.A class for storing binary {-1, 1} vectors, as the current implementation uses {0, 1}

Could you provide guidance on the contribution process and point me to any documentation for writing tests? Will work on a fork and then submit a PR once I'm done.

Thanks
Harshavardhan Shahi

Linear Codes for Hyperdimensional Computing
https://arxiv.org/pdf/2403.03278
Efficient Vector Symbolic Architectures from Histogram Recovery
https://arxiv.org/pdf/2511.01838

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions