Extracting Formulae in Many-Valued Logic from Deep Neural Networks

Yani Zhang, Helmut Bölcskei
1/22/2024
cs.LGcs.AIcs.LO

No confident code implementation found for this paper.

Abstract

We propose a new perspective on deep ReLU networks, namely as circuit counterparts of Lukasiewicz infinite-valued logic -- a many-valued (MV) generalization of Boolean logic. An algorithm for extracting formulae in MV logic from deep ReLU networks is presented. As the algorithm applies to networks with general, in particular also real-valued, weights, it can be used to extract logical formulae from deep ReLU networks trained on data.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Complete Identification of Deep ReLU Neural Networks by Many-Valued Logic

87% match2026

Cellular automata, many-valued logic, and deep neural networks

86% match2024

Categorical Construction of Logically Verifiable Neural Architectures

80% match2025

Unwrapping All ReLU Networks

80% match2023

Quantifying The Limits of AI Reasoning: Systematic Neural Network Representations of Algorithms

80% match2025

Efficiently Learning Probabilistic Logical Models by Cheaply Ranking Mined Rules

79% match2024

Cite this paper

@article{zhang2024extracting,
  title  = {Extracting Formulae in Many-Valued Logic from Deep Neural Networks},
  author = {Yani Zhang and Helmut Bölcskei},
  year   = {2024},
  eprint = {2401.12113},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2401.12113v2}
}

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/2401.12113.svg)](https://codeofpaper.com/paper/2401.12113)

Discussion