AND: Audio Network Dissection for Interpreting Deep Acoustic Models

TL;DR

This paper introduces AND, a novel framework for interpreting deep acoustic models at the neuron level by generating natural language explanations of responsive audio features, enabling better understanding and unlearning of audio neural networks.

Contribution

AND is the first framework to automatically generate natural language explanations for neurons in deep acoustic models, bridging interpretability gaps in audio neural network research.

Findings

Models rely on basic acoustic features rather than high-level concepts.

Training strategies influence neuron interpretability and model behavior.

AND enables concept-specific audio unlearning through description-based pruning.

Abstract

Neuron-level interpretations aim to explain network behaviors and properties by investigating neurons responsive to specific perceptual or structural input patterns. Although there is emerging work in the vision and language domains, none is explored for acoustic models. To bridge the gap, we introduce $\textit{AND}$, the first $\textbf{A}$udio $\textbf{N}$etwork $\textbf{D}$issection framework that automatically establishes natural language explanations of acoustic neurons based on highly-responsive audio. $\textit{AND}$ features the use of LLMs to summarize mutual acoustic features and identities among audio. Extensive experiments are conducted to verify $\textit{AND}$'s precise and informative descriptions. In addition, we demonstrate a potential use of $\textit{AND}$ for audio machine unlearning by conducting concept-specific pruning based on the generated descriptions. Finally, we highlight two acoustic model behaviors with analysis by $\textit{AND}$: (i) models discriminate audio with a combination of basic acoustic features rather than high-level abstract concepts; (ii) training strategies affect model behaviors and neuron interpretability -- supervised training guides neurons to gradually narrow their attention, while self-supervised learning encourages neurons to be polysemantic for exploring high-level features.