# From Black Box to Biomarker: Sparse Autoencoders for Interpreting Speech Models of Parkinson's Disease

arXiv: https://arxiv.org/abs/2507.16836 · HTML: https://mlsanity.com/paper/2507.16836/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Interpreting biomedical and science models, one of 13 directions in Mechanistic interpretability.
- Builds on 14 of its 44 references that are in the map.
- First posted on arXiv 16 Jul 2025.

## Placed in

- Mechanistic interpretability / [Interpreting biomedical and science models](https://mlsanity.com/mechinterp/interpreting-biomedical-and-science-models/)

## Builds on (most cited first, 14 of 14)

- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [Scaling and evaluating sparse autoencoders](https://mlsanity.com/paper/2406.04093/) (2024)
- [Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders](https://mlsanity.com/paper/2407.14435/) (2024)
- [BatchTopK Sparse Autoencoders](https://mlsanity.com/paper/2412.06410/) (2024)
- [Open-Source Conversational AI with SpeechBrain 1.0](https://mlsanity.com/paper/2407.00463/) (2024)
- [Self-attention encoding and pooling for speaker recognition](https://mlsanity.com/paper/2008.01077/) (2020)
- [Listen to Interpret: Post-hoc Interpretability for Audio Networks with NMF](https://mlsanity.com/paper/2202.11479/) (2022)
- [Prototype Learning for Interpretable Respiratory Sound Analysis](https://mlsanity.com/paper/2110.03536/) (2021)
- [Listenable Maps for Audio Classifiers](https://mlsanity.com/paper/2403.13086/) (2024)
- [Exploiting Foundation Models and Speech Enhancement for Parkinson's Disease Detection from Speech in Real-World Operative Conditions](https://mlsanity.com/paper/2406.16128/) (2024)
- [Unveiling Interpretability in Self-Supervised Speech Representations for Parkinson's Diagnosis](https://mlsanity.com/paper/2412.02006/) (2024)
- [Bilingual Dual-Head Deep Model for Parkinson's Disease Detection from Speech](https://mlsanity.com/paper/2503.10301/) (2025)
- [Interpretable Prognostics with Concept Bottleneck Models](https://mlsanity.com/paper/2405.17575/) (2024)
- [Investigating the Effectiveness of Explainability Methods in Parkinson's Detection from Speech](https://mlsanity.com/paper/2411.08013/) (2024)

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

- [Mechanistic Interpretability of Brain-to-Speech Models Across Speech Modes](https://mlsanity.com/paper/2602.01247/) (2026)

## Abstract (arXiv)

Speech holds promise as a cost-effective and non-invasive biomarker for neurological conditions such as Parkinson's disease (PD). While deep learning systems trained on raw audio can find subtle signals not available from hand-crafted features, their black-box nature hinders clinical adoption. To address this, we apply sparse autoencoders (SAEs) to uncover interpretable internal representations from a speech-based PD detection system. We introduce a novel mask-based activation for adapting SAEs to small biomedical datasets, creating sparse disentangled dictionary representations. These dictionary entries are found to have strong associations with characteristic articulatory deficits in PD speech, such as reduced spectral flux and increased spectral flatness in the low-energy regions highlighted by the model attention. We further show that the spectral flux is related to volumetric measurements of the putamen from MRI scans, demonstrating the potential of SAEs to reveal clinically relevant biomarkers for disease monitoring and diagnosis.