Pith. sign in

REVIEW 2 major objections 5 minor 31 references

What You Read Isn't What You Hear: Linguistic Sensitivity in Deepfake Speech Detection

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Rewording the transcript before text-to-speech synthesis can flip audio deepfake detectors from “spoof” to “real,” with one commercial detector dropping from 100% to 32% accuracy.

desk verdict Transcript-level perturbations flip anti-spoofing detectors; open-source numbers are clouded by batch-norm calibration, but the commercial drop and case study make the phenomenon real. read the letter →

arxiv 2505.17513 v1 pith:UZEM3H2Q submitted 2025-05-23 cs.LG cs.CLcs.SDeess.AS

Binh Nguyen , Shuji Shi , Ryan Ofman , Thai Le This is my paper
classification cs.LGcs.CLcs.SDeess.AS
keywords deepfakeaudiodetectionanti-spoofingtranscript-leveladversarialattackslinguisticsensitivitytext-to-speechsynthesisrobustnessvoicecloningfraud
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that audio deepfake detectors are sensitive to the wording of the transcript that is read aloud by text-to-speech, not only to the acoustics of the audio. It shows that small, meaning-preserving word substitutions can flip a detector's verdict from “spoofed” to “real” for many detector–voice pairs, with attack success rates above 60% on several open-source systems and one commercial API falling from 100% to 32% accuracy. If true, this means current anti-spoofing defenses, which are evaluated almost entirely on acoustic perturbations, can be bypassed by simply editing the scam script. The paper also argues the vulnerability is systematic: a regression analysis of linguistic, acoustic, and model-level features links susceptibility to linguistic complexity and to how clustered the detector's audio embeddings are for a given synthetic voice.

What carries the argument

The load-bearing machinery is a black-box, model-agnostic transcript perturbation algorithm: it scores each word by how much its removal changes the detector's bona-fide probability, then greedily substitutes the highest-impact words with synonym or masked-language-model candidates that maximize the bona-fide score while passing a semantic-similarity threshold and part-of-speech check. The analysis side introduces Audio Encoder Similarity, the mean cosine similarity between a voice's synthesized audio embeddings and their centroid in the detector's representation space, and uses it to show that detectors whose embeddings tightly cluster a given text-to-speech voice are far less vulnerable. This metric, together with the detector's own spoof and bona-fide F1 scores, gives the strongest predictive signal for whether a transcript attack will succeed.

What would settle it

Run the same transcript-level attack against the three open-source detectors without any per-voice batch-norm calibration, or with a single fixed calibration learned on one voice, and check whether attack success rates still exceed 60%; if they collapse, the linguistic sensitivity is a calibration artifact rather than a property of the detectors.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that linguistic variation in a transcript propagates through a text-to-speech pipeline and changes the acoustic realization enough that state-of-the-art anti-spoofing models change their decision, even though the audio's semantic content and the synthesis system are untouched. The authors formalize this as an optimization problem: find a minimally edited transcript, constrained to preserve syntax and meaning, whose synthesized audio is labeled bona-fide. Their empirical evaluation across 1,439 transcripts, three open-source detectors, several text-to-speech voices, and two commercial APIs reports attack success rates up to 82% in open-source settings and a drop from 100% to 32% for one commercial detector when paired with a commercial TTS service. A case study replicating a celebrity voice-cloning scam shows per-utterance bona-fide probabilities rising from below 1% to 58–90% after a few lexical substitutions, which the paper presents as evidence that the vulnerability is exploitable in real-world fraud scenarios.

Load-bearing premise

The open-source attack numbers come from detectors that were batch-norm calibrated to each text-to-speech voice until accuracy exceeded 90%, and without a comparison to uncalibrated detectors it is possible that this calibration, not the detector itself, is what makes rewording so effective.

Editorial extensions

If this is right

  • A fraudster can evade a deployed audio deepfake detector by paraphrasing the script rather than manipulating the audio, so anti-spoofing evaluations that only add acoustic noise miss a practical attack surface.
  • Detector–voice pairing matters: the same detector can be nearly immune for one synthetic voice and highly vulnerable for another, so robustness claims should be reported per text-to-speech system and voice.
  • Commercial services are not exempt: one commercial API in the paper's tests falls from 100% to 32% accuracy under transcript attacks, and the celebrity-scam case study flips spoof utterances from below 1% to 58–90% bona-fide probability.
  • Feature analysis points to a usable defense signal: detectors with high Audio Encoder Similarity for a voice and balanced spoof and bona-fide F1 scores are harder to attack, so these properties could guide model selection or adaptation.
  • The predictive models trained on the engineered features reach about 76% F1 in approximating detector decisions, suggesting gray-box or black-box attackers could optimize transcript edits through a proxy without querying the real detector.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the paper does not test is whether a transcript-only proxy, using linguistic features like readability and perplexity change without querying the detector, can select effective rewording, which would let an attacker operate fully offline.
  • The finding implies that training corpora with systematic linguistic differences between spoofed and real speech, as the paper shows for the ASVSpoof 2019 data, may teach detectors a linguistic prior; paraphrasing toward the bona-fide register could transfer across text-to-speech voices.
  • If the effect is driven by acoustic embedding geometry, a defense could regularize detectors to keep embeddings of paraphrases close, or text-to-speech pipelines could normalize prosody across synonym substitutions; neither direction is explored in the paper.
  • The paper leaves open whether human listeners would notice the edited transcripts; a user study testing naturalness and suspicion would clarify whether the attack is realistic in live conversation rather than only in automated detection pipelines.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper introduces transcript-level adversarial attacks against audio anti-spoofing detectors (AASs). The authors formulate a black-box optimization problem in which a transcript is minimally perturbed with synonym or MLM-based substitutions, synthesized through a TTS model, and scored by the target detector. They evaluate three open-source detectors (AASIST-2, CLAD, RawNet-2) and two commercial APIs across multiple TTS voices, reporting attack success rates above 60% for several open-source pairs and a drop for one commercial API from 100% to 32% accuracy. They also present a logistic-regression feature analysis over linguistic, acoustic, and model-level features, and a case study simulating the Brad Pitt voice-cloning scam. The central claim is that AASs are vulnerable to minor transcript-level linguistic perturbations, not just acoustic manipulations.

Significance. If substantiated, the paper identifies a genuinely under-explored attack surface: transcript variation that propagates through TTS synthesis and flips the decisions of state-of-the-art anti-spoofing systems. The work is broad in scope: multiple detectors, TTS models, voice profiles, and perturbation strategies, plus a black-box attack formulation and a concrete commercial-API case study. The commercial API-A result (100% to 32%) and the case study provide vivid, practically relevant evidence. The feature analysis is a useful first step toward explainable vulnerability, and the authors are explicit about limitations. However, the quantitative open-source claims currently rest on a per-voice batch-norm calibration protocol whose effect is not isolated, so the strength of the evidence is not yet commensurate with the broad conclusions.

major comments (2)
  1. [Appendix B.1; Tables 3 and 5] The open-source results are measured on detectors that are batch-norm calibrated to each TTS voice until detection accuracy exceeds 90% on a small non-overlapping set. The paper reports no comparison with the uncalibrated detectors, no calibration-set size or selection procedure, and no sensitivity analysis around the 90% threshold. This matters because calibration moves the decision boundary toward the synthetic-voice manifold, and the transcript substitutions may be exploiting a boundary created by the adaptation protocol rather than an intrinsic property of the detectors. Since all 108 open-source experiments and the feature regression in Table 5 use these adapted models, the reported ASR values are not yet grounded as evidence of inherent linguistic sensitivity. Please report (a) attack success rates on uncalibrated detectors, (b) the calibration-set size and how it is sampled, and (c) how ASR varies with the calibration accuracy target. The commercial API-A result in Table 4 is independent of this issue and would keep the qualitative finding alive, but the open-source numbers and the 'from over 90% to just below 20%' contribution claim require this comparison.
  2. [Table A5 (CLAD, American Male, PWWS row)] The CLAD American Male PWWS row reports OC = 34.0%, while the other three attack methods for the same detector-voice pair report OC = 99.8%. Original accuracy before attack should be identical across attack methods because it is measured on the same unperturbed transcripts. This entry appears erroneous, and it propagates to the averaged row in Table 3 (OC = 83.3%, ASR = 56.6%) and to aggregate statements about detector vulnerability. Please correct the entry and recompute all affected averages and any claims derived from them. The presence of this inconsistency also raises a data-integrity concern for the other appendix tables.
minor comments (5)
  1. [Eq. (2), Section 4] The semantic preservation constraint depends on a threshold δ, but the paper never reports the value of δ or the sensitivity of ASR to this threshold. Please state δ and, ideally, show how attack success varies with δ.
  2. [Table 4, Section 5.2.2] The commercial-API results are based on 100 randomly sampled items per pair, but no confidence intervals are reported. Given the headline claim of a drop from 100% to 32%, exact binomial confidence intervals would help quantify the uncertainty, especially for the 100% entries.
  3. [Section 3, Problem Formulation] In the definition of the audio input, the text says 'L is the wavelength of an audio'; this should be 'length of an audio.'
  4. [Appendix B.4] The appendix headers and text use both 'RawNet-2' and 'Rawnet-2'; please standardize the spelling.
  5. [Section 6, Table 5] The Spoof F1 and Bona-fide F1 features are computed on the same VoiceWukong evaluation set and are therefore strongly tied to overall detector quality. Their high-magnitude coefficients are not surprising and should be interpreted as descriptive of detector quality rather than as evidence of a distinct linguistic mechanism.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: all central claims are empirical measurements; the batch-norm calibration is a protocol choice, not a fitted prediction.

full rationale

The paper's central claims are attack success rates and accuracy drops measured empirically by running transcript perturbations through TTS and anti-spoofing detectors. The attack algorithm (Alg. 1) optimizes substitutions using detector outputs, and the reported ASR/AUA values are held-out measurements, not quantities derived from fitted parameters. The only potentially concerning passage is Appendix B.1, where open-source detectors are adapted via batch-norm calibration until detection accuracy exceeds 90%; however, calibration is fit to unperturbed detection accuracy, not to attack outcomes, so the measured ASR is not forced by the calibration procedure. This is a methodological robustness concern, not circularity. The feature-attribution analysis (Table 5, Table 7) uses logistic regression and surrogate models to explain detector decisions, but these models are not used to produce the headline ASR numbers and do not feed back into the attack success computation. The self-citations (Le et al. 2022; Uchendu et al. 2023) are background motivation for text-domain attacks and are not load-bearing for the audio-specific empirical results. No uniqueness theorem is imported, no ansatz is smuggled via citation, and no known result is merely renamed. The derivation is self-contained and the conclusions rest on direct experimentation.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces new metrics (Audio Encoder Similarity, DTW distance, readability and aesthetics deltas) but no new physical or model entities, so no invented-entity burden applies.

free parameters (2)
  • Semantic similarity threshold delta (Eq. 2) = not reported
    Controls how much meaning can change between original and adversarial transcript; every attack result depends on it, but the value is never given.
  • Batch-norm calibration accuracy target = 90%
    Detectors are adapted to each TTS voice until original accuracy exceeds 90% (Appendix B.1). This hand-chosen threshold selects the evaluated model states and may affect measured vulnerability.
assumptions (4)
  • ad hoc to paper Batch-norm calibration on a small non-overlapping set is a fair deployment procedure that does not change the detector's linguistic sensitivity.
    Invoked in Section 5.1 and Appendix B.1. If the calibration itself induces sensitivity, the attack rates are inflated.
  • domain assumption Universal Sentence Encoder cosine similarity above an unreported threshold is a sufficient proxy for semantic preservation and human imperceptibility.
    Used in Section 4 Step 3, Eq. (2). The Limitation section admits human detectability was not measured.
  • domain assumption Attacker has black-box query access to the target anti-spoofing system and can synthesize speech from arbitrary transcripts.
    Standard threat model for the algorithm in Section 4; the attack requires querying F(G(T')) for each candidate replacement.
  • domain assumption Linguistic disparities between spoof and bona-fide samples in ASVSpoof 2019 LA transfer to the evaluated models and explain vulnerability.
    Used as motivation in Section 2.2 and as interpretation in Section 6.2, but not tested as a causal mechanism.

how reviews work

0 comments
Cite this review

Pith. "Pith review of What You Read Isn't What You Hear: Linguistic Sensitivity in Deepfake Speech Detection." pith.science (2026). https://pith.science/paper/UZEM3H2Q

@misc{pith2026250517513,
  author       = {Pith},
  title        = {Pith review of: What You Read Isn't What You Hear: Linguistic Sensitivity in Deepfake Speech Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UZEM3H2Q}},
  note         = {Machine review of arXiv:2505.17513}
}
read the original abstract

Recent advances in text-to-speech technologies have enabled realistic voice generation, fueling audio-based deepfake attacks such as fraud and impersonation. While audio anti-spoofing systems are critical for detecting such threats, prior work has predominantly focused on acoustic-level perturbations, leaving the impact of linguistic variation largely unexplored. In this paper, we investigate the linguistic sensitivity of both open-source and commercial anti-spoofing detectors by introducing transcript-level adversarial attacks. Our extensive evaluation reveals that even minor linguistic perturbations can significantly degrade detection accuracy: attack success rates surpass 60% on several open-source detector-voice pairs, and notably one commercial detection accuracy drops from 100% on synthetic audio to just 32%. Through a comprehensive feature attribution analysis, we identify that both linguistic complexity and model-level audio embedding similarity contribute strongly to detector vulnerability. We further demonstrate the real-world risk via a case study replicating the Brad Pitt audio deepfake scam, using transcript adversarial attacks to completely bypass commercial detectors. These results highlight the need to move beyond purely acoustic defenses and account for linguistic variation in the design of robust anti-spoofing systems. All source code will be publicly available.

Figures

Figures reproduced from arXiv: 2505.17513 by the authors.

Figure 1
Figure 1. Linguistic variation of the transcript can swing [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Feature impact on bona-fide probability prediction. A positive effect means the feature increases the [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

31 extracted references · 8 canonical work pages

  1. [1]

    Rohit Arora, Anmol Arora, and Rohit Singh Rathore. 2022. https://arxiv.org/abs/2109.14900 Impact of channel variation on one-class learning for spoof detection . Preprint, arXiv:2109.14900

    work page Pith review arXiv 2022

  2. [2]

    Luigi Attorresi, Davide Salvi, Clara Borrelli, Paolo Bestagini, and Stefano Tubaro. 2022. https://arxiv.org/abs/2210.17222 Combining automatic speaker verification and prosody analysis for synthetic speech detection . Preprint, arXiv:2210.17222

    work page Pith review arXiv 2022

  3. [3]

    Zhongjie Ba, Qing Wen, Peng Cheng, Yuwei Wang, Feng Lin, Li Lu, and Zhenguang Liu. 2023. https://doi.org/10.1145/3543507.3583222 Transferring audio deepfake detection capability across languages . In Proceedings of the ACM Web Conference 2023, WWW '23, page 2033–2044, New York, NY, USA. Association for Computing Machinery

    arXiv 2023

  4. [4]

    John, Noah Constant, Mario Guajardo-Cespedes, Steve Yuan, Chris Tar, Yun-Hsuan Sung, Brian Strope, and Ray Kurzweil

    Daniel Cer, Yinfei Yang, Sheng yi Kong, Nan Hua, Nicole Limtiaco, Rhomni St. John, Noah Constant, Mario Guajardo-Cespedes, Steve Yuan, Chris Tar, Yun-Hsuan Sung, Brian Strope, and Ray Kurzweil. 2018. https://arxiv.org/abs/1803.11175 Universal sentence encoder . Preprint, arXiv:1803.11175

    arXiv 2018

  5. [5]

    Yushen Chen, Zhikang Niu, Ziyang Ma, Keqi Deng, Chunhui Wang, Jian Zhao, Kai Yu, and Xie Chen. 2024. https://arxiv.org/abs/2410.06885 F5-tts: A fairytaler that fakes fluent and faithful speech with flow matching . Preprint, arXiv:2410.06885

    arXiv 2024

  6. [6]

    Chantal Da Silva. 2024. French woman falls victim to online romance scam by ai brad pitt, report says. https://www.nbcnews.com/news/world/ai-brad-pitt-woman-romance-scam-france-tf1-rcna187745. NBC News, Accessed: 2024-06-13

    work page 2024

  7. [7]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://arxiv.org/abs/1810.04805 Bert: Pre-training of deep bidirectional transformers for language understanding . Preprint, arXiv:1810.04805

    arXiv 2019

  8. [8]

    Siddhant Garg and Goutham Ramakrishnan. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.498 BAE : BERT -based adversarial examples for text classification . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 6174--6181, Online. Association for Computational Linguistics

    doi:10.18653/v1/2020.emnlp-main.498 2020

Show all 31 references
  1. [9]

    Di Jin, Zhijing Jin, Joey Tianyi Zhou, and Peter Szolovits. 2020. https://arxiv.org/abs/1907.11932 Is bert really robust? a strong baseline for natural language attack on text classification and entailment . Preprint, arXiv:1907.11932

    2020 arXiv

  2. [10]

    Jee-weon Jung Jung, Hee-Soo Heo, Hemlata Tak, Hye jin Shim, Joon Son Chung, Bong-Jin Lee, Ha-Jin Yu, and Nicholas Evans. 2021. https://arxiv.org/abs/2110.01200 Aasist: Audio anti-spoofing using integrated spectro-temporal graph attention networks . Preprint, arXiv:2110.01200

    2021 arXiv

  3. [11]

    Thai Le, Jooyoung Lee, Kevin Yen, Yifan Hu, and Dongwon Lee. 2022. https://doi.org/10.18653/v1/2022.findings-acl.232 Perturbations in the wild: Leveraging human-written text perturbations for realistic adversarial attack and defense . In Findings of the Association for Computa...

    2022 doi

  4. [12]

    Linyang Li, Ruotian Ma, Qipeng Guo, Xiangyang Xue, and Xipeng Qiu. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.500 BERT - ATTACK : Adversarial attack against BERT using BERT . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNL...

    2020 doi

  5. [13]

    So-jeong Lim Lim, Adam Jatowt, and Masatoshi Yoshikawa. 2018. https://api.semanticscholar.org/CorpusID:208194893 Understanding characteristics of biased sentences in news articles . In CIKM Workshops

    2018

  6. [14]

    Florian Lux, Sarina Meyer, Lyonel Behringer, Frank Zalkow, Phat Do, Matt Coler, Emanuël A. P. Habets, and Ngoc Thang Vu. 2024. https://arxiv.org/abs/2406.06403 Meta learning text-to-speech synthesis in over 7000 languages . Preprint, arXiv:2406.06403

    2024 arXiv

  7. [15]

    John Morris, Eli Lifland, Jin Yong Yoo, Jake Grigsby, Di Jin, and Yanjun Qi. 2020. Textattack: A framework for adversarial attacks, data augmentation, and adversarial training in nlp. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: Sy...

    2020

  8. [16]

    Müller, Philip Sperl, and Konstantin Böttinger

    Nicolas M. Müller, Philip Sperl, and Konstantin Böttinger. 2023. https://arxiv.org/abs/2308.11800 Complex-valued neural networks for voice anti-spoofing . Preprint, arXiv:2308.11800

    2023 arXiv

  9. [17]

    Uwe Peters Peters. 2024. https://doi.org/10.1080/09515089.2023.2261976 The philosophical debate on linguistic bias: A critical perspective . Philosophical Psychology, 37(6):1513--1538

    2024

  10. [18]

    Shuhuai Ren, Yihe Deng, Kun He, and Wanxiang Che. 2019. https://doi.org/10.18653/v1/P19-1103 Generating natural language adversarial examples through probability weighted word saliency . In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics...

    2019 doi

  11. [19]

    Gil Shomron and Uri Weiser. 2020. https://arxiv.org/abs/2010.05625 Post-training batchnorm recalibration . Preprint, arXiv:2010.05625

    2020 arXiv

  12. [20]

    Signicat. 2024. Fraud attempts with deepfakes have increased by 2137\ https://www.signicat.com/press-releases/fraud-attempts-with-deepfakes-have-increased-by-2137-over-the-last-three-year. Signicat, Accessed: 2024-06-13

    2024

  13. [21]

    Hemlata Tak, Jose Patino, Massimiliano Todisco, Andreas Nautsch, Nicholas Evans, and Anthony Larcher. 2021. https://arxiv.org/abs/2011.01108 End-to-end anti-spoofing with rawnet2 . Preprint, arXiv:2011.01108

    2021 arXiv

  14. [22]

    Hemlata Tak, Massimiliano Todisco, Xin Wang, Jee weon Jung, Junichi Yamagishi, and Nicholas Evans. 2022. https://arxiv.org/abs/2202.12233 Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation . Preprint, arXiv:2202.12233

    2022 arXiv

  15. [23]

    Andros Tjandra, Yi-Chiao Wu, Baishan Guo, John Hoffman, Brian Ellis, Apoorv Vyas, Bowen Shi, Sanyuan Chen, Matt Le, Nick Zacharov, Carleigh Wood, Ann Lee, and Wei-Ning Hsu. 2025. https://arxiv.org/abs/2502.05139 Meta audiobox aesthetics: Unified automatic quality assessment fo...

    2025 arXiv

  16. [24]

    Adaku Uchendu, Thai Le, and Dongwon Lee. 2023. Attribution and obfuscation of neural text authorship: A data mining perspective. ACM SIGKDD Explorations Newsletter, 25(1):1--18

    2023

  17. [25]

    Haibin Wu, Songxiang Liu, Helen Meng, and Hung yi Lee. 2020. https://arxiv.org/abs/2003.03065 Defense against adversarial attacks on spoofing countermeasures of asv . Preprint, arXiv:2003.03065

    2020 arXiv

  18. [26]

    Haolin Wu, Jing Chen, Ruiying Du, Cong Wu, Kun He, Xingcan Shang, Hao Ren, and Guowen Xu. 2024. https://arxiv.org/abs/2404.15854 Clad: Robust audio deepfake detection against manipulation attacks with contrastive learning . Preprint, arXiv:2404.15854

    2024 arXiv

  19. [27]

    Junichi Yamagishi, Christophe Veaux, and Kirsten MacDonald. 2019. https://doi.org/10.7488/ds/2645 CSTR VCTK Corpus: English Multi-speaker Corpus for CSTR Voice Cloning Toolkit (version 0.92), [sound] . https://doi.org/10.7488/ds/2645. University of Edinburgh. The Centre for Sp...

    2019 doi

  20. [28]

    Ziwei Yan, Yanjie Zhao, and Haoyu Wang. 2024. https://arxiv.org/abs/2409.06348 Voicewukong: Benchmarking deepfake voice detection . Preprint, arXiv:2409.06348

    2024 arXiv

  21. [29]

    Jian Zhu, Cong Zhang, and David Jurgens. 2022. https://arxiv.org/abs/2204.03067 Byt5 model for massively multilingual grapheme-to-phoneme conversion . Preprint, arXiv:2204.03067

    2022 arXiv

  22. [30]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  23. [31]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.