Recall to Imagine, a model-based RL algorithm with superhuman memory. Oral (1.2%) @ ICLR 2024

Simple and fast parallel scan over sequences of tensors, with any binary associative function you specify, for PyTorch (Franz A. Heinsen, 2024)

Network Inventory Scanner

Golang DynamoDB parallel scan

Reference implementation of selective resetting for parallel prefix scans, as proposed in "Generalized Orders of Magnitude for Scalable, Parallel, High-Dynamic-Range Computation" (Heinsen and Kozachkov, 2025).

Triton kernels for linear RNN (currently recurrentgemma)

Chunked parallel associative scan in Triton: varlen packing, bidirectional, torch.compile. Drop-in primitive for gated linear RNNs.

A JAX / NNX implementation of a linear GRU for language modeling

This repository contains three challenges presented on the course of Advanced Algorithms and Parallel Programming (095946) at Politecnico di Milano.

Impl of Dyn Sys Parallelization Techniques on RNNs and ODE Solvers Using pyTorch.

GSSM: a reproducing-kernel framework for the linear-SSM family (Mamba/S5/LRU as parametric special cases) + key-conditioned holographic recall. Research disclosure.

Scan Windows networks quickly to identify devices, detect OS, find MAC vendors, analyze risks, and generate detailed HTML reports.