mlsanityData to 14 Oct 2026

Perceptual compensation for tonal context in self-supervised speech models

arXiv 2606.17835 ↗First posted 16 Jun 2026cs.CL cs.AI eess.AScited by 0

Not on the map. Why?

Abstract (arXiv)

This study examines the extent to which the wav2vec2.0 architecture exhibits evidence of compensation for phonological context. We conducted a pseudo-replication of a perceptional compensation experiment on Mandarin Chinese tones, and compared the embedding similarities and probing classifier outputs between a purely self-supervised pre-trained model and a model fine-tuned for Mandarin ASR. No evidence of compensation was found in the embedding similarities of the purely pre-trained model. Probing classifiers showed some evidence of compensation in addition to the expected layer-wise improvements in categorization, but failed to replicate human performance on isolated test syllables. Our findings contrast with previous reports of sensitivity to phonological structure emerging through pre-training alone, and suggest that supervised objectives may be necessary to encourage the abstraction of at least some types of phonological regularities.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 35 references that are in the map. First posted on arXiv 16 Jun 2026.

Builds on

15 of 35 references.

References · 15

  1. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
  2. 2021Probing Classifiers: Promises, Shortcomings, and Advancescited by 876
  3. 2022Self-Supervised Speech Representation Learning: A Reviewcited by 260
  4. 2023What Do Self-Supervised Speech Models Know About Words?cited by 64
  5. 2022Do self-supervised speech models develop human-like perception biases?cited by 28
  6. 2020Probing Acoustic Representations for Phonetic Propertiescited by 26
  7. 2024Encoding of lexical tone in self-supervised models of spoken languagecited by 23
  8. 2024Human-like Linguistic Biases in Neural Speech Models: Phonetic Categorization and Phonotactic Constraints in Wav2Vec2.0cited by 19
  9. 2023What Can an Accent Identifier Learn? Probing Phonetic and Prosodic Information in a Wav2vec2-based Accent Identification Modelcited by 18
  10. 2024A layer-wise analysis of Mandarin and English suprasegmentals in SSL speech modelscited by 18
  11. 2025Effective Context in Neural Speech Modelscited by 5
  12. 2025What do self-supervised speech models know about Dutch? Analyzing advantages of language-specific pre-trainingcited by 4
3 more
  1. 2024Perception of Phonological Assimilation by Neural Speech Recognition Modelscited by 3
  2. 2025Word stress in self-supervised speech models: A cross-linguistic comparisoncited by 3
  3. 2023Bigger is not Always Better: The Effect of Context Size on Speech Pre-Trainingcited by 2

Built on by

Not cited yet.