Pith. sign in

REVIEW 4 major objections 5 minor 34 references

From Scarcity to Capability: Empowering Fake News Detection in Low-Resource Languages with LLMs

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A fine-tuned BLOOM 560M model reaches 89% macro F1 on Bangla fake-news detection, powered by a new 60,000-article dataset that the paper introduces.

desk verdict Useful dataset buried under self-contradictory numbers: the paper's own external test shows SVM beating BLOOM, and the tables don't add up. read the letter →

arxiv 2501.09604 v1 pith:GU3WJGNE submitted 2025-01-16 cs.CL

Hrithik Majumdar Shibu , Shrestha Datta , Md. Sumon Miah , Nasrullah Sami , Mahruba Sharmin Chowdhury , Md. Saiful Islam This is my paper · ORCID
classification cs.CL
keywords Banglafakenewslow-resourcelanguagesdetectionBanNews-2.0BLOOMfine-tuningQLoRAtransformerbenchmarkdatasetconstruction
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 introduces BanFakeNews-2.0, a 60,000-article Bangla fake-news dataset with 13,000 verified fake items and 47,000 authentic items across 13 categories. It argues that this larger, more balanced resource lets transformer models and fine-tuned large language models detect Bangla fake news substantially better than traditional linguistic-feature classifiers and better than models trained on the original BanFakeNews dataset. The best result is a fine-tuned BLOOM 560M model reaching 89% macro F1 on an independent 1,000-article test set, against 86% for the strongest SVM baseline. The paper also releases the dataset and models to support further work on low-resource fake-news detection.

What carries the argument

The central object is BanFakeNews-2.0, a 60,000-article Bangla news corpus in which the fake class is built from articles debunked by fact-checking sites and the authentic class from 30 mainstream Bangladeshi news portals, with labels cross-checked by three annotators using majority voting (inter-annotator agreement 0.93). The other load-bearing mechanism is QLoRA fine-tuning, which adapts a 560M-parameter BLOOM model in 4-bit precision with low-rank adapters, letting a small language model reach an 89% macro F1 score. The dataset's category balance and duplicate removal (dropping items with over 50% token overlap) are what the paper credits for the improved generalization seen on the independent test set.

What would settle it

Manually audit a random sample of the 13,000 fake articles by checking each against the fact-checking source cited in Table 6; if a substantial portion cannot be traced to a debunked claim, the dataset's labels do not support the reported F1 scores. Alternatively, compute n-gram or document-level overlap between the 1,000-article test set and the training portion; if the overlap is high, the external 'generalization' result in Table 4 is an artifact of test contamination.

Watch

Extended reading notes

Core claim

BanFakeNews-2.0 extends the earlier BanFakeNews resource by adding 11,700 manually verified fake articles gathered from fact-checking platforms, producing a 60,000-item corpus with a 47,000-to-13,000 authentic-to-fake split. The paper claims that this proportional dataset, combined with an independently curated 1,000-article test set, enables fine-tuned transformer and LLM classifiers to outperform SVM models using TF-IDF, character n-grams, and FastText embeddings. Specifically, fine-tuning BLOOM 560M with QLoRA achieves an 89% macro F1 score, and models trained on BanFakeNews-2.0 generalize better on the external test set than models trained on the original BanFakeNews (macro F1 67 vs 29 for BLOOM, 91 vs 39 for SVM). The authors take this as evidence that dataset scale and balance, not just model architecture, are the key to moving low-resource languages from scarcity to capability.

Load-bearing premise

The load-bearing premise is that the ground-truth 'fake' labels, drawn from fact-checking sites and blogs, are correct and that the 1,000-article external test set is genuinely independent and representative of the real Bangla news stream.

Editorial extensions

If this is right

  • A larger, more balanced Bangla fake-news corpus lets fine-tuned small LLMs reach 89% macro F1, compared with 86% for the best SVM baseline and 78% for a majority classifier.
  • Models trained on BanFakeNews-2.0 generalize better to a manually curated external test set than models trained on the original BanFakeNews (SVM: 91 vs 39 macro F1; BLOOM: 67 vs 29).
  • Character n-gram features (C3-C5) and word unigrams are stronger traditional baselines than bigrams or trigrams for Bangla fake-news detection.
  • The public release of dataset and models provides a reproducible benchmark for future Bangla fake-news detection research.

Reading between the lines

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

  • Because the fake-news class is sourced from fact-checking and debunking sites, the dataset may under-represent fake stories that have not been publicly debunked; a live deployment would need to test on never-seen rumor types.
  • The 89% macro F1 hides a class asymmetry: authentic-news F1 is 96 while fake-news F1 is 81, so a more balanced fake/authentic split or a focus on the fake class could raise practical usefulness for moderation.
  • The QLoRA fine-tuning recipe (4-bit quantization, rank 8, alpha 32) is a parameter-efficient template that could be transferred to other low-resource languages with similar dataset construction.
  • Releasing the external test set is only useful if its independence is audited; a simple overlap check between test articles and training sources would make the claimed generalization much easier to verify.
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

4 major / 5 minor

Summary. This paper introduces BanFakeNews-2.0, an extension of the authors' earlier BanFakeNews dataset, claimed to contain 60,000 Bangla news articles (47,000 authentic and 13,000 fake) across 13 categories, together with a manually curated independent test set of 1,000 articles. The authors benchmark several models on this resource: TF-IDF character and word n-grams with linear SVM, five BERT-family transformers, and four LLMs fine-tuned with QLoRA, and report precision, recall, and F1 for authentic and fake classes as well as macro F1. The abstract's headline claim is that fine-tuned BLOOM 560M achieves the highest macro F1 of 89 and significantly outperforms traditional methods, and the paper further claims that models trained on BanFakeNews-2.0 generalize better than those trained on the original BanFakeNews. The manuscript also includes a brief limitations section acknowledging dataset imbalance and the future threat of generative-model fake news.

Significance. If the benchmark were reliable, the release of a 60,000-article Bangla fake news dataset with an independent test set would be a useful contribution for low-resource NLP, and the comparison of QLoRA-fine-tuned LLMs with traditional classifiers would interest the community. The authors' commitment to publicly release the dataset and models is commendable. However, the reported evaluations contain internal contradictions that prevent the results from being relied upon, and the headline claim about BLOOM's superiority is directly contradicted by the paper's own external evaluation.

major comments (4)
  1. [Abstract and Section 5, Table 4] The abstract and Introduction claim that fine-tuned BLOOM 560M 'achieves the highest performance, with a macro F1 score of 89' and 'significantly outperforms traditional methods.' Table 4 shows the opposite on the independently curated external test set: SVM (All) trained on BanFakeNews-2.0 reaches macro F1 91, while BLOOM trained on the same data reaches 67. Since Table 4 is presented as the rigorous cross-comparison for generalization, this internal contradiction is load-bearing and the headline claim is unsupported.
  2. [Table 3] The macro F1 values for the Majority and Random baselines are arithmetically impossible under the standard equal-weight macro averaging. The Majority baseline has per-class F1 of 88 for authentic and 0 for fake, so the macro average is 44, not the reported 78. The Random baseline has per-class F1s 61 and 30, giving 45.5, not 63. Unless a different averaging scheme is intended, the table's baseline numbers cannot be correct; if a different metric is used, it must be explicitly defined.
  3. [Table 2, Abstract, and Section 2] The category counts in Table 2 sum to 45,839 authentic and 13,993 fake articles (59,832 total), not the stated 47,000 authentic, 13,000 fake, and 60,000 total. Either the table or the headline numbers are wrong. This discrepancy affects the central description of the resource and must be resolved before the dataset can be trusted. The claimed addition of 11,700 fake articles to the original 1,300 also yields 13,000, not 13,993.
  4. [Section 2 and Abstract (external test set)] The paper repeatedly relies on a 'manually curated independent test set of 1,000 news articles' for the external evaluation in Table 4, but Section 2 provides no details about how this test set was constructed, selected, or annotated. No inter-annotator agreement is reported for this test set, and the source list in Table 6 includes blogs and satirical sites without explaining how they were validated as providers of fake news. Without this information, the credibility of the external evaluation is unestablished.
minor comments (5)
  1. [Section 2] The sentence 'we filtered out duplicates and removed items with over 50% or 300 words of token overlap' is ambiguous; please specify whether the threshold is 50% overlap or 300 words, and how overlap is computed.
  2. [Section 3.3] The statement that 'alpha is used as a regularization parameter' is not accurate for QLoRA, where alpha is a scaling hyperparameter for the low-rank adaptation, not a regularization parameter in the usual sense.
  3. [Table 1] The dataset name appears as 'BanFakeNews-2' in Table 1 and as 'BanFakeNews-2.0' elsewhere; please use one consistent name throughout.
  4. [Section 4.1] The validation procedure should clarify whether hyperparameters were selected on the validation half before evaluating on the test half; as written, the split description does not rule out leakage of validation information into the reported test numbers.
  5. [Table 4] The column heading 'Mac. F1' should be spelled out as 'Macro F1' and the averaging scheme should be defined somewhere in the experimental setup.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical benchmark with held-out evaluation; its internal/external contradiction is a correctness issue, not a circular derivation.

full rationale

This paper is an empirical benchmark and dataset contribution, not a derivation chain. The headline result (BLOOM 560M macro F1 89, Abstract and Table 3) is a measured number on a 30% holdout split of the proposed dataset; hyperparameters were chosen on a validation subset (Section 4.1 and Section 4.3), so the test numbers are not fitted inputs renamed as predictions. The external 1,000-article test set is presented as an independent cross-check (Abstract; Section 5; Table 4), which is the opposite of circular: it can and does contradict the internal ranking, with SVM at 91 vs BLOOM at 67. That contradiction is a serious validity/correctness problem for the claim that BLOOM 'significantly outperforms traditional methods,' but it is not circularity. The paper builds on the authors' own earlier BanFakeNews dataset (Hossain et al., 2020), and several authors overlap, but this is normal incremental research: BanFakeNews-2.0 is a newly collected artifact, and the evaluation is against held-out and external data rather than against the cited prior work's outputs. No uniqueness theorem, ansatz-by-citation, or self-definitional equivalence is used. The dataset count inconsistency (Table 2 sums to 59,832 vs the stated 60,000; 45,839 authentic vs 47,000) is a reporting/arithmetic error, not a circular step. I find no circular step.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper's central claims rest on the accuracy of the dataset labels and the independence of the test split. There are no theoretical derivations, so free parameters are experimental choices and curation thresholds. The label accuracy and source credibility are assumed rather than independently verified.

free parameters (4)
  • QLoRA rank and alpha = rank=8, alpha=32
    Chosen by hand in §3.3; these hyperparameters directly affect the reported BLOOM F1 of 89.
  • Deduplication threshold = 50% token overlap or 300 words
    Chosen in §2; affects dataset composition and the separation between train and test sets.
  • Maximum input tokens = 512
    Set in §4.3; standard transformer constraint, influences which articles can be used in full.
  • N-gram ranges = char: 3,4,5; word: 1,2,3
    Chosen in §3.1 for the feature-based SVM baselines.
assumptions (5)
  • domain assumption Fact-checking platforms (e.g., bdfactcheck.com, jaachai.com) provide accurate verdicts on which articles are fake.
    The fake news labels are derived from these sources (§2). If these platforms are wrong, the dataset labels are wrong.
  • domain assumption The top 30 Bangladeshi news portals listed in Table 5 publish only authentic news.
    Authentic labels are assigned based on source credibility (§2).
  • domain assumption Majority voting by three annotators yields the correct class label.
    Ground truth for the independent test set and new fake labels is based on this protocol (§2).
  • domain assumption The deduplication rule (removing >50% overlap or >300-word overlap) prevents train-test leakage.
    The paper assumes this ensures the external test set is independent (§2, §4).
  • domain assumption The independent test set of 1,000 articles is representative of real-world Bangla news.
    Used to claim generalization in Table 4; no sampling strategy is described (§4, §5).

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Scarcity to Capability: Empowering Fake News Detection in Low-Resource Languages with LLMs." pith.science (2026). https://pith.science/paper/GU3WJGNE

@misc{pith2026250109604,
  author       = {Pith},
  title        = {Pith review of: From Scarcity to Capability: Empowering Fake News Detection in Low-Resource Languages with LLMs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GU3WJGNE}},
  note         = {Machine review of arXiv:2501.09604}
}
read the original abstract

The rapid spread of fake news presents a significant global challenge, particularly in low-resource languages like Bangla, which lack adequate datasets and detection tools. Although manual fact-checking is accurate, it is expensive and slow to prevent the dissemination of fake news. Addressing this gap, we introduce BanFakeNews-2.0, a robust dataset to enhance Bangla fake news detection. This version includes 11,700 additional, meticulously curated fake news articles validated from credible sources, creating a proportional dataset of 47,000 authentic and 13,000 fake news items across 13 categories. In addition, we created a manually curated independent test set of 460 fake and 540 authentic news items for rigorous evaluation. We invest efforts in collecting fake news from credible sources and manually verified while preserving the linguistic richness. We develop a benchmark system utilizing transformer-based architectures, including fine-tuned Bidirectional Encoder Representations from Transformers variants (F1-87\%) and Large Language Models with Quantized Low-Rank Approximation (F1-89\%), that significantly outperforms traditional methods. BanFakeNews-2.0 offers a valuable resource to advance research and application in fake news detection for low-resourced languages. We publicly release our dataset and model on Github to foster research in this direction.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 20 canonical work pages

  1. [1]

    Hewett, Jamie Huynh, Mojan Javaheripi, Xin Jin, Piero Kauffmann, Nikos Karampatziakis, Dongwoo Kim, Mahoud Khademi, Lev Kurilenko, James R

    Marah Abdin, Sam Ade Jacobs, Ammar Ahmad Awan, Jyoti Aneja, Ahmed Awadallah, Hany Awadalla, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Jianmin Bao, Harkirat Behl, Alon Benhaim, Misha Bilenko, Johan Bjorck, Sébastien Bubeck, Qin Cai, Martin Cai, Caio César Teodoro Mendes, Weizhu Chen, Vishrav Chaudhary, Dong Chen, Dongdong Chen, Yen-Chun Chen, Yi-Ling Chen...

    arXiv 2024

  2. [2]

    Sayeed Al-Zaman and Mridha Md

    Md. Sayeed Al-Zaman and Mridha Md. Shiblee Noman. 2023. https://doi.org/10.1016/j.dib.2023.109439 A dataset on social media users’ engagement with religious misinformation . Data in Brief, 49:109439

    arXiv 2023

  3. [3]

    R a zvan Andonie. 2019. Hyperparameter optimization in learning systems. Journal of Membrane Computing, 1(4):279--291

    work page 2019

  4. [4]

    Marco Bellagente, Jonathan Tow, Dakota Mahan, Duy Phung, Maksym Zhuravinskyi, Reshinth Adithyan, James Baicoianu, Ben Brooks, Nathan Cooper, Ashish Datta, et al. 2024. Stable lm 2 1.6 b technical report. arXiv preprint arXiv:2402.17834

    arXiv 2024

  5. [5]

    Sohel Rahman, and Rifat Shahriyar

    Abhik Bhattacharjee, Tahmid Hasan, Wasi Ahmad, Kazi Samin Mubasshir, Md Saiful Islam, Anindya Iqbal, M. Sohel Rahman, and Rifat Shahriyar. 2022. https://aclanthology.org/2022.findings-naacl.98 B angla BERT : Language model pretraining and benchmarks for low-resource language understanding evaluation in B angla . In Findings of the Association for Computat...

    2022

  6. [6]

    Pragyananda Bhikkhu. 2014. https://en.prothomalo.com/opinion/Who-will-be-tried-for-Ramu-destruction Who will be tried for ramu destruction? Published: 30 Sep 2014, 16: 58

    work page 2014

  7. [7]

    Aengus Bridgman, Eric Merkley, Peter John Loewen, Taylor Owen, Derek Ruths, Lisa Teichmann, and Oleg Zhilin. 2020. The causes and consequences of covid-19 misperceptions: Understanding the role of news and social media. Harvard Kennedy School Misinformation Review, 1(3)

    work page 2020

  8. [8]

    Juan Cao, Peng Qi, Qiang Sheng, Tianyun Yang, Junbo Guo, and Jintao Li. 2020. Exploring the role of visual content in fake news detection. Disinformation, Misinformation, and Fake News in Social Media: Emerging Research Challenges and Opportunities, pages 141--161

    work page 2020

Show all 34 references
  1. [9]

    Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzm \' a n, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2019. https://arxiv.org/abs/1911.02116 Unsupervised cross-lingual representation learning at scale . C...

    2019 arXiv

  2. [10]

    Saiful Islam, and Md

    Sowmen Das, Md. Saiful Islam, and Md. Ruhul Amin. 2022. Gca-net: Utilizing gated context attention for improving image forgery localization and detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pages 81--90

    2022

  3. [11]

    Saiful Islam, and Md

    Sowmen Das, Selim Seferbekov, Arup Datta, Md. Saiful Islam, and Md. Ruhul Amin. 2021. Towards solving the deepfake problem: An analysis on improving deepfake detection using dynamic face augmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (...

    2021

  4. [13]

    Jacob Devlin, Ming - Wei Chang, Kenton Lee, and Kristina Toutanova. 2018 b . https://arxiv.org/abs/1810.04805 BERT: pre-training of deep bidirectional transformers for language understanding . CoRR, abs/1810.04805

    2018 arXiv

  5. [14]

    Joseph L Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological bulletin, 76(5):378

    1971

  6. [15]

    Hearst, Susan T Dumais, Edgar Osuna, John Platt, and Bernhard Scholkopf

    Marti A. Hearst, Susan T Dumais, Edgar Osuna, John Platt, and Bernhard Scholkopf. 1998. Support vector machines. IEEE Intelligent Systems and their applications, 13(4):18--28

    1998

  7. [16]

    Md Zobaer Hossain, Md Ashraful Rahman, Md Saiful Islam, and Sudipta Kar. 2020. https://aclanthology.org/2020.lrec-1.349 B an F ake N ews: A dataset for detecting fake news in B angla . In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 2862--2871...

    2020

  8. [17]

    Md Gulzar Hussain, Md Rashidul Hasan, Mahmuda Rahman, Joy Protim, and Sakib Al Hasan. 2020. https://doi.org/10.1109/iCCECE49321.2020.9231167 Detection of bangla fake news using mnb and svm classifier . In 2020 International Conference on Computing, Electronics & Communications...

    2020

  9. [18]

    Hakan Inan, Kartikeya Upasani, Jianfeng Chi, Rashi Rungta, Krithika Iyer, Yuning Mao, Michael Tontchev, Qing Hu, Brian Fuller, Davide Testuggine, and Madian Khabsa. 2023. https://arxiv.org/abs/2312.06674 Llama guard: Llm-based input-output safeguard for human-ai conversations ...

    2023 arXiv

  10. [19]

    Khondoker Ittehadul Islam, Tanvir Yuvraz, Md Saiful Islam, and Enamul Hassan. 2022. https://aclanthology.org/2022.aacl-short.17 E mo N o B a: A dataset for analyzing fine-grained emotions on noisy B angla texts . In Proceedings of the 2nd Conference of the Asia-Pacific Chapter...

    2022

  11. [20]

    Diederik P Kingma and Jimmy Ba. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

    2014 arXiv

  12. [21]

    Ramez Kouzy, Joseph Abi Jaoude, Afif Kraitem, Molly B El Alam, Basil Karam, Elio Adib, Jabra Zarka, Cindy Traboulsi, Elie W Akl, and Khalil Baddour. 2020. Coronavirus goes viral: quantifying the covid-19 misinformation epidemic on twitter. Cureus, 12(3)

    2020

  13. [22]

    Sun Kyong Lee, Juhyung Sun, Seulki Jang, and Shane Connelly. 2022. Misinformation of covid-19 vaccines and vaccine hesitancy

    2022

  14. [23]

    Reza Moradi, Reza Berangi, and Behrouz Minaei. 2020. A survey of regularization strategies for deep models. Artificial Intelligence Review, 53(6):3947--3986

    2020

  15. [24]

    Cathal O’Connor and Michelle Murphy. 2020. Going viral: doctors must tackle fake news in the covid-19 pandemic. Bmj, 369(10.1136)

    2020

  16. [25]

    Jiayi Pan, Chengcan Wang, Kaifu Zheng, Yangguang Li, Zhenyu Wang, and Bin Feng. 2023. https://arxiv.org/abs/2312.03788 Smoothquant+: Accurate and efficient 4-bit post-training weightquantization for llm . Preprint, arXiv:2312.03788

    2023 arXiv

  17. [26]

    Nauros Romim, Mosahed Ahmed, Md Saiful Islam, Arnab Sen Sharma, Hriteshwar Talukder, and Mohammad Ruhul Amin. 2022. Bd-shs: A benchmark dataset for learning to detect online bangla hate speech in different social contexts. In Proceedings of the Thirteenth Language Resources an...

    2022

  18. [27]

    SadikAlJarif. 2022. bangla fake news dataset. https://www.kaggle.com/datasets/sadikaljarif/bangla-fake-news-detection-dataset?select=final_bn_data.csv

    2022

  19. [28]

    Sagor Sarker. 2020. https://github.com/sagorbrur/bangla-bert Banglabert: Bengali mask language model for bengali language understanding

    2020

  20. [30]

    Tabiur Rahman Prodhan

    Sharifa Umma Shirina and Md. Tabiur Rahman Prodhan. 2020. https://doi.org/10.34257/GJHSSAVOL20IS17PG11 Spreading fake news in the virtual realm in bangladesh: Assessment of impact . Global Journal of Human-Social Science, 20(A17):11–25

    2020 doi

  21. [31]

    Antoine Simoulin, Namyong Park, Xiaoyi Liu, and Grey Yang. 2024. Memory-efficient fine-tuning of transformers via token selection. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 21565--21580

    2024

  22. [32]

    Bhuvanesh Singh and Dilip Kumar Sharma. 2022. Predicting image credibility in fake news over social media using multi-modal approach. Neural Computing and Applications, 34(24):21503--21517

    2022

  23. [33]

    Borhan Uddin, Nahid Reza, Md Saiful Islam, Hasib Ahsan, and Mohammad Ruhul Amin. 2021. Fighting against fake news during pandemic era: Does providing related news help student internet users to detect covid-19 misinformation? In Proceedings of the 13th ACM Web Science Conferen...

    2021

  24. [34]

    BigScience Workshop, :, Teven Le Scao, Angela Fan, Christopher Akiki, Ellie Pavlick, Suzana Ilić, Daniel Hesslow, Roman Castagné, Alexandra Sasha Luccioni, François Yvon, Matthias Gallé, Jonathan Tow, Alexander M. Rush, Stella Biderman, Albert Webson, Pawan Sasanka Ammanamanch...

    2023 arXiv

  25. [35]

    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...

  26. [36]

    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 10, 2026 · model on record in the stance chip above.