Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

CoheMark: A Novel Sentence-Level Watermark for Enhanced Text Quality

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read CoheMark claims a sentence-level watermark that keeps text readable by steering each sentence into the topic clusters of the previous sentence while maintaining detection rates above 97%.

desk verdict A solid incremental sentence-level watermarking paper whose headline detection numbers are undercut by a cross-model null distribution, but the method is coherent and worth refereeing. read the letter →

arxiv 2504.17309 v1 pith:S5OMT74X submitted 2025-04-24 cs.CL

Junyan Zhang , Shuliang Liu , Aiwei Liu , Yubo Gao , Jungang Li , Xiaojie Gu , Xuming Hu This is my paper · ORCID
classification cs.CL
keywords sentence-levelwatermarkingfuzzyc-meansclusteringcohesiontextqualityLLMrejectionsamplingwatermarkrobustnesssemanticcoherence
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 claims that a sentence-level watermark can be embedded without sacrificing text coherence by making the acceptance of each generated sentence depend on its semantic proximity to the previous sentence. CoheMark clusters a domain's sentence embeddings with fuzzy c-means, represents each sentence by a ranking of cluster memberships, and accepts only sentences whose top cluster falls in the 'green' slots defined by the previous sentence's ranking. On OpenGen and LFQA with OPT-2.7B and Llama-3-8B, the method reports true positive rates at 1 percent false positive of 97 to 100 percent, comparable to the strongest baselines, and pairwise GPT-4o evaluations rate its text above KGW, EXP, Unbiased, and SemStamp on coherence, relevance, interestingness, and integrity. The paper's point is that the quality-strength trade-off in watermarking is not fixed: exploiting inter-sentence cohesion can push both at once.

What carries the argument

The machinery is the pair formed by the FuzzyClusterer and the CoheSampler, linked by the Membership Index: a ranking of a sentence's fuzzy c-means membership degrees across clusters. The FuzzyClusterer learns soft clusters over sentence embeddings from the target domain, and the Membership Index turns any new sentence into a ranked list of cluster affinities. The CoheSampler uses a hand-specified Next Sentence Selection Criterion to read the previous sentence's ranking and declare certain rank positions green; the next sentence is accepted only if its top-ranked cluster sits in one of those positions. The switching between NSSC v1 and v2 is intended to keep text locally coherent while periodically pivoting to less-represented topics. This rejection-sampling loop carries the argument: it replaces arbitrary semantic-region division with a cohesion-guided one.

What would settle it

Replace the hand-picked green clusters with random ones of the same size and re-measure quality and detection: if quality holds, cohesion is not the source of the advantage.

Watch

Extended reading notes

Core claim

The central discovery is that choosing the valid semantic spaces for the next sentence by the membership profile of the previous sentence—rather than by a random red/green split—preserves textual cohesion while maintaining detectability. CoheMark first trains fuzzy c-means clusters on sentences from the target domain, then at each generation step embeds the previous sentence and ranks the clusters by membership degree. The next sentence is generated and re-embedded until its highest-membership cluster falls in the green set defined by the current Next Sentence Selection Criterion: NSSC v1 takes the 1st and 3rd clusters in the previous sentence's ranking, NSSC v2 takes the 2nd, 4th, 5th, and 6th, with a switch after five consecutive matches. Detection checks what fraction of sentences in a text obey the same criterion. The paper reports TPR@1% of 99.3 and 97.3 across models on OpenGen, 100 and 97.3 on LFQA, and GPT-4o soft win rates of 61-95 percent over four baseline families.

Load-bearing premise

The paper's central claim rests on hand-picked rules for which topic clusters count as green for the next sentence, and those rules are never varied or tested.

Editorial extensions

If this is right

  • If CoheMark works as reported, sentence-level watermarks can avoid modifying the model's logit distribution and still exceed 97 percent true positive rate at 1 percent false positive rate.
  • The method makes text quality a design input rather than an afterthought: generated text is kept on the topic established by the previous sentence, which the paper supports with GPT-4o pairwise win rates over KGW, EXP, Unbiased, and SemStamp.
  • CoheMark reports higher true positive rates than the baselines after sentence-level paraphrase attacks, indicating the cohesive selection survives rewording.
  • The runtime cost is real but bounded: with OPT-2.7B on OpenGen, CoheMark averages about 27 seconds per generation, roughly seven times slower than token-level watermarks and about three times faster than SemStamp.
  • The paper argues traditional automatic metrics such as perplexity and BertScore conflict with LLM-based evaluation of watermarked text, so quality conclusions should not rest on the traditional measures alone.

Reading between the lines

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

  • The exact Next Sentence Selection Criteria appear unoptimized: nothing in the paper tests alternatives, so a natural next experiment is varying which rank positions are green and how many matches trigger the switch.
  • If the cohesion hypothesis is the real driver, the fuzzy c-means layer is replaceable by any soft clusterer that yields the same membership rankings, so the method could transfer to other embedding spaces.
  • The robustness evaluation uses a particular sentence-by-sentence paraphrase attack; stronger adversarial rewrites that deliberately break topic continuity would stress the watermark differently.
  • The LLM-judge results are reported without a human preference study, so a direct human evaluation would settle whether the quality advantage is as large as GPT-4o says.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes CoheMark, a sentence-level watermarking method that uses fuzzy c-means clustering over sentence embeddings and hand-crafted next-sentence selection criteria (NSSC v1/v2) to define green/red semantic spaces. Generation proceeds by rejection sampling: a candidate sentence is accepted if its primary membership cluster falls in the current green set, where the green set is derived from the membership index of the previous sentence. Detection computes the proportion of sentences satisfying the rule and thresholds it to achieve a target false-positive rate. The authors report TPR@1% values of 97.3–100% across OPT-2.7B and Llama-3-8B on OpenGen and LFQA, claim competitive or superior text quality based on GPT-4o pairwise preferences, and show robustness to sentence-level paraphrase attacks. The central claims are that CoheMark achieves strong watermark strength while minimally impacting text quality.

Significance. If the detectability result were established with a properly matched null distribution, the idea of using coherence-driven sentence selection for watermarking would be a modest but useful contribution to the sentence-level watermarking literature. The paper is thorough in comparing against six baselines, two base models, and two datasets, and it honestly reports failure rates, runtime, and the absence of a k-SemStamp comparison. However, the current evidence does not yet support the headline claim: the detection metric is computed against a mismatched unwatermarked text distribution, the quality advantage rests on a single 100-prompt GPT-4o preference evaluation with no statistical analysis, and the central NSSC design choices are never ablated. The contribution is plausible but needs substantial additional validation.

major comments (4)
  1. [Section B / Table 1] The TPR@1% values in Table 1 are not comparable across models because the unwatermarked negative set is generated by Mistral-7B-Instruct-v0.3 while the watermarked positives are generated by OPT-2.7B and Llama-3-8B. As described in Section B, the unwatermarked text is produced by asking the aligned Mistral model to 'mimic human-written text', so the 1% FPR threshold is calibrated on Mistral output. Since sentence embeddings and fuzzy-cluster memberships are sensitive to model identity and style, the reported TPR may partly reflect cross-model differences rather than the watermark rule. The authors must report TPR@1% using a same-model unwatermarked negative set (e.g., unwatermarked generations from OPT-2.7B and Llama-3-8B with the same sampling settings) for each model/dataset combination. Until this is done, the central claim of 'strong watermark strength' is not established.
  2. [Section 4 / Table 6] The failure-generation rates of 0.5%–4.5% in Table 6, combined with the maximum-trial stopping rule described in Section A, mean that TPR@1% is computed only on successful generations. If prompts that hit the trial limit are dropped from the detection evaluation, the reported TPRs are conditioned on an outcome that may correlate with text difficulty, length, or topic. The authors should report intent-to-treat rates (counting failures as undetected) or demonstrate that failures are statistically independent of the detection outcome. Without this, the detection numbers may overstate the usable watermark strength in practice.
  3. [Section 4.2 / Table 2] The text-quality claim relies almost entirely on GPT-4o pairwise preference scores over 100 OpenGen prompts, with no confidence intervals, no inter-judge agreement, and no significance tests for the soft win rates. Moreover, the evaluation prompt explicitly asks the judge to assess 'coherence', while CoheMark is explicitly designed to enforce topical overlap via NSSC; the evaluation is therefore partly measuring the method's intended design rather than an independent property. The authors should validate the LLM judge against human ratings, report variance across prompts and judges, and include a non-circular quality metric such as perplexity under a held-out LM or task-specific automatic metrics, rather than dismissing the traditional metrics in Table 1 on the basis of a single unvalidated judge.
  4. [Section 4 / NSSC definition] The NSSC v1 and v2 green/red index sets and the switching rule are hand-crafted and never ablated; the paper explicitly states that optimization of these rules is left for future work. Because the entire watermark depends on these specific choices (e.g., green indices {1,3} vs. {2,4,5,6}, and switching after 5 matches), a reader cannot rule out that the strong detection and quality results are artifacts of a particular configuration. At minimum, an ablation varying the green-index set, the membership ranking cardinality, and the switching threshold should be provided for one model/dataset pair to show the method is not a knife-edge construction.
minor comments (4)
  1. [Throughout] There are several typos and inconsistencies, including 'Qusetion 1' and 'Qusetion 2' in Section 4.2, 'explaination' in Section 3.1, and inconsistent capitalization of 'Semstamp' vs. 'SemStamp'.
  2. [Section 4.2] The authors state that CoheMark outperforms baselines on 'accuracy' and 'reliability' in the case study, but the CoheMark example in Table 4 itself contains fabricated historical details and incorrect elevation data, as acknowledged in the text; the claimed superiority in factual accuracy is therefore not persuasive.
  3. [Figure 2] The robustness results after paraphrasing are presented only as a line plot without numerical values, error bars, or the number of prompts used; reporting the raw TPR values and sample sizes would allow readers to assess the significance of the robustness advantage.
  4. [Section 4 / Baselines] SemStamp results are omitted for Llama-3-8B because of infinite loops, and k-SemStamp is not compared at all; the paper should at least discuss how the missing baselines affect the comparison, especially since SemStamp is the closest prior sentence-level method.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: CoheMark's detection results are benchmarked against unwatermarked text and its quality evaluation measures an independently defined objective, so the paper's claims are not reduced to fitted inputs or self-citations.

full rationale

CoheMark is an insertion algorithm defined by fixed modules (Embedder, FuzzyClusterer, CoheSampler), and its detection test compares watermarked text against unwatermarked text, so the TPR@1% scores are not derived from the fitted cluster parameters by construction. The fuzzy c-means cluster count and the Next Sentence Selection Criteria are hand-chosen inputs that the paper explicitly leaves unoptimized, but these choices are parameters of the method rather than quantities the paper claims to predict. The GPT-4o judge is asked to rate coherence, relevance, interestingness, and integrity, which overlaps with CoheMark's stated objective of preserving cohesion; however, this is an evaluation of whether the design achieves its target, not a hidden reuse of the same data to manufacture the result. The potential confound that unwatermarked negatives are Mistral-7B output while watermarked positives are OPT and Llama output is a benchmark-validity concern, not a circular derivation from the paper's own equations. No load-bearing self-citation chain is present: MarkLLM and prior watermarking works are used only for experimental settings, baselines, and comparison, and the central watermarked-versus-unwatermarked discrimination is externally grounded. Therefore no significant circularity is identified.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claims rest on empirical design choices (8 clusters, NSSC index sets, switching threshold) and on the unstated transferability of the clustering domain. No theoretical derivation or external benchmark is provided for these choices.

free parameters (6)
  • Number of fuzzy clusters C = 8
    Set to 8 based on preliminary experiments (Section 4), affects both detection and quality trade-off.
  • Fuzziness parameter m = not reported
    Standard FCM parameter in Eq. (1), not specified in the paper, needed to reproduce clusters.
  • NSSC v1 green cluster indices = {1,3}
    Hand-chosen criteria in Section 4: accept next sentence if its primary cluster ranks 1st or 3rd in the previous sentence's membership index.
  • NSSC v2 green cluster indices = {2,4,5,6}
    Hand-chosen criteria in Section 4 for occasional topic pivots.
  • Switching rule threshold = 5
    Switch to NSSC v2 after 5 cumulative primary-cluster matches; arbitrary and unoptimized.
  • Generation temperature / repetition penalty = 0.9 / 1.05
    Appendix A; sampling hyperparameters used in experiments, part of the method's operational settings.
assumptions (5)
  • domain assumption Sentence-BERT embeddings capture semantic topical structure sufficient for clustering sentences into meaningful semantic spaces.
    Section 3.1 assumes the embedder's representations support fuzzy c-means clustering that reflects textual coherence.
  • domain assumption Fuzzy c-means clusters trained on a sample from a domain transfer to OpenGen and LFQA prompts.
    The training domain and N for clusters are not specified, yet the same clusters are used across datasets.
  • ad hoc to paper The hand-crafted NSSC rules preserve coherence and quality without degrading detectability.
    Section 4 states the rules and defers optimization to future work, making this a postulate of the method.
  • domain assumption GPT-4o pairwise preferences are a valid and comprehensive text-quality measure.
    Section 4.2 uses GPT-4o as judge and prioritizes its results over traditional metrics, which the paper argues are outdated.
  • domain assumption Unwatermarked reference texts generated by Mistral-7B-Instruct are an appropriate baseline for detection false-positive rates.
    Appendix B describes the reference text generation; fairness of this baseline is assumed without analysis.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CoheMark: A Novel Sentence-Level Watermark for Enhanced Text Quality." pith.science (2026). https://pith.science/paper/S5OMT74X

@misc{pith2026250417309,
  author       = {Pith},
  title        = {Pith review of: CoheMark: A Novel Sentence-Level Watermark for Enhanced Text Quality},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S5OMT74X}},
  note         = {Machine review of arXiv:2504.17309}
}
read the original abstract

Watermarking technology is a method used to trace the usage of content generated by large language models. Sentence-level watermarking aids in preserving the semantic integrity within individual sentences while maintaining greater robustness. However, many existing sentence-level watermarking techniques depend on arbitrary segmentation or generation processes to embed watermarks, which can limit the availability of appropriate sentences. This limitation, in turn, compromises the quality of the generated response. To address the challenge of balancing high text quality with robust watermark detection, we propose CoheMark, an advanced sentence-level watermarking technique that exploits the cohesive relationships between sentences for better logical fluency. The core methodology of CoheMark involves selecting sentences through trained fuzzy c-means clustering and applying specific next sentence selection criteria. Experimental evaluations demonstrate that CoheMark achieves strong watermark strength while exerting minimal impact on text quality.

Figures

Figures reproduced from arXiv: 2504.17309 by the authors.

Figure 1
Figure 1. The CoheMark watermarking procedure. Due to the inherent characteristic of rejection [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. As illustrated, CoheMark demonstrates a higher True Positive Rate following a paraphras [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SAMark: A Self-Anchored Text Watermarking with Paragraph-Level Paraphrase Robustness

    cs.CR 2026-05 unverdicted novelty 6.5 of 10

    SAMark uses self-anchored semantic green regions, multi-channel hyperbolic scoring, and diversity-aware filtering to reach 90.2% TP@FP1% detection under paragraph paraphrasing while preserving text quality.

Reference graph

Works this paper leans on

54 extracted references · 17 canonical work pages · cited by 1 Pith paper

  1. [1]

    Watermarking gpt outputs, 2023

    Scott Aaronson and Hendrik Kirchner. Watermarking gpt outputs, 2023

    work page 2023

  2. [2]

    Llama 3 model card

    AI@Meta. Llama 3 model card. 2024. URL https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md

    2024

  3. [3]

    Natural language watermarking: Design, analysis, and a proof-of-concept implementation

    Mikhail J Atallah, Victor Raskin, Michael Crogan, Christian Hempelmann, Florian Kerschbaum, Dina Mohamed, and Sanket Naik. Natural language watermarking: Design, analysis, and a proof-of-concept implementation. In Information Hiding: 4th International Workshop, IH 2001 Pittsburgh, PA, USA, April 25--27, 2001 Proceedings 4, pp.\ 185--200. Springer, 2001

    2001

  4. [4]

    Fcm: The fuzzy c-means clustering algorithm

    James C Bezdek, Robert Ehrlich, and William Full. Fcm: The fuzzy c-means clustering algorithm. Computers & geosciences, 10 0 (2-3): 0 191--203, 1984

    work page 1984

  5. [5]

    Postmark: A robust blackbox watermark for large language models

    Yapei Chang, Kalpesh Krishna, Amir Houmansadr, John Wieting, and Mohit Iyyer. Postmark: A robust blackbox watermark for large language models. arXiv preprint arXiv:2406.14517, 2024

    arXiv 2024

  6. [6]

    Beyond factuality: A comprehensive evaluation of large language models as knowledge generators

    Liang Chen, Yang Deng, Yatao Bian, Zeyu Qin, Bingzhe Wu, Tat-Seng Chua, and Kam-Fai Wong. Beyond factuality: A comprehensive evaluation of large language models as knowledge generators. arXiv preprint arXiv:2310.07289, 2023 a

    arXiv 2023

  7. [7]

    Watme: Towards lossless watermarking through lexical redundancy

    Liang Chen, Yatao Bian, Yang Deng, Deng Cai, Shuaiyi Li, Peilin Zhao, and Kam-Fai Wong. Watme: Towards lossless watermarking through lexical redundancy. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp.\ 9166--9180, 2024

    work page 2024

  8. [8]

    Exploring the use of large language models for reference-free text quality evaluation: An empirical study

    Yi Chen, Rui Wang, Haiyun Jiang, Shuming Shi, and Ruifeng Xu. Exploring the use of large language models for reference-free text quality evaluation: An empirical study. arXiv preprint arXiv:2304.00723, 2023 b

    arXiv 2023

Show all 54 references
  1. [9]

    Undetectable watermarks for language models

    Miranda Christ, Sam Gunn, and Or Zamir. Undetectable watermarks for language models. In The Thirty Seventh Annual Conference on Learning Theory, pp.\ 1125--1139. PMLR, 2024

    2024

  2. [10]

    Fuzzy c-means clustering with spatial information for image segmentation

    Keh-Shih Chuang, Hong-Long Tzeng, Sharon Chen, Jay Wu, and Tzong-Jer Chen. Fuzzy c-means clustering with spatial information for image segmentation. computerized medical imaging and graphics, 30 0 (1): 0 9--15, 2006

    2006

  3. [11]

    A fuzzy relative of the isodata process and its use in detecting compact well-separated clusters

    Joseph C Dunn. A fuzzy relative of the isodata process and its use in detecting compact well-separated clusters. 1973

    1973

  4. [12]

    The role of coherence and cohesion in text comprehension: an event-related fmri study

    Evelyn C Ferstl and D Yves Von Cramon. The role of coherence and cohesion in text comprehension: an event-related fmri study. Cognitive brain research, 11 0 (3): 0 325--340, 2001

    2001

  5. [13]

    Watermax: breaking the llm watermark detectability-robustness-quality trade-off

    Eva Giboulot and Teddy Furon. Watermax: breaking the llm watermark detectability-robustness-quality trade-off. arXiv preprint arXiv:2403.04808, 2024

    2024 arXiv

  6. [14]

    Cohesion in english

    Michael Alexander Kirkwood Halliday and Ruqaiya Hasan. Cohesion in english. Routledge, 2014

    2014

  7. [15]

    Less is more: Sparse watermarking in llms with enhanced text quality

    Duy C Hoang, Hung TQ Le, Rui Chu, Ping Li, Weijie Zhao, Yingjie Lao, and Khoa D Doan. Less is more: Sparse watermarking in llms with enhanced text quality. arXiv preprint arXiv:2407.13803, 2024

    2024 arXiv

  8. [16]

    Semstamp: A semantic watermark with paraphrastic robustness for text generation

    Abe Bohan Hou, Jingyu Zhang, Tianxing He, Yichen Wang, Yung-Sung Chuang, Hongwei Wang, Lingfeng Shen, Benjamin Van Durme, Daniel Khashabi, and Yulia Tsvetkov. Semstamp: A semantic watermark with paraphrastic robustness for text generation. arXiv preprint arXiv:2310.03991, 2023

    arXiv

  9. [17]

    k-semstamp: A clustering-based semantic watermark for detection of machine-generated text

    Abe Bohan Hou, Jingyu Zhang, Yichen Wang, Daniel Khashabi, and Tianxing He. k-semstamp: A clustering-based semantic watermark for detection of machine-generated text. arXiv preprint arXiv:2402.11399, 2024

    2024 arXiv

  10. [18]

    Unbiased watermark for large language models

    Zhengmian Hu, Lichang Chen, Xidong Wu, Yihan Wu, Hongyang Zhang, and Heng Huang. Unbiased watermark for large language models. arXiv preprint arXiv:2310.10669, 2023

    2023 arXiv

  11. [19]

    Mistral 7b

    Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, et al. Mistral 7b. arXiv preprint arXiv:2310.06825, 2023

    arXiv

  12. [20]

    A watermark for large language models

    John Kirchenbauer, Jonas Geiping, Yuxin Wen, Jonathan Katz, Ian Miers, and Tom Goldstein. A watermark for large language models. In International Conference on Machine Learning, pp.\ 17061--17084. PMLR, 2023 a

    2023

  13. [21]

    On the reliability of watermarks for large language models

    John Kirchenbauer, Jonas Geiping, Yuxin Wen, Manli Shu, Khalid Saifullah, Kezhi Kong, Kasun Fernando, Aniruddha Saha, Micah Goldblum, and Tom Goldstein. On the reliability of watermarks for large language models. arXiv preprint arXiv:2306.04634, 2023 b

    2023 arXiv

  14. [22]

    Paraphrasing evades detectors of ai-generated text, but retrieval is an effective defense

    Kalpesh Krishna, Yixiao Song, Marzena Karpinska, John Wieting, and Mohit Iyyer. Paraphrasing evades detectors of ai-generated text, but retrieval is an effective defense. Advances in Neural Information Processing Systems, 36, 2024

    2024

  15. [23]

    Booksum: A collection of datasets for long-form narrative summarization

    Wojciech Kry \'s ci \'n ski, Nazneen Rajani, Divyansh Agarwal, Caiming Xiong, and Dragomir Radev. Booksum: A collection of datasets for long-form narrative summarization. arXiv preprint arXiv:2105.08209, 2021

    2021 arXiv

  16. [24]

    Robust distortion-free watermarks for language models

    Rohith Kuditipudi, John Thickstun, Tatsunori Hashimoto, and Percy Liang. Robust distortion-free watermarks for language models. arXiv preprint arXiv:2307.15593, 2023

    2023 arXiv

  17. [25]

    An entropy-based text watermarking detection method

    Yijian Lu, Aiwei Liu, Dianzhi Yu, Jingjing Li, and Irwin King. An entropy-based text watermarking detection method. arXiv preprint arXiv:2403.13485, 2024

    2024 arXiv

  18. [26]

    Longwanjuan: Towards systematic measurement for long text quality

    Kai Lv, Xiaoran Liu, Qipeng Guo, Hang Yan, Conghui He, Xipeng Qiu, and Dahua Lin. Longwanjuan: Towards systematic measurement for long text quality. arXiv preprint arXiv:2402.13583, 2024

    2024 arXiv

  19. [27]

    Harmbench: A standardized evaluation framework for automated red teaming and robust refusal

    Mantas Mazeika, Long Phan, Xuwang Yin, Andy Zou, Zifan Wang, Norman Mu, Elham Sakhaee, Nathaniel Li, Steven Basart, Bo Li, et al. Harmbench: A standardized evaluation framework for automated red teaming and robust refusal. arXiv preprint arXiv:2402.04249, 2024

    2024 arXiv

  20. [28]

    The threat of offensive ai to organizations

    Yisroel Mirsky, Ambra Demontis, Jaidip Kotak, Ram Shankar, Deng Gelei, Liu Yang, Xiangyu Zhang, Maura Pintor, Wenke Lee, Yuval Elovici, et al. The threat of offensive ai to organizations. Computers & Security, 124: 0 103006, 2023

    2023

  21. [29]

    Digital medical image segmentation using fuzzy c-means clustering

    Bakhtyar Ahmed Mohammed and Muzhir Shaban Al-Ani. Digital medical image segmentation using fuzzy c-means clustering. UHD Journal of Science and Technology, 4 0 (1): 0 51--58, 2020

    2020

  22. [30]

    Waterjudge: Quality-detection trade-off when watermarking large language models

    Piotr Molenda, Adian Liusie, and Mark JF Gales. Waterjudge: Quality-detection trade-off when watermarking large language models. arXiv preprint arXiv:2403.19548, 2024

    2024 arXiv

  23. [31]

    Deeptextmark: A deep learning-driven text watermarking approach for identifying large language model generated text

    Travis Munyer, Abdullah Tanvir, Arjon Das, and Xin Zhong. Deeptextmark: A deep learning-driven text watermarking approach for identifying large language model generated text. arXiv preprint arXiv:2305.05773, 2023

    2023 arXiv

  24. [32]

    Gpt-4 technical report

    R OpenAI. Gpt-4 technical report. arxiv 2303.08774. View in Article, 2 0 (5), 2023

    2023 arXiv

  25. [33]

    Markllm: An open-source toolkit for llm watermarking

    Leyi Pan, Aiwei Liu, Zhiwei He, Zitian Gao, Xuandong Zhao, Yijian Lu, Binglin Zhou, Shuliang Liu, Xuming Hu, Lijie Wen, et al. Markllm: An open-source toolkit for llm watermarking. arXiv preprint arXiv:2405.10051, 2024

    2024 arXiv

  26. [34]

    Advanced image segmentation for precision agriculture using cnn-gat fusion and fuzzy c-means clustering

    Min Peng, Yunxiang Liu, Intakhab Alam Qadri, Uzair Aslam Bhatti, Bilal Ahmed, Nadia M Sarhan, and EM Awwad. Advanced image segmentation for precision agriculture using cnn-gat fusion and fuzzy c-means clustering. Computers and Electronics in Agriculture, 226: 0 109431, 2024

    2024

  27. [35]

    Natural language watermarking via paraphraser-based lexical substitution

    Jipeng Qiang, Shiyu Zhu, Yun Li, Yi Zhu, Yunhao Yuan, and Xindong Wu. Natural language watermarking via paraphraser-based lexical substitution. Artificial Intelligence, 317: 0 103859, 2023

    2023

  28. [36]

    Exploring the limits of transfer learning with a unified text-to-text transformer

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of machine learning research, 21 0 (140): 0 1--67, 2020

    2020

  29. [37]

    Sentence-bert: Sentence embeddings using siamese bert-networks

    N Reimers. Sentence-bert: Sentence embeddings using siamese bert-networks. arXiv preprint arXiv:1908.10084, 2019

    1908 arXiv

  30. [38]

    Clustering sentence-level text using a novel fuzzy relational clustering algorithm

    Andrew Skabar and Khaled Abdalgader. Clustering sentence-level text using a novel fuzzy relational clustering algorithm. IEEE transactions on knowledge and data engineering, 25 0 (1): 0 62--75, 2011

    2011

  31. [39]

    Words are not enough: sentence level natural language watermarking

    Mercan Topkara, Umut Topkara, and Mikhail J Atallah. Words are not enough: sentence level natural language watermarking. In Proceedings of the 4th ACM international workshop on Contents protection and security, pp.\ 37--46, 2006 a

    2006

  32. [40]

    The hiding virtues of ambiguity: quantifiably resilient watermarking of natural language text through synonym substitutions

    Umut Topkara, Mercan Topkara, and Mikhail J Atallah. The hiding virtues of ambiguity: quantifiably resilient watermarking of natural language text through synonym substitutions. In Proceedings of the 8th workshop on Multimedia and security, pp.\ 164--174, 2006 b

    2006

  33. [41]

    Llama 2: Open foundation and fine-tuned chat models

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023

    2023 arXiv

  34. [42]

    The importance of accurate cohesion and coherence in the text

    Olda Xhepa. The importance of accurate cohesion and coherence in the text. 2016

    2016

  35. [43]

    Hallucination is inevitable: An innate limitation of large language models

    Ziwei Xu, Sanjay Jain, and Mohan Kankanhalli. Hallucination is inevitable: An innate limitation of large language models. arXiv preprint arXiv:2401.11817, 2024

    2024 arXiv

  36. [44]

    Qwen2 technical report

    An Yang, Baosong Yang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Zhou, Chengpeng Li, Chengyuan Li, Dayiheng Liu, Fei Huang, et al. Qwen2 technical report. arXiv preprint arXiv:2407.10671, 2024

    2024 arXiv

  37. [45]

    Tracing text provenance via context-aware lexical substitution

    Xi Yang, Jie Zhang, Kejiang Chen, Weiming Zhang, Zehua Ma, Feng Wang, and Nenghai Yu. Tracing text provenance via context-aware lexical substitution. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.\ 11613--11621, 2022

    2022

  38. [46]

    Watermarking text generated by black-box language models

    Xi Yang, Kejiang Chen, Weiming Zhang, Chang Liu, Yuang Qi, Jie Zhang, Han Fang, and Nenghai Yu. Watermarking text generated by black-box language models. arXiv preprint arXiv:2305.08883, 2023

    2023 arXiv

  39. [47]

    Opt: Open pre-trained transformer language models

    Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen, Christopher Dewan, Mona Diab, Xian Li, Xi Victoria Lin, et al. Opt: Open pre-trained transformer language models. arXiv preprint arXiv:2205.01068, 2022

    2022 arXiv

  40. [48]

    Bertscore: Evaluating text generation with bert

    Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q Weinberger, and Yoav Artzi. Bertscore: Evaluating text generation with bert. arXiv preprint arXiv:1904.09675, 2019

    1904 arXiv

  41. [49]

    Provable robust watermarking for ai-generated text

    Xuandong Zhao, Prabhanjan Ananth, Lei Li, and Yu-Xiang Wang. Provable robust watermarking for ai-generated text. arXiv preprint arXiv:2306.17439, 2023

    2023 arXiv

  42. [50]

    Judging llm-as-a-judge with mt-bench and chatbot arena

    Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, et al. Judging llm-as-a-judge with mt-bench and chatbot arena. Advances in Neural Information Processing Systems, 36: 0 46595--46623, 2023

    2023

  43. [51]

    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...

  44. [52]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...

  45. [53]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...

  46. [54]

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.