Convolutional Differentiable Logic Gate Networks

Felix Petersen, Hilde Kuehne, Christian Borgelt, Julian Welzel, Stefano Ermon
11/7/2024
cs.LGcs.CV

No official code found. Best implementation: Felix-Petersen/difflogic (786 stars, MIT, last commit 2024-03-19). 1 more below.

Abstract

With the increasing inference cost of machine learning models, there is a growing interest in models with fast and efficient inference. Recently, an approach for learning logic gate networks directly via a differentiable relaxation was proposed. Logic gate networks are faster than conventional neural network approaches because their inference only requires logic gate operators such as NAND, OR, and XOR, which are the underlying building blocks of current hardware and can be efficiently executed. We build on this idea, extending it by deep logic gate tree convolutions, logical OR pooling, and residual initializations. This allows scaling logic gate networks up by over one order of magnitude and utilizing the paradigm of convolution. On CIFAR-10, we achieve an accuracy of 86.29% using only 61 million logic gates, which improves over the SOTA while being 29x smaller.

View on arXivView PDF

Code Implementations(2)

Felix-Petersen/difflogicCommunity100%
MIT

A Library for Differentiable Logic Gate Networks

78693C++, CudaNov 28, 20222 years agoMIT
Grahmindol/Deep-LogicCommunity67%
MIT

Ce projet implémente un réseau de neurones basé sur des portes logiques différentiables, inspiré du concept de Logic Gate Networks (LGN) décrit dans l'article arXiv:2411.04732.

10C, MakefileNov 2, 202511 months agoMIT
adam-optimizerailogic-gatesmachine-learningnetworking+1 more

Similar Papers

A Method for Optimizing Connections in Differentiable Logic Gate Networks

88% match2025

LILogic Net: Compact Logic Gate Networks with Learnable Connectivity for Efficient Hardware Deployment

87% match2025

WARP-LUTs -- Walsh-Assisted Relaxation for Probabilistic Look Up Tables

87% match2025

Recurrent Deep Differentiable Logic Gate Networks

85% match2025

Learning Interpretable Differentiable Logic Networks

83% match2 repos12024

Polynomial Surrogate Training for Differentiable Ternary Logic Gate Networks

83% match2026

Cite this paper

@article{petersen2024convolutional,
  title  = {Convolutional Differentiable Logic Gate Networks},
  author = {Felix Petersen and Hilde Kuehne and Christian Borgelt and Julian Welzel and Stefano Ermon},
  year   = {2024},
  eprint = {2411.04732},
  archivePrefix = {arXiv},
  doi    = {10.48550/arXiv.2411.04732},
  url    = {http://arxiv.org/abs/2411.04732v1},
  journal = {NEURIPS 2024 2024}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2411.04732.svg)](https://codeofpaper.com/paper/2411.04732)

Discussion