# [b] = [d] - [t] + [p]: Self-supervised Speech Models Discover Phonological Vector Arithmetic

arXiv: https://arxiv.org/abs/2602.18899 · HTML: https://mlsanity.com/paper/2602.18899/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 38 of its 84 references that are in the map.
- First posted on arXiv 21 Feb 2026.

## Builds on (most cited first, 30 of 38)

- [Efficient Estimation of Word Representations in Vector Space](https://mlsanity.com/paper/1301.3781/) (2013)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [Steering Language Models With Activation Engineering](https://mlsanity.com/paper/2308.10248/) (2023)
- [How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings](https://mlsanity.com/paper/1909.00512/) (2019)
- [LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speech](https://mlsanity.com/paper/1904.02882/) (2019)
- [Unsupervised Cross-lingual Representation Learning for Speech Recognition](https://mlsanity.com/paper/2006.13979/) (2020)
- [MLS: A Large-Scale Multilingual Dataset for Speech Research](https://mlsanity.com/paper/2012.03411/) (2020)
- [Libri-Light: A Benchmark for ASR with Limited or No Supervision](https://mlsanity.com/paper/1912.07875/) (2019)
- [VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretation](https://mlsanity.com/paper/2101.00390/) (2021)
- [GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio](https://mlsanity.com/paper/2106.06909/) (2021)
- [Layer-wise Analysis of a Self-supervised Speech Representation Model](https://mlsanity.com/paper/2107.04734/) (2021)
- [Comparative layer-wise analysis of self-supervised speech models](https://mlsanity.com/paper/2211.03929/) (2022)
- [Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis](https://mlsanity.com/paper/2306.00814/) (2023)
- [Simple and Effective Zero-shot Cross-lingual Phoneme Recognition](https://mlsanity.com/paper/2109.11680/) (2021)
- [Self-Supervised Speech Representations are More Phonetic than Semantic](https://mlsanity.com/paper/2406.08619/) (2024)
- [What Do Self-Supervised Speech Models Know About Words?](https://mlsanity.com/paper/2307.00162/) (2023)
- [Towards Understanding Linear Word Analogies](https://mlsanity.com/paper/1810.04882/) (2018)
- [Word Discovery in Visually Grounded, Self-Supervised Speech Models](https://mlsanity.com/paper/2203.15081/) (2022)
- [Evidence of Vocal Tract Articulation in Self-Supervised Learning of Speech](https://mlsanity.com/paper/2210.11723/) (2022)
- [SyllableLM: Learning Coarse Semantic Units for Speech Language Models](https://mlsanity.com/paper/2410.04029/) (2024)
- [Phonological Features for 0-shot Multilingual Speech Synthesis](https://mlsanity.com/paper/2008.04107/) (2020)
- [Opening the Black Box of wav2vec Feature Encoder](https://mlsanity.com/paper/2210.15386/) (2022)
- [Wave to Syntax: Probing spoken language models for syntax](https://mlsanity.com/paper/2305.18957/) (2023)
- [Coding Speech through Vocal Tract Kinematics](https://mlsanity.com/paper/2406.12998/) (2024)
- [Analogies minus analogy test: measuring regularities in word embeddings](https://mlsanity.com/paper/2010.03446/) (2020)
- [Universal Automatic Phonetic Transcription into the International Phonetic Alphabet](https://mlsanity.com/paper/2308.03917/) (2023)
- [Self-Supervised Models of Speech Infer Universal Articulatory Kinematics](https://mlsanity.com/paper/2310.10788/) (2023)

## Built on by (most cited first, 12 of 12)

- [Self-Supervised Speech Models Encode Phonetic Context via Position-dependent Orthogonal Subspaces](https://mlsanity.com/paper/2603.12642/) (2026)
- [An Empirical Recipe for Universal Phone Recognition](https://mlsanity.com/paper/2603.29042/) (2026)
- [Adapting Self-Supervised Speech Representations for Cross-lingual Dysarthria Detection in Parkinson's Disease](https://mlsanity.com/paper/2603.22225/) (2026)
- [Speech Playground: An Interactive Tool for Speech Analysis and Comparison](https://mlsanity.com/paper/2607.00418/) (2026)
- [Language Orthogonalization for Zero-Shot Cross-Lingual Audio Deepfake Detection](https://mlsanity.com/paper/2609.16458/) (2026)
- [A Native-Reference Coordinate Geometry for L2 Pronunciation Deviation Using Self-Supervised Speech Models](https://mlsanity.com/paper/2609.28060/) (2026)
- [Phone Segmentation and Recognition through Phonological Activation Mapping](https://mlsanity.com/paper/2607.09020/) (2026)
- [Quantifying Consonant Contributions to Word Intelligibility via Acoustic Masking](https://mlsanity.com/paper/2609.12122/) (2026)
- [A Native-Reference Phone-Class Geometry for Second-Language Pronunciation Analysis](https://mlsanity.com/paper/2609.30075/) (2026)
- [Do Audio Language Models Hear and Read Distinctive Features Alike?](https://mlsanity.com/paper/2609.30167/) (2026)
- [Child-Adapted Structured Phonological Representations for Interpretable Speech Sound Analysis](https://mlsanity.com/paper/2610.00852/) (2026)
- [Deep Speech Representations Track Word Timing along a Low-Dimensional Circular Trajectory](https://mlsanity.com/paper/2610.16273/) (2026)

## Abstract (arXiv)

Self-supervised speech models (S3Ms) are known to encode rich phonetic information, yet how this information is structured remains underexplored. We conduct a comprehensive study across 96 languages to analyze the underlying structure of S3M representations, with particular attention to phonological vectors. We first show that there exist linear directions within the model's representation space that correspond to phonological features. We further demonstrate that the scale of these phonological vectors correlate to the degree of acoustic realization of their corresponding phonological features in a continuous manner. For example, the difference between [d] and [t] yields a voicing vector: adding this vector to [p] produces [b], while scaling it results in a continuum of voicing. Together, these findings indicate that S3Ms encode speech using phonologically interpretable and compositional vectors, demonstrating phonological vector arithmetic. All code and interactive demos are available at https://github.com/juice500ml/phonetic-arithmetic .