Pith. sign in

REVIEW 4 major objections 5 minor 43 references

Legal Document Summarization: Enhancing Judicial Efficiency through Automation Detection

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A Legal Document Summarization framework built on GPT-4 and BERT claims to automate key-information extraction from legal texts, reducing review time from 15 to 5 hours and tripling daily case capacity.

desk verdict A legal-summarization paper whose evaluation uses no legal data and whose workflow numbers are asserted, so the central claim is untested. read the letter →

arxiv 2507.18952 v1 pith:ODITYQ3J submitted 2025-07-25 cs.CL

Yongjie Li , Ruilin Nong , Jianan Liu , Lucas Evans This is my paper · ORCID
classification cs.CL
keywords legaldocumentsummarizationjudicialefficiencyautomationdetectionkeyinformationextractionlargelanguagemodelsGPT-4BERTROUGE
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

Legal Document Summarization is a proposed framework for automatically finding and extracting the key information in legal texts so that attorneys can review a summary instead of the full document. The paper claims that combining a relevance-threshold extraction step with GPT-4 or BERT summarization, plus a hybrid key-information detector, produces concise summaries while cutting the average review time from 15 hours to 5 hours. It reports that detection of crucial details rises from 65% to 90%, summary error rates fall from 20% to 5%, and daily case capacity grows from 5 to 15 cases. A sympathetic reader would take the contribution as evidence that today's large language models can take over the triage portion of legal document review, freeing practitioners for analysis and decisions. The experimental evidence is drawn from four public summarization corpora paired with a 500-document annotated set.

What carries the argument

The central mechanism is a formal extraction-summarization loop. Given document $D$, the framework computes $K = \mathrm{Extract}(D) = \{k_i \mid k_i \in D,\ \mathrm{Relevance}(k_i) \geq \theta\}$, a set of sentences whose relevance score passes a threshold $\theta$, and then generates $S = \mathrm{Summarize}(K)$, a concise summary parameterized by the trained model. Sentence selection is the same operation in scoring form: $S_{\mathrm{summary}} = \arg\max_S \sum_{s \in S} \mathrm{score}(s,M)$, and efficiency is defined as $E = (T_{\mathrm{manual}} - T_{\mathrm{automated}})/T_{\mathrm{manual}} \times 100\%$. This machinery carries the whole argument because every claimed gain, from review time to daily case capacity, is the reported output of this two-stage process running with GPT-4 or BERT and a hybrid key-information detector. The equations specify the pieces, but the actual performance comes from the model choices and the threshold design rather than from an analytic guarantee.

What would settle it

Run the same GPT-4 and BERT pipelines on a held-out set of actual court opinions or legal briefs with human reference summaries and measure lawyer review time against manual reading; if the review-time reduction does not approach 66.67% or the key-information detection rate does not climb from 65% to 90%, the central efficiency claim is not supported.

Watch

Extended reading notes

Core claim

The paper's central claim is that automated legal summarization is not just a convenience but a workflow transformation: a two-stage pipeline that extracts relevance-scored segments and then compresses them can preserve the content a lawyer needs while dramatically reducing the time spent reading. The reported numbers are concrete: 66.67% less review time (15 to 5 hours), a 38.46% increase in key-information detection (65% to 90%), a 75% drop in summary errors (20% to 5%), and a 200% rise in cases handled per day (5 to 15). The paper attributes these gains to the framework's use of large language models and machine-learned relevance scoring, with GPT-4 outperforming BERT across every dataset and a hybrid detector giving the best precision and recall. If the framework works as described, legal professionals could delegate document triage to software and reserve human judgment for the analytical decisions that follow reading a summary.

Load-bearing premise

The load-bearing assumption is that the four evaluation corpora, WikiLingua, WikiWeb2M, GameWikiSum, and FinDSum, are close enough to genuine legal documents that their summary quality and speed numbers carry over to real judicial review.

Editorial extensions

If this is right

  • Legal staff could cut per-document review from 15 hours to 5 hours, reallocating the saved time to analysis and decision-making.
  • A key-information detection rise from 65% to 90% would mean fewer overlooked clauses, citations, or case elements during triage.
  • A summary error-rate drop from 20% to 5% would make automated summaries dependable enough to use as the first-pass review artifact.
  • Tripling daily case capacity from 5 to 15 would allow a legal team to handle a larger docket without adding review hours.

Reading between the lines

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

  • A direct testable extension would rerun the method on real court opinions with lawyer-written reference summaries; the four public datasets are generic summarization benchmarks, so legal-domain transfer is the unconfirmed step.
  • The 66.67% time saving is computed from assumed baselines of 15 manual hours and 5 automated hours rather than measured end-to-end workflow data, so the real ratio depends on the verification work that still surrounds automated summaries.
  • A natural deployment the paper does not spell out is a confidence threshold: route only high-confidence summaries around human review, leaving low-confidence documents to lawyers, which would preserve accuracy while realizing most of the speed gain.
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 / 5 minor

Summary. The paper proposes a Legal Document Summarization framework that combines automated key-segment extraction and summarization (Eqs. 1 and 2) with machine-learning pattern recognition, and it claims that this automation substantially improves judicial efficiency. Experiments are reported for GPT-4 and BERT on WikiLingua, WikiWeb2M, GameWikiSum, and FinDSum, with ROUGE, accuracy, relevance, completeness, and F1 scores in Table I, and Section V-F reports workflow gains including a 66.67% reduction in review time and a 200% increase in cases processed per day. The abstract, introduction, and conclusion repeatedly state that the method was validated on real-world legal datasets, but the paper describes no legal-document corpus and no experimental protocol for the workflow-efficiency claims.

Significance. If the central claims were substantiated, the paper would address an important applied problem: automated legal summarization could reduce the time lawyers and judges spend on document triage. The problem framing is relevant and the paper identifies a plausible use of LLMs for extraction and summarization. However, the manuscript provides no machine-checked proofs, reproducible code, or complete experimental protocol, and the evidence that connects the experiments to legal documents is absent. The claimed efficiency gains are not tied to any measurement, and several tables report methods never introduced in the methodology. As it stands, the contribution is a set of formal definitions and performance numbers on non-legal corpora, not a demonstrated legal summarization system.

major comments (4)
  1. [Section IV-A and Table I] The four datasets used in Table I—WikiLingua, WikiWeb2M, GameWikiSum, and FinDSum—are not legal document corpora, yet the abstract, Section I, and Section VI repeatedly state that experiments were conducted on real-world legal datasets. Since every ROUGE, accuracy, and F1 score in Table I is computed on non-legal text, the central claim that the framework summarizes legal documents with high quality is untested.
  2. [Section V-F and Eq. (6)] The workflow gains reported in Section V-F (review time reduced from 15 to 5 hours, detection rate from 65% to 90%, error rate from 20% to 5%, and cases processed per day from 5 to 15) are stated without any measurement protocol, subject population, document sample, or link to the experiments in Section V-A. Equation (6) defines an efficiency percentage but does not supply the values of Tmanual and Tautomated for these claims, so the 66.67% and 200% figures have no evidentiary basis.
  3. [Section II-B and Tables II-III] The baselines in Section II-B are not summarization systems (voice recognition, GitHub workflow automation, hyper-automation, testbench generation, and defect detection), so there is no comparison against existing summarization methods. In addition, Tables II and III report results for methods (Rule-Based, Supervised Learning, Deep Learning, Hybrid Approach; SVM, Random Forest, Naive Bayes, Decision Tree, Logistic Regression) that are not described in the methodology, making the numbers in these tables uninterpretable and disconnected from the proposed framework.
  4. [Section III-A through III-C, Eqs. (1)-(5)] Equations (1), (2), (3), and (5) are definitions of the desired extraction and summarization behavior, not derivations of a concrete trained system: the relevance threshold theta in Eq. (1) is never specified, the model parameters P and theta are not tied to a training procedure on legal data, and the scoring function score(s, M) is not defined. The claim that models are trained on annotated legal data (Section III-A) is unsupported because Section IV-A lists only non-legal datasets; the formal apparatus therefore does not establish a working method.
minor comments (5)
  1. [Section IV-D] Section IV-D states that the experiments use 500 annotated legal documents averaging 2,000 tokens, but no such legal dataset is described in Section IV-A or anywhere else in the paper; this contradicts the dataset list and needs clarification.
  2. [Tables II and III] Table II presents precision, recall, and F1 as percentages while Table III uses decimal values; the units and their relationship should be unified and stated explicitly.
  3. [Eq. (4)] The summation notation in Eq. (4) appears as 'mX' and should be rendered as a standard sum over i = 1 to m; the current typesetting makes the loss definition unclear.
  4. [Figures 2 and 3] Figures 2 and 3 lack axis labels, error bars, and any indication of statistical significance, so the claimed upward trend in pattern recognition performance cannot be verified from the figures alone.
  5. [References] Reference formatting is inconsistent (for example, 'V oice' and 'teusz Litwin' are misspelled), and references [38] and [39] are cited in the dataset section but are not datasets used in the summarization evaluation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the paper's equations are generic definitions and its self-citations are not load-bearing; the central problems are unsubstantiated claims and dataset mismatch, not circularity.

full rationale

Walking the claimed derivation chain, no step makes a prediction or first-principles result that is equivalent to its inputs by construction. Eq. (1) defines extraction as a relevance-thresholded subset of a document, Eq. (3) is a generic classifier form, and Eq. (5) defines summarization as maximizing a sentence-scoring function; these are definitions rather than derivations, and none of the reported scores is shown to be forced by them. Eq. (6) is the standard percentage-reduction formula; the later statement that review time fell from 15 to 5 hours is arithmetic (66.67%) applied to asserted, unmeasured inputs, which is an evidentiary and reproducibility problem rather than a circular reduction. The self-citations ([5], [16], [32]) occur only in related-work annotations and do not carry the framework's central legal-efficiency claim, so they are not load-bearing. The most serious defect is external validity: Section IV-A lists only WikiLingua, WikiWeb2M, GameWikiSum, and FinDSum, while the abstract and Section VI claim experiments on 'real-world legal datasets'; that contradiction, plus the absence of any measurement protocol for the workflow figures, undermines the legal-domain conclusion, but it is not a circularity because the evaluation is not an input to any derivation. Therefore the appropriate circularity finding is no significant circularity.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The framework's claimed effectiveness rests on an unspecified relevance threshold, unstated trained parameters, the assumption that four non-legal summarization datasets proxy legal documents, and an undocumented manual assessment. None of these is given independent support, so the central judicial-efficiency claim is not backed by the described evidence.

free parameters (2)
  • Relevance threshold θ
    Equation (1) defines extracted key segments as those with Relevance(k_i) ≥ θ, but the paper never states how θ is chosen, learned, or validated.
  • Model parameters P/θ
    Equations (2) to (4) assume a trained summarizer and a learned scoring function; no architecture, data split, or training details beyond batch size and learning rate are provided.
assumptions (3)
  • ad hoc to paper Non-legal summarization corpora stand in for legal documents.
    Section IV-A lists WikiLingua, WikiWeb2M, GameWikiSum, and FinDSum as the evaluation datasets, none of which is a legal corpus; the legal-domain claim assumes transfer from these corpora.
  • domain assumption A learned relevance function can identify legally important content from annotated data.
    Equation (1) depends on a relevance measure trained on annotated legal data, but no annotation protocol or legal annotation set is described.
  • domain assumption Manual relevance and completeness scores are reliable.
    Tables I and II report manually assessed relevance and completeness, but no rubric, evaluator training, or inter-annotator agreement is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Legal Document Summarization: Enhancing Judicial Efficiency through Automation Detection." pith.science (2026). https://pith.science/paper/ODITYQ3J

@misc{pith2026250718952,
  author       = {Pith},
  title        = {Pith review of: Legal Document Summarization: Enhancing Judicial Efficiency through Automation Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ODITYQ3J}},
  note         = {Machine review of arXiv:2507.18952}
}
read the original abstract

Legal document summarization represents a significant advancement towards improving judicial efficiency through the automation of key information detection. Our approach leverages state-of-the-art natural language processing techniques to meticulously identify and extract essential data from extensive legal texts, which facilitates a more efficient review process. By employing advanced machine learning algorithms, the framework recognizes underlying patterns within judicial documents to create precise summaries that encapsulate the crucial elements. This automation alleviates the burden on legal professionals, concurrently reducing the likelihood of overlooking vital information that could lead to errors. Through comprehensive experiments conducted with actual legal datasets, we demonstrate the capability of our method to generate high-quality summaries while preserving the integrity of the original content and enhancing processing times considerably. The results reveal marked improvements in operational efficiency, allowing legal practitioners to direct their efforts toward critical analytical and decision-making activities instead of manual reviews. This research highlights promising technology-driven strategies that can significantly alter workflow dynamics within the legal sector, emphasizing the role of automation in refining judicial processes.

Figures

Figures reproduced from arXiv: 2507.18952 by the authors.

Figure 1
Figure 1. Two Different legal document auditing framework for LLMs [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Performance comparison of models on selected sum [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 2
Figure 2. Evaluation of pattern recognition techniques in legal [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Impact of automated legal document summarization on [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 22 canonical work pages

  1. [1]

    Language models are few-shot learners,

    T. B. Brown, B. Mann, N. Ryder, M. Subbiah, J. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, S. Agarwal, A. Herbert- V oss, G. Krueger, T. Henighan, R. Child, A. Ramesh, D. M. Ziegler, J. Wu, C. Winter, C. Hesse, M. Chen, E. Sigler, M. teusz Litwin, S. Gray, B. Chess, J. Clark, C. Berner, S. McCandlish, A. Radford, I. Sutskever, and...

    arXiv 2005

  2. [2]

    Palm: Scaling language modeling with pathways,

    A. Chowdhery, S. Narang, J. Devlin, M. Bosma, G. Mishra, A. Roberts, P. Barham, H. W. Chung, C. Sutton, S. Gehrmann, P. Schuh, K. Shi, S. Tsvyashchenko, J. Maynez, A. Rao, P. Barnes, Y . Tay, N. M. Shazeer, V . Prabhakaran, E. Reif, N. Du, B. Hutchinson, R. Pope, J. Bradbury, J. Austin, M. Isard, G. Gur-Ari, P. Yin, T. Duke, A. Levskaya, S. Ghemawat, S. D...

    arXiv 2022

  3. [3]

    Training language models to follow instructions with human feedback,

    L. Ouyang, J. Wu, X. Jiang, D. Almeida, C. L. Wainwright, P. Mishkin, C. Zhang, S. Agarwal, K. Slama, A. Ray, J. Schulman, J. Hilton, F. Kelton, L. E. Miller, M. Simens, A. Askell, P. Welinder, P. F. Christiano, J. Leike, and R. J. Lowe, “Training language models to follow instructions with human feedback,” ArXiv, vol. abs/2203.02155, 2022

    arXiv 2022

  4. [4]

    Artificial intelligence for context-aware visual change detection in software test automation,

    M. Moradi, K. Yan, D. Colwell, and R. Asgari, “Artificial intelligence for context-aware visual change detection in software test automation,” ArXiv, vol. abs/2405.00874, 2024

    work page arXiv 2024

  5. [5]

    Analysis of collective response reveals that covid-19-related activities start from the end of 2019 in mainland china,

    J. Liu, T. Huang, H. Xiong, J. Huang, J. Zhou, H. Jiang, G. Yang, H. Wang, and D. Dou, “Analysis of collective response reveals that covid-19-related activities start from the end of 2019 in mainland china,” medRxiv, pp. 2020–10, 2020

    2019

  6. [6]

    Voice Recognition Robot with Real-Time Surveillance and Automation

    L. Basyal, “V oice recognition robot with real-time surveillance and automation,” ArXiv, vol. abs/2312.04072, 2023

    work page Pith review arXiv 2023

  7. [7]

    HADES: Homologous Automated Document Exploration and Summarization

    P. Wilczy’nski, A. .Z’olkowski, M. Krzyzi’nski, E. Wi´snios, B. Pieli’nski, S. Gizi’nski, J. Sienkiewicz, and P. Biecek, “Hades: Homologous automated document exploration and summarization,” ArXiv, vol. abs/2302.13099, 2023

    work page Pith review arXiv 2023

  8. [8]

    Semi-automatic data enhancement for document-level relation extraction with distant supervision from large language models,

    J. Li, Z. Jia, and Z. Zheng, “Semi-automatic data enhancement for document-level relation extraction with distant supervision from large language models,” pp. 5495–5505, 2023

    work page 2023

Show all 43 references
  1. [9]

    Enhancing legal document retrieval: A multi-phase approach with large language models,

    H.-L. Nguyen, D.-M. Nguyen, T.-M. Nguyen, H.-T. Nguyen, T.-H.-Y . Vuong, and K. Satoh, “Enhancing legal document retrieval: A multi-phase approach with large language models,” ArXiv, vol. abs/2403.18093, 2024

    2024 arXiv

  2. [10]

    Improving vietnamese legal document retrieval using synthetic data,

    S. P. Tien, H. N. Doan, A. N. Dai, and S. D. Viet, “Improving vietnamese legal document retrieval using synthetic data,”ArXiv, vol. abs/2412.00657, 2024

    2024 arXiv

  3. [11]

    Large language model prompt chaining for long legal document classification,

    D. Trautmann, “Large language model prompt chaining for long legal document classification,” ArXiv, vol. abs/2308.04138, 2023

    2023 arXiv

  4. [12]

    Automated argument generation from legal facts,

    O. Tuvey and P. Sen, “Automated argument generation from legal facts,” ArXiv, vol. abs/2310.05680, 2023

    2023 arXiv

  5. [13]

    Docinfer: Document-level natural language inference using optimal evidence selection,

    P. Mathur, G. Kunapuli, R. A. Bhat, M. Shrivastava, D. Manocha, and M. Singh, “Docinfer: Document-level natural language inference using optimal evidence selection,” pp. 809–824, 2022

    2022

  6. [14]

    Docetl: Agentic query rewriting and evaluation for complex document processing,

    S. Shankar, A. G. Parameswaran, and E. Wu, “Docetl: Agentic query rewriting and evaluation for complex document processing,” ArXiv, vol. abs/2410.12189, 2024

    2024 arXiv

  7. [15]

    Anls* - a universal document processing metric for generative large language models,

    D. Peer, P. Sch ¨opf, V . Nebendahl, A. Rietzler, and S. Stabinger, “Anls* - a universal document processing metric for generative large language models,” ArXiv, vol. abs/2402.03848, 2024

    2024 arXiv

  8. [16]

    Chulo: Chunk-level key information representation for long document processing,

    Y . Li, S. Han, Y . Dai, and F. Cao, “Chulo: Chunk-level key information representation for long document processing,” ArXiv, vol. abs/2410.11119, 2024

    2024 arXiv

  9. [17]

    Unifying vision, text, and layout for universal document processing,

    Z. Tang, Z. Yang, G. Wang, Y . Fang, Y . Liu, C. Zhu, M. Zeng, C.-Y . Zhang, and M. Bansal, “Unifying vision, text, and layout for universal document processing,” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 19 254–19 264, 2022

    2023

  10. [18]

    Domain-agnostic neural architecture for class incre- mental continual learning in document processing platform,

    M. W ´ojcik, W. Kosciukiewicz, M. Baran, T. Kajdanowicz, and A. Gonczarek, “Domain-agnostic neural architecture for class incre- mental continual learning in document processing platform,” ArXiv, vol. abs/2307.05399, 2023

    2023 arXiv

  11. [19]

    The future of document indexing: Gpt and donut revolutionize table of content processing,

    D. W. Feyisa, H. Berihun, A. Zewdu, M. Najimoghadam, and M. Zare, “The future of document indexing: Gpt and donut revolutionize table of content processing,” ArXiv, vol. abs/2403.07553, 2024

    2024 arXiv

  12. [20]

    Raptor: Recursive abstractive processing for tree-organized retrieval,

    P. Sarthi, S. Abdullah, A. Tuli, S. Khanna, A. Goldie, and C. D. Manning, “Raptor: Recursive abstractive processing for tree-organized retrieval,” ArXiv, vol. abs/2401.18059, 2024

    2024 arXiv

  13. [21]

    Adapting large language models for document-level machine translation,

    M. Wu, T.-T. Vu, L. Qu, G. Foster, and G. Haffari, “Adapting large language models for document-level machine translation,” ArXiv, vol. abs/2401.06468, 2024

    2024 arXiv

  14. [22]

    Semantic similarity matching for patent documents using ensemble bert-related model and novel text processing method,

    L. Yu, B. Liu, Q. Lin, X. Zhao, and C. Che, “Semantic similarity matching for patent documents using ensemble bert-related model and novel text processing method,” ArXiv, vol. abs/2401.06782, 2024

    2024 arXiv

  15. [23]

    Intelligent system for automated molecular patent infringement assessment,

    Y . Shi, S. Li, T. Zhang, X. Fang, J. Wang, Z. Liu, G. Zhao, Z. Zhu, Z. Gao, R. Zhong, L. Zhang, G. Ke, E. Weinan, H. Cai, and X. Wang, “Intelligent system for automated molecular patent infringement assessment,” ArXiv, vol. abs/2412.07819, 2024

    2024 arXiv

  16. [24]

    blocklaw: Blockchain technology for legal automation and workflow - cyber ethics and cybersecurity platforms,

    B. Pokharel and N. Kshetri, “blocklaw: Blockchain technology for legal automation and workflow - cyber ethics and cybersecurity platforms,” ArXiv, vol. abs/2410.06143, 2024

    2024 arXiv

  17. [25]

    Identification of regulatory requirements relevant to business processes: A comparative study on generative ai, embedding-based ranking, crowd and expert- driven methods,

    C. Sai, S. Sadiq, L. Han, G. Demartini, and S. Rinderle-Ma, “Identification of regulatory requirements relevant to business processes: A comparative study on generative ai, embedding-based ranking, crowd and expert- driven methods,” ArXiv, vol. abs/2401.02986, 2024

    2024 arXiv

  18. [26]

    Enhancing legal case retrieval via scaling high-quality synthetic query-candidate pairs,

    C. Gao, C. Xiao, Z. Liu, H. Chen, Z. Liu, and M. Sun, “Enhancing legal case retrieval via scaling high-quality synthetic query-candidate pairs,” ArXiv, vol. abs/2410.06581, 2024

    2024 arXiv

  19. [27]

    Enabling discriminative rea- soning in llms for legal judgment prediction,

    C. Deng, K. Mao, Y . Zhang, and Z. Dou, “Enabling discriminative rea- soning in llms for legal judgment prediction,” ArXiv, vol. abs/2407.01964, 2024

    2024 arXiv

  20. [28]

    Better call gpt, comparing large language models against lawyers,

    L. Martin, N. Whitehouse, S. Yiu, L. Catterson, and R. Perera, “Better call gpt, comparing large language models against lawyers,” ArXiv, vol. abs/2401.16212, 2024

    2024 arXiv

  21. [29]

    Agents on the bench: Large language model based multi agent framework for trustworthy digital justice,

    C. Jiang and X. Yang, “Agents on the bench: Large language model based multi agent framework for trustworthy digital justice,” ArXiv, vol. abs/2412.18697, 2024

    2024 arXiv

  22. [30]

    An element is worth a thousand words: Enhancing legal case retrieval by incorporating legal elements,

    C. Deng, Z. Dou, Y . Zhou, P. Zhang, and K. Mao, “An element is worth a thousand words: Enhancing legal case retrieval by incorporating legal elements,” pp. 2354–2365, 2024

    2024

  23. [31]

    Nowj1@alqac 2023: Enhancing legal task performance with classic statistical models and pre-trained language models,

    T.-M. Nguyen, X.-H. Nguyen, N.-D. Mai, M.-Q. Hoang, V .-H. Nguyen, H.-V . Nguyen, N. H. Thanh, and T.-H.-Y . Vuong, “Nowj1@alqac 2023: Enhancing legal task performance with classic statistical models and pre-trained language models,” ArXiv, vol. abs/2309.09070, 2023

    2023 arXiv

  24. [32]

    Unlocking context constraints of llms: Enhancing context efficiency of llms with self-information-based content filtering,

    Y . Li, “Unlocking context constraints of llms: Enhancing context efficiency of llms with self-information-based content filtering,” ArXiv, vol. abs/2304.12102, 2023

    2023 arXiv

  25. [33]

    Eragent: Enhancing retrieval-augmented language models with improved accuracy, efficiency, and personalization,

    Y . Shi, X. Zi, Z. Shi, H. Zhang, Q. Wu, and M. Xu, “Eragent: Enhancing retrieval-augmented language models with improved accuracy, efficiency, and personalization,” ArXiv, vol. abs/2405.06683, 2024

    2024 arXiv

  26. [34]

    Wikilingua: A new benchmark dataset for multilingual abstractive summarization,

    F. Ladhak, E. Durmus, C. Cardie, and K. McKeown, “Wikilingua: A new benchmark dataset for multilingual abstractive summarization,” ArXiv, vol. abs/2010.03093, 2020

    2010 arXiv

  27. [35]

    Wikiweb2m: A page-level multimodal wikipedia dataset,

    A. Burns, K. Srinivasan, J. Ainslie, G. Brown, B. A. Plummer, K. Saenko, J. Ni, and M. Guo, “Wikiweb2m: A page-level multimodal wikipedia dataset,” ArXiv, vol. abs/2305.05432, 2023

    2023 arXiv

  28. [36]

    Gamewikisum: a novel large multi- document summarization dataset,

    D. Antognini and B. Faltings, “Gamewikisum: a novel large multi- document summarization dataset,” pp. 6645–6650, 2020

    2020

  29. [37]

    Long text and multi-table summarization: Dataset and method,

    S. Liu, J. Cao, R. Yang, and Z. Wen, “Long text and multi-table summarization: Dataset and method,” ArXiv, vol. abs/2302.03815, 2023

    2023 arXiv

  30. [38]

    Solving multiple-instance and multiple- part learning problems with decision trees and rule sets. application to the mutagenesis problem,

    Y . Chevaleyre and J.-D. Zucker, “Solving multiple-instance and multiple- part learning problems with decision trees and rule sets. application to the mutagenesis problem,” pp. 204–214, 2001

    2001

  31. [39]

    Reading digits in natural images with unsupervised feature learning,

    Y . Netzer, T. Wang, A. Coates, A. Bissacco, B. Wu, and A. Ng, “Reading digits in natural images with unsupervised feature learning,” 2011

    2011

  32. [40]

    The github devel- opment workflow automation ecosystems,

    M. Wessel, T. Mens, A. Decan, and P. R. Mazrae, “The github devel- opment workflow automation ecosystems,” ArXiv, vol. abs/2305.04772, 2023

    2023 arXiv

  33. [41]

    Hyper-automation-the next peripheral for automation in it industries,

    A. Rajput and R. Gupta, “Hyper-automation-the next peripheral for automation in it industries,” ArXiv, vol. abs/2305.11896, 2023

    2023 arXiv

  34. [42]

    Llm- aided testbench generation and bug detection for finite-state machines,

    J. Bhandari, J. Knechtel, R. Narayanaswamy, S. Garg, and R. Karri, “Llm- aided testbench generation and bug detection for finite-state machines,” ArXiv, vol. abs/2406.17132, 2024

    2024 arXiv

  35. [43]

    A comprehensive survey on machine learning driven material defect detection: Challenges, solutions, and future prospects,

    J. Bai, D. Wu, T. Shelley, P. Schubel, D. Twine, J. Russell, X. Zeng, and J. Zhang, “A comprehensive survey on machine learning driven material defect detection: Challenges, solutions, and future prospects,” ArXiv, vol. abs/2406.07880, 2024

    2024 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.