# Towards Dys-XAI: Influence-Based Explanations for Dysarthria Severity Assessment

arXiv: https://arxiv.org/abs/2606.21306 · HTML: https://mlsanity.com/paper/2606.21306/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 31 references that are in the map.
- First posted on arXiv 19 Jun 2026.

## Builds on (most cited first, 12 of 12)

- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Understanding Black-box Predictions via Influence Functions](https://mlsanity.com/paper/1703.04730/) (2017)
- [Estimating Training Data Influence by Tracing Gradient Descent](https://mlsanity.com/paper/2002.08484/) (2020)
- [Exploring Explainability: A Definition, a Model, and a Knowledge Catalogue](https://mlsanity.com/paper/2108.03012/) (2021)
- [Explainable Attribute-Based Speaker Verification](https://mlsanity.com/paper/2405.19796/) (2024)
- [Can We Trust Explainable AI Methods on ASR? An Evaluation on Phoneme Recognition](https://mlsanity.com/paper/2305.18011/) (2023)
- [Explanations for Automatic Speech Recognition](https://mlsanity.com/paper/2302.14062/) (2023)
- [CDSD: Chinese Dysarthria Speech Database](https://mlsanity.com/paper/2310.15930/) (2023)
- [Easy, Interpretable, Effective: openSMILE for voice deepfake detection](https://mlsanity.com/paper/2408.15775/) (2024)
- [Probing Whisper for Dysarthric Speech in Detection and Assessment](https://mlsanity.com/paper/2510.04219/) (2025)
- [Multilingual Dysarthric Speech Assessment Using Universal Phone Recognition and Language-Specific Phonemic Contrast Modeling](https://mlsanity.com/paper/2601.21205/) (2026)
- [SAND Challenge: Four Approaches for Dysartria Severity Classification](https://mlsanity.com/paper/2512.02669/) (2025)

## Abstract (arXiv)

Dysarthria severity assessment is essential for therapy planning and longitudinal monitoring, yet manual perceptual rating is time-consuming and variable across clinicians. Although deep learning models achieve strong performance, their black-box nature limits clinical adoption. Existing speech explainability methods typically provide acoustic feature importance scores that are difficult for end-users to interpret. We propose an influence-based, instance-level explainability framework that explains each decision through supportive and competing training samples. Using gradient-based influence approximations, we compute per-utterance influence scores to identify supportive and competing training samples for each prediction. Controlled deletion experiments from 5 to 20 percent validate the explanations, showing that removing highly influential samples systematically shifts predictions. This approach provides auditable explanations by linking decisions to perceptible reference cases.