mlsanityData to 7 Oct 2026

Probing self-supervised speech models for phonetic and phonemic information: a case study in aspiration

arXiv 2306.06232 ↗First posted 9 Jun 2023cs.CL cs.SD eess.AScited by 35

Not on the map. Why?

Abstract (arXiv)

Textless self-supervised speech models have grown in capabilities in recent years, but the nature of the linguistic information they encode has not yet been thoroughly examined. We evaluate the extent to which these models' learned representations align with basic representational distinctions made by humans, focusing on a set of phonetic (low-level) and phonemic (more abstract) contrasts instantiated in word-initial stops. We find that robust representations of both phonetic and phonemic distinctions emerge in early layers of these models' architectures, and are preserved in the principal components of deeper layer representations. Our analyses suggest two sources for this success: some can only be explained by the optimization of the models on speech data, while some can be attributed to these models' high-dimensional architectures. Our findings show that speech-trained HuBERT derives a low-noise and low-dimensional subspace corresponding to abstract phonological distinctions.

Key facts. Cited by 35 papers in the map (data through 7 Oct 2026). Builds on 14 of its 23 references that are in the map. First posted on arXiv 9 Jun 2023.

Builds on

14 of 23 references.

References · 14

  1. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,243
  2. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,853
  3. 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,301
  4. 2021Probing Classifiers: Promises, Shortcomings, and Advancescited by 831
  5. 2021Layer-wise Analysis of a Self-supervised Speech Representation Modelcited by 293
  6. 2020TERA: Self-Supervised Learning of Transformer Encoder Representation for Speechcited by 186
  7. 2022Comparative layer-wise analysis of self-supervised speech modelscited by 155
  8. 2019Generative Pre-Training for Speech with Autoregressive Predictive Codingcited by 82
  9. 2022Toward a realistic model of speech processing in the brain with self-supervised learningcited by 63
  10. 2022Self-supervised models of audio effectively explain human cortical responses to speechcited by 37
  11. 2022Do self-supervised speech models develop human-like perception biases?cited by 28
  12. 2020Similarity Analysis of Self-Supervised Speech Representationscited by 26
2 more
  1. 2020Self-Supervised Representations Improve End-to-End Speech Translationcited by 21
  2. 2020Exploring the Use of an Unsupervised Autoregressive Model as a Shared Encoder for Text-Dependent Speaker Verificationcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 8; 2025: 14; 2026: 13.

35 papers.

Sort:

By area:

  1. 2024Self-Supervised Speech Representations are More Phonetic than Semanticcited by 70
  2. 2025How Linguistics Learned to Stop Worrying and Love the Language Modelscited by 50
  3. 2024Encoding of lexical tone in self-supervised models of spoken languagecited by 22
  4. 2024Human-like Linguistic Biases in Neural Speech Models: Phonetic Categorization and Phonotactic Constraints in Wav2Vec2.0cited by 19
  5. 2024A layer-wise analysis of Mandarin and English suprasegmentals in SSL speech modelscited by 18
  6. 2025Leveraging Allophony in Self-Supervised Speech Models for Atypical Pronunciation Assessmentcited by 15
  7. 2025Emergent morpho-phonological representations in self-supervised speech modelscited by 11
  8. 2026Self-Supervised Speech Models Encode Phonetic Context via Position-dependent Orthogonal Subspacescited by 7
  9. 2024Self-supervised Speech Representations Still Struggle with African American Vernacular Englishcited by 6
  10. 2025Layer-wise Minimal Pair Probing Reveals Contextual Grammatical-Conceptual Hierarchy in Speech Representationscited by 5
  11. 2025PASE: Leveraging the Phonological Prior of WavLM for Low-Hallucination Generative Speech Enhancementcited by 5
  12. 2025Iterative refinement, not training objective, makes HuBERT behave differently from wav2vec 2.0cited by 4
23 more
  1. 2024Probing self-attention in self-supervised speech models for cross-linguistic differencescited by 3
  2. 2024Do Discrete Self-Supervised Representations of Speech Capture Tone Distinctions?cited by 3
  3. 2025Analyzing the relationships between pretraining language, phonetic, tonal, and speaker information in self-supervised speech modelscited by 3
  4. 2024A predictive learning model can simulate temporal dynamics and context effects found in neural representations of continuous speechcited by 2
  5. 2025Echoes of Phonetics: Unveiling Relevant Acoustic Cues for ASR via Feature Attributioncited by 2
  6. 2025From Words to Waves: Analyzing Concept Formation in Speech and Text-Based Foundation Modelscited by 1
  7. 2025Identifying Speaker Information in Feed-Forward Layers of Self-Supervised Speech Transformerscited by 1
  8. 2025Language models as tools for investigating the distinction between possible and impossible natural languagescited by 0
  9. 2025Large language models have learned to use languagecited by 0
  10. 2025Large language models are not about natural languagecited by 0
  11. 2025Linguists should learn to love speech-based deep learning modelscited by 0
  12. 2026Categorize Early, Integrate Late: Divergent Processing Strategies in Automatic Speech Recognitioncited by 0
  13. 2026Are Language Models Models?cited by 0
  14. 2026Tracking the emergence of linguistic structure in self-supervised models learning from speechcited by 0
  15. 2026Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated Probabilitiescited by 0
  16. 2026You Can't Fight in Here! This is BBS!cited by 0
  17. 2026A framework for analyzing concept representations in neural modelscited by 0
  18. 2026Not-So-Strange Love: Language Models and Generative Linguistic Theories are More Compatible than They Appearcited by 0
  19. 2026Probing in the Wild: A Case Study of Self-Supervised Speech Representations on Mandarin Sub-dialects with Unsupervised Articulatory Analysiscited by 0
  20. 2026Structured Phonological Representations for Audio-Articulatory rtMRI Speech Classificationcited by 0
  21. 2026Cluster Assignments in Soft Targets Shape Speech Representations: Evidence from S-JEPAcited by 0
  22. 2026Do speech foundation models really learn words?cited by 0
  23. 2026Phonological Interference in Multilingual Speech Modelscited by 0