Probing in the Wild: A Case Study of Self-Supervised Speech Representations on Mandarin Sub-dialects with Unsupervised Articulatory Analysis

2026-06-24

AI Summary

This study investigates the sensitivity and stability of internal representations in self-supervised speech models when handling fine-grained phonetic variation across Chinese sub-dialects. We propose an unsupervised probing approach that integrates a language-agnostic universal phone recognizer with articulatory feature mappings, enabling the first analysis of cross-dialectal decodability of phonetic features on real-world, unlabeled dialectal corpora. Our findings reveal that acoustically salient features—such as labiality and frication—are stably represented across sub-dialects, whereas fine-grained spectral characteristics exhibit significantly higher decodability in Beijing Mandarin. Moreover, different phonetic dimensions display distinct layer-wise dynamics within the model architecture, uncovering an uneven sensitivity of self-supervised representations to dialectal variation.

Abstract

While self-supervised speech models have achieved strong performance across speech tasks, relatively little is known about how their internal phonetic representations behave under fine-grained dialect variation. Existing probing studies typically rely on curated corpora with manual phonetic annotations, limiting their applicability to naturally occurring dialect speech. We present a case study of articulatory feature representations in a Mandarin self-supervised speech model using an entirely unlabeled probing pipeline. Phone sequences are generated using a language-agnostic universal phone recognizer and mapped to articulatory feature vectors, enabling frame-level probing without manual annotation. Our results reveal a structured pattern in articulatory feature decodability across Mandarin sub-dialects. Acoustically salient features such as labiality and stridency remain comparatively stable, whereas features associated with finer spectral distinctions exhibit larger dialect-dependent variation. This variation is driven primarily by elevated decodability for Beijing speech relative to other Mandarin sub-dialects. Layer-wise analyses further show distinct representational dynamics for these feature groups. These findings suggest that language-agnostic articulatory probing can be applied to real-world dialect corpora and that dialect sensitivity in self-supervised speech representations is unevenly distributed across articulatory dimensions.

Problem

• self-supervised speech representations
• Mandarin sub-dialects
• articulatory features
• probing
• dialect variation

Innovation

• self-supervised speech representation
• unsupervised articulatory probing
• Mandarin sub-dialects
• language-agnostic phone recognition
• frame-level phonetic decoding