Torchhd is a Python library for Hyperdimensional Computing and Vector Symbolic Architectures

Cognitive Computing with Associative Memory

Hyperdimensional Computing Library for building Vector-Symbolic Architectures in Python 3

GPU-accelerated neural network operations using Vulkan compute shaders.

Hyperdimensional computing with statistical guarantees: calibrated probabilities, conformal prediction sets, anomaly detection with a guaranteed false-positive rate

Nazgul is a unified C++/Python framework for time-optimal multi-joint trajectory planning (inherited from LongTermPlanner) and edge-efficient hyperdimensional computing (imported from Arthedain), providing a VSA backend factory with 8 algebras, HDC energy analysis at 45nm CMOS

FPGA-accelerated event-based normal flow estimation and real-time collision avoidance for drones — 100Hz end-to-end with microsecond latency.

A chef's palate for AI agent memory — un-mix any day's work into its exact projects, and detect workstreams nobody has named yet. Hyperdimensional fingerprints, zero dependencies.

Repository for HYPERDOA: Robust and Efficient DoA Estimation using Hyperdimensional Computing

"VSA, Analogy, and Dynamic Similarity" presentation given at the Workshop on Developments in Hyperdimensional Computing and Vector Symbolic Architectures, Heidelberg, Germany, 2020-03-16.

Publications by Peter Overmann

Hyperprobe is the Python implementation of the framework proposed in the paper "Hyperdimensional Probe: Decoding LLM Representations via Vector Symbolic Architectures".

Keynote presentation for the Midnight Sun Workshop on Vector Symbolic Architectures

This project aims to develop a very basic Vector Symbolic Architecture model to use as a default model in my other VSA projects.

Composition-episodic cognitive memory for AI agents (VSA + lattice geometry)

A quantum Hyper-Dimensional Computing (qHDC) framework in Qiskit.

Source code of the slides for the lecture "Analogical Reasoning" given on 2021⁠-⁠10⁠-⁠06 as Module 6 of Neuroscience 299: Computing with High-Dimensional Vectors at the Redwood Center for Theoretical Neuroscience, University of California, Berkeley

Research to use Vector Symbolic Architectures to implement altitude hold in a simulated multicopter.

VSALab — Interactive Hyperdimensional Computing & Vector Symbolic Architectures Laboratory. 6 modules: HD vectors, language recognition, SDM, graph analogy, resonator networks, VSA vs NN arena. Zero dependencies.