Efficient and Transparent Machine Learning: Exploring Applications of Differentiable Logic Gate Networks
Undergraduate Honors Thesis, University of Florida (UF), 2025
This undergraduate thesis advances neurosymbolic AI by operationalizing Differentiable Logic Gate Networks (DiffLogic) across hardware, compression, and visualization pipelines. The work deploys DiffLogic on FPGAs for efficient inference, proposes a compression strategy that preserves decision fidelity, and delivers interpretability tooling that exposes the logical structures learned by the network. Due to an embargo, the manuscript will be released at a later date.
@phdthesis{maldaner2025difflogic,
title={Efficient and Transparent Machine Learning: Exploring Applications of Differentiable Logic Gate Networks},
author={Maldaner, Matheus Kunzler},
school={University of Florida},
year={2025}
}