Neural Interpretable Reasoning

TL;DR

This paper introduces a new neural modeling framework that enhances interpretability and scalability in deep learning by leveraging inference equivariance and neural re-parametrization, enabling transparent reasoning.

Contribution

It proposes a novel paradigm of neural generation and interpretable execution for scalable, interpretable neural reasoning models.

Findings

Scalable verification of interpretability via Markovian properties.

Neural re-parametrization reduces complexity of interpretability verification.

Framework achieves both expressiveness and transparency in neural reasoning.

Abstract

We formalize a novel modeling framework for achieving interpretability in deep learning, anchored in the principle of inference equivariance. While the direct verification of interpretability scales exponentially with the number of variables of the system, we show that this complexity can be mitigated by treating interpretability as a Markovian property and employing neural re-parametrization techniques. Building on these insights, we propose a new modeling paradigm -- neural generation and interpretable execution -- that enables scalable verification of equivariance. This paradigm provides a general approach for designing Neural Interpretable Reasoners that are not only expressive but also transparent.