# I Guess That's Why They Call it the Blues: Causal Analysis for Audio Classifiers

arXiv: https://arxiv.org/abs/2601.16675 · HTML: https://mlsanity.com/paper/2601.16675/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 24 references that are in the map.
- First posted on arXiv 23 Jan 2026.

## Builds on (most cited first, 11 of 11)

- [One pixel attack for fooling deep neural networks](https://mlsanity.com/paper/1710.08864/) (2017)
- [Audio Adversarial Examples: Targeted Attacks on Speech-to-Text](https://mlsanity.com/paper/1801.01944/) (2018)
- [Targeted Adversarial Examples for Black Box Audio Systems](https://mlsanity.com/paper/1805.07820/) (2018)
- [SirenAttack: Generating Adversarial Audio for End-to-End Acoustic Systems](https://mlsanity.com/paper/1901.07846/) (2019)
- [audioLIME: Listenable Explanations Using Source Separation](https://mlsanity.com/paper/2008.00582/) (2020)
- [Explanations for Occluded Images](https://mlsanity.com/paper/2103.03622/) (2021)
- [Causal Explanations for Image Classifiers](https://mlsanity.com/paper/2411.08875/) (2024)
- [FACE: Faithful Automatic Concept Extraction](https://mlsanity.com/paper/2510.11675/) (2025)
- [Sufficient, Necessary and Complete Causal Explanations in Image Classification](https://mlsanity.com/paper/2507.23497/) (2025)
- [I Am Big, You Are Little; I Am Right, You Are Wrong](https://mlsanity.com/paper/2507.23509/) (2025)
- [Out-of-the-box: Black-box Causal Attacks on Object Detectors](https://mlsanity.com/paper/2512.03730/) (2025)

## Built on by (most cited first, 1 of 1)

- [Mask-Induced Displacement in Audio XAI via Logit Trajectory Decomposition](https://mlsanity.com/paper/2609.33486/) (2026)

## Abstract (arXiv)

It is well-known that audio classifiers often rely on non-musically relevant features and spurious correlations to classify audio. Hence audio classifiers are easy to manipulate or confuse, resulting in wrong classifications. While inducing a misclassification is not hard, until now the set of features that the classifiers rely on was not well understood. In this paper we introduce a new method that uses causal reasoning to discover features of the frequency space that are sufficient and necessary for a given classification. We describe an implementation of this algorithm in the tool FreqReX and provide experimental results on a number of standard benchmark datasets. Our experiments show that causally sufficient and necessary subsets allow us to manipulate the outputs of the models in a variety of ways by changing the input very slightly. Namely, a change to one out of 240,000 frequencies results in a change in classification 58% of the time, and the change can be so small that it is practically inaudible. These results show that causal analysis is useful for understanding the reasoning process of audio classifiers and can be used to successfully manipulate their outputs.