mlsanityData to 14 Oct 2026

PRiSM: Benchmarking Phone Realization in Speech Models

arXiv 2601.14046 ↗First posted 20 Jan 2026cs.CL cs.SDcited by 4

Not on the map. Why?

Abstract (arXiv)

Phone recognition (PR) serves as the atomic interface for language-agnostic modeling for cross-lingual speech processing and phonetic analysis. Despite prolonged efforts in developing PR systems, current evaluations only measure surface-level transcription accuracy. We introduce PRiSM, the first open-source benchmark designed to expose blind spots in phonetic perception through intrinsic and extrinsic evaluation of PR systems. PRiSM standardizes transcription-based evaluation and assesses downstream utility in clinical, educational, and multilingual settings with transcription and representation probes. We find that diverse language exposure during training is key to PR performance, encoder-CTC models are the most stable, and specialized PR models still outperform Large Audio Language Models. PRiSM releases code, recipes, and datasets to move the field toward multilingual speech models with robust phonetic ability: https://github.com/changelinglab/prism.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 40 of its 74 references that are in the map. First posted on arXiv 20 Jan 2026.

Builds on

40 of 74 references.

References · 40

  1. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
  2. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
  3. 2017Axiomatic Attribution for Deep Networkscited by 3,384
  4. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  5. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
  6. 2021SUPERB: Speech processing Universal PERformance Benchmarkcited by 666
  7. 2022FLEURS: Few-shot Learning Evaluation of Universal Representations of Speechcited by 469
  8. 2025On The Landscape of Spoken Language Models: A Comprehensive Surveycited by 118
  9. 2020The Zero Resource Speech Benchmark 2021: Metrics and baselines for unsupervised spoken language modelingcited by 101
  10. 2021Intermediate Loss Regularization for CTC-based Speech Recognitioncited by 96
  11. 2020Universal Phone Recognition with a Multilingual Allophone Systemcited by 81
  12. 2021Simple and Effective Zero-shot Cross-lingual Phoneme Recognitioncited by 79
28 more
  1. 2024Self-Supervised Speech Representations are More Phonetic than Semanticcited by 71
  2. 2024A Survey on Speech Large Language Models for Understandingcited by 71
  3. 2024Dynamic-SUPERB Phase-2: A Collaboratively Expanding Benchmark for Measuring the Capabilities of Spoken Language Models with 180 Taskscited by 71
  4. 2021speechocean762: An Open-Source Non-native English Speech Corpus For Pronunciation Assessmentcited by 58
  5. 2022ByT5 model for massively multilingual grapheme-to-phoneme conversioncited by 49
  6. 2025Towards Holistic Evaluation of Large Audio-Language Models: A Comprehensive Surveycited by 39
  7. 2023The Edinburgh International Accents of English Corpus: Towards the Democratization of English ASRcited by 36
  8. 2023BabySLM: language-acquisition-friendly benchmark of self-supervised spoken language modelscited by 29
  9. 2022ASR2K: Speech Recognition for Around 2000 Languages without Audiocited by 26
  10. 2021The Accented English Speech Recognition Challenge 2020: Open Datasets, Tracks, Baselines, Results and Methodscited by 24
  11. 2020Phonotactic Complexity and its Trade-offscited by 23
  12. 2025ZIPA: A family of efficient models for multilingual phone recognitioncited by 19
  13. 2023What Can an Accent Identifier Learn? Probing Phonetic and Prosodic Information in a Wav2vec2-based Accent Identification Modelcited by 18
  14. 2024Whistle: Data-Efficient Multilingual and Crosslingual Speech Recognition via Weakly Phonetic Supervisioncited by 17
  15. 2024CR-CTC: Consistency regularization on CTC for improved speech recognitioncited by 17
  16. 2019UltraSuite: A Repository of Ultrasound and Acoustic Data from Child Speech Therapy Sessionscited by 16
  17. 2023Self-Supervised Models of Speech Infer Universal Articulatory Kinematicscited by 16
  18. 2021EasyCall corpus: a dysarthric speech datasetcited by 15
  19. 2025Leveraging Allophony in Self-Supervised Speech Models for Atypical Pronunciation Assessmentcited by 15
  20. 2023The taste of IPA: Towards open-vocabulary keyword spotting and forced alignment in any languagecited by 14
  21. 2025POWSM: A Phonetic Open Whisper-Style Speech Foundation Modelcited by 14
  22. 2024Cross-Lingual Transfer Learning for Speech Translationcited by 11
  23. 2024Phonetic Segmentation of the UCLA Phonetics Lab Archivecited by 9
  24. 2018Investigating the role of L1 in automatic pronunciation evaluation of L2 speechcited by 5
  25. 2021Tusom2021: A Phonetically Transcribed Speech Dataset from an Endangered Language for Universal Phone Recognition Experimentscited by 5
  26. 2024Small Language Models Also Work With Small Vocabularies: Probing the Linguistic Abilities of Grapheme- and Phoneme-Based Baby Llamascited by 5
  27. 2025Finding My Voice: Generative Reconstruction of Disordered Speech for Automated Clinical Evaluationcited by 3
  28. 2025BabyLM's First Words: Word Segmentation as a Phonological Probing Taskcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026An Empirical Recipe for Universal Phone Recognitioncited by 5
  2. 2026Multilingual Dysarthric Speech Assessment Using Universal Phone Recognition and Language-Specific Phonemic Contrast Modelingcited by 4
  3. 2026Phone Segmentation and Recognition through Phonological Activation Mappingcited by 0
  4. 2026MoLGE: Mixture of Language Group Experts for Efficient Scaling of Massively Multilingual Speech Recognitioncited by 0