Pith. sign in

REVIEW 65 references

AUTOLAW: Enhancing Legal Compliance in Large Language Models via Case Law Generation and Jury-Inspired Deliberation

T0 review · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read AutoLaw's verifier-ranked legal-role jury with a similar-case demonstration beats majority voting for violation detection on three law and policy benchmarks.

arxiv 2505.14015 v2 pith:Z7S4VZMZ submitted 2025-05-20 cs.CL

Tai D. Nguyen , Long H. Pham , Jun Sun This is my paper
classification cs.CL
keywords legalautolawcompliancedeliberationdetectiongenerationllmsaddress
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

AutoLaw attacks the problem of LLMs giving legally wrong answers in Singapore. First it turns regulations into lists of punishable acts, and then uses a strong generation model to write thousands of tricky stories in which one of those acts is hidden. Four test models find obvious violations easily but struggle with the tricky stories, which is why the authors say adversarial data is useful for telling models apart.

The second half is a voting scheme. The authors create a pool of 21 model-role combinations, such as Qwen as judge or Phi as prosecutor. A GPT-4.1 verifier scores each juror on past synthetic cases, and the best-scoring jurors are selected for each new case. Each juror votes yes or no, and the majority wins. A retrieved similar case is included in the prompt as an example.

The paper reports large gains over random majority voting, and near-perfect scores on 58 real newspaper cases. The main weakness is that the reported metric counts only detection of violations; almost all test examples are violations, so a model that always says yes would look perfect. The synthetic labels are also assigned by the same pipeline that creates the scenarios, so the headline improvement is hard to interpret as genuine legal compliance rather than alignment with the pipeline's own bias.

Extended reading notes

Core claim

Main claim: AutoLaw, by ranking jurors with a verifier on synthesized case law and having them vote with a retrieved demonstration, achieves the highest detection rates across all pools, with gaps over majority voting from 11.37 to 22.27 percentage points on Law-SG and Unfair-TOS, and 97.62 to 100 on Case-SG. The paper states the gaps are substantial, ranging from 11.37 (Vote-3, P2) to 22.27 (Vote-1, P1). If correct, verifier-ranked legal-role juries plus a similar-case demonstration improve violation detection substantially over unranked majority voting in this setting.

Load-bearing premise

Load-bearing premise: the Law-SG test inputs are disjoint from the case-law database C* used for retrieval. Algorithm 2 chooses x_hat = argmax sim(x, x_hat) over C* and puts x_hat and its Yes answer into every juror's prompt; if the test scenario is itself in C*, the system is handed the answer. The paper never states this disjointness, and Law-SG is generated by the same ALI process over the same five regulations as C* (Sections 3.1, 3.3, Appendix A.2). A related but separate premise is that every ALI-refined scenario remains a true violation, which is why a No vote in the Table 1 medical-condition example is treated as an error.

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.

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

The central claim rests on self-labeled synthetic data generated by the same ALI pipeline used for ranking and testing, on the unvalidated GPT-4.1 verifier, and on a detection-rate metric that lacks a negative class. No new physical or mathematical entity is postulated.

free parameters (2)
  • majority threshold theta = 0.5
    Chosen by convention in Algorithm 2 and Section 3.3; not fitted, but it determines the verdict and therefore the reported detection-rate gaps.
  • jury size k = 1, 3, and 5 (each reported)
    The paper reports Vote-1, Vote-3, and Vote-5 separately, so k is an experimenter-chosen design parameter rather than a hidden fitted value.
assumptions (4)
  • domain assumption Every scenario generated by ALI(m) from a listed misconduct is a true violation of that misconduct, and its Yes label is correct.
    Used throughout Section 2.1, Section 3.1, and Appendix Table 6 to build Law-SG and the case-law database; no human validation of labels is reported, and the medical-condition example in Table 1 shows a scenario where the label may be contestable.
  • domain assumption The verifier V(m,r), powered by GPT-4.1, assigns correct 0 to 1 scores to juror answers on synthetic case law.
    Section 3.2 relies on V to rank and select jurors; the paper does not measure verifier agreement with legal experts or with human labels.
  • domain assumption Detection rate on positive-only test sets is an adequate measure of legal compliance.
    Section 4 defines detection rate as the metric for all three datasets; Law-SG, Case-SG, and Unfair-TOS are evaluated without a negative class, so a yes-only classifier would achieve 100 percent by construction.
  • domain assumption Law-SG test scenarios are disjoint from the case-law database C* used for retrieval demonstrations.
    Algorithm 2 retrieves the most similar case-law item without excluding the test input; if a Law-SG input is in C*, the juror is shown the exact scenario with the correct answer, conflating retrieval with deliberation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AUTOLAW: Enhancing Legal Compliance in Large Language Models via Case Law Generation and Jury-Inspired Deliberation." pith.science (2026). https://pith.science/paper/Z7S4VZMZ

@misc{pith2026250514015,
  author       = {Pith},
  title        = {Pith review of: AUTOLAW: Enhancing Legal Compliance in Large Language Models via Case Law Generation and Jury-Inspired Deliberation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z7S4VZMZ}},
  note         = {Machine review of arXiv:2505.14015}
}
read the original abstract

The rapid advancement of domain-specific large language models (LLMs) in fields like law necessitates frameworks that account for nuanced regional legal distinctions, which are critical for ensuring compliance and trustworthiness. Existing legal evaluation benchmarks often lack adaptability and fail to address diverse local contexts, limiting their utility in dynamically evolving regulatory landscapes. To address these gaps, we propose AutoLaw, a novel violation detection framework that combines adversarial data generation with a jury-inspired deliberation process to enhance legal compliance of LLMs. Unlike static approaches, AutoLaw dynamically synthesizes case law to reflect local regulations and employs a pool of LLM-based "jurors" to simulate judicial decision-making. Jurors are ranked and selected based on synthesized legal expertise, enabling a deliberation process that minimizes bias and improves detection accuracy. Evaluations across three benchmarks: Law-SG, Case-SG (legality), and Unfair-TOS (policy), demonstrate AutoLaw's effectiveness: adversarial data generation improves LLM discrimination, while the jury-based voting strategy significantly boosts violation detection rates. Our results highlight the framework's ability to adaptively probe legal misalignments and deliver reliable, context-aware judgments, offering a scalable solution for evaluating and enhancing LLMs in legally sensitive applications.

Figures

Figures reproduced from arXiv: 2505.14015 by the authors.

Figure 1
Figure 1. AUTOLAW generates adversarial data and im￾proves violation detection. Figure (a) shows detection rates (DR, the lower the better at discriminating LLMs) on explicit (trivial) and implicit (nuanced) data. Figure (b) presents violation detection rates (the higher the better per￾formance) on three datasets. We evaluate AUTOLAW on three benchmarks spanning legality (Law-SG, Case-SG) and policy violation (Unfair-TOS [2])… view at source ↗
Figure 2
Figure 2. LLM performance on the violation detection task and editing performance on Llama3.1:8b. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Performance comparison of LLMs with DPO and RAG enhancements. The [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: An overview of the existing majority vote (a) and the proposed [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The impact of each component (e.g., jury selection) on Law-SG dataset. [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

65 extracted references · 23 canonical work pages

  1. [1]

    U. AI. Unsloth library, 2024. Accessed: November 8, 2024

    work page 2024

  2. [2]

    B. S. Akash, A. Kupireddy, and L. B. Murthy. Unfair tos: An automated approach using customized bert.arXiv preprint arXiv:2401.11207, 2024

    work page Pith review arXiv 2024

  3. [3]

    Rapid transit systems regulations, 2025

    Attorney-General’s Chambers. Rapid transit systems regulations, 2025. Singapore Statutes Online

    work page 2025

  4. [4]

    S. R. Bowman and G. E. Dahl. What will it take to fix benchmarking in natural language understanding?arXiv preprint arXiv:2104.02145, 2021

    arXiv 2021

  5. [5]

    Chalkidis, I

    I. Chalkidis, I. Androutsopoulos, and N. Aletras. Neural legal judgment prediction in english. arXiv preprint arXiv:1906.02059, 2019

    arXiv 1906

  6. [6]

    Chalkidis, M

    I. Chalkidis, M. Fergadiotis, D. Tsarapatsanis, N. Aletras, I. Androutsopoulos, and P. Malaka- siotis. Paragraph-level rationale extraction through regularization: A case study on european court of human rights cases.arXiv preprint arXiv:2103.13084, 2021

    arXiv 2021

  7. [7]

    Chalkidis, A

    I. Chalkidis, A. Jana, D. Hartung, M. Bommarito, I. Androutsopoulos, D. M. Katz, and N. Aletras. Lexglue: A benchmark dataset for legal language understanding in english.arXiv preprint arXiv:2110.00976, 2021

    arXiv 2021

  8. [8]

    Chang, X

    Y . Chang, X. Wang, J. Wang, Y . Wu, L. Yang, K. Zhu, H. Chen, X. Yi, C. Wang, Y . Wang, et al. A survey on evaluation of large language models.ACM transactions on intelligent systems and technology, 15(3):1–45, 2024

    2024

Show all 65 references
  1. [9]

    Chase and L

    H. Chase and L. contributors. Langchain, 2022

    2022

  2. [10]

    X. Chen, R. Aksitov, U. Alon, J. Ren, K. Xiao, P. Yin, S. Prakash, C. Sutton, X. Wang, and D. Zhou. Universal self-consistency for large language model generation.arXiv preprint arXiv:2311.17311, 2023

    2023 arXiv

  3. [11]

    Cheng, S

    D. Cheng, S. Huang, and F. Wei. Adapting large language models via reading comprehension. InThe Twelfth International Conference on Learning Representations, 2023

    2023

  4. [12]

    Y . Du, S. Li, A. Torralba, J. B. Tenenbaum, and I. Mordatch. Improving factuality and reasoning in language models through multiagent debate. InForty-first International Conference on Machine Learning, 2023

    2023

  5. [13]

    D. Dua, Y . Wang, P. Dasigi, G. Stanovsky, S. Singh, and M. Gardner. Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs.arXiv preprint arXiv:1903.00161, 2019

    1903 arXiv

  6. [14]

    Gu, H.-X

    J.-C. Gu, H.-X. Xu, J.-Y . Ma, P. Lu, Z.-H. Ling, K.-W. Chang, and N. Peng. Model editing harms general abilities of large language models: Regularization to the rescue.arXiv preprint arXiv:2401.04700, 2024

    2024 arXiv

  7. [15]

    N. Guha, J. Nyarko, D. Ho, C. Ré, A. Chilton, A. Chohlas-Wood, A. Peters, B. Waldon, D. Rockmore, D. Zambrano, et al. Legalbench: A collaboratively built benchmark for measuring legal reasoning in large language models.Advances in Neural Information Processing Systems, 36:4412...

    2023

  8. [16]

    N. Guha, J. Nyarko, D. Ho, C. Ré, A. Chilton, A. Chohlas-Wood, A. Peters, B. Waldon, D. Rockmore, D. Zambrano, et al. Legalbench: A collaboratively built benchmark for measuring legal reasoning in large language models.Advances in Neural Information Processing Systems, 36, 2024

    2024

  9. [17]

    Guldimann, A

    P. Guldimann, A. Spiridonov, R. Staab, N. Jovanovi´c, M. Vero, V . Vechev, A.-M. Gueorguieva, M. Balunovi ´c, N. Konstantinov, P. Bielik, et al. Compl-ai framework: A technical inter- pretation and llm benchmarking suite for the eu artificial intelligence act.arXiv preprint ar...

    arXiv

  10. [18]

    Gupta, A

    A. Gupta, A. Rao, and G. Anumanchipalli. Model editing at scale leads to gradual and catastrophic forgetting.arXiv preprint arXiv:2401.07453, 2024

    2024 arXiv

  11. [19]

    Gupta, D

    A. Gupta, D. Sajnani, and G. Anumanchipalli. A unified framework for model editing.arXiv preprint arXiv:2403.14236, 2024

    2024 arXiv

  12. [20]

    Gururangan, A

    S. Gururangan, A. Marasovi ´c, S. Swayamdipta, K. Lo, I. Beltagy, D. Downey, and N. A. Smith. Don’t stop pretraining: Adapt language models to domains and tasks.arXiv preprint arXiv:2004.10964, 2020

    2004 arXiv

  13. [21]

    Hartvigsen, S

    T. Hartvigsen, S. Sankaranarayanan, H. Palangi, Y . Kim, and M. Ghassemi. Aging with grace: Lifelong model editing with discrete key-value adaptors. InProceedings of the 37th Conference on Neural Information Processing Systems (NeurIPS 2023), 2023

    2023

  14. [22]

    J. Hong, N. Lee, and J. Thorne. Orpo: Monolithic preference optimization without reference model.arXiv preprint arXiv:2403.07691, 2024

    2024 arXiv

  15. [23]

    Irving, P

    G. Irving, P. Christiano, and D. Amodei. Ai safety via debate.arXiv preprint arXiv:1805.00899, 2018

    2018 arXiv

  16. [24]

    Jiang, Y

    Y . Jiang, Y . Huang, X. Wang, Y . Qin, S. Deng, N. Zhang, and H. Chen. Easyedit: An easy-to-use knowledge editing framework for large language models. InProceedings of the 62nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, 2024

    2024

  17. [25]

    D. M. Katz, M. J. Bommarito II, and J. Blackman. A general approach for predicting the behavior of the supreme court of the united states.PloS one, 12(4):e0174698, 2017

    2017

  18. [26]

    A. Khan, J. Hughes, D. Valentine, L. Ruis, K. Sachan, A. Radhakrishnan, E. Grefenstette, S. R. Bowman, T. Rocktäschel, and E. Perez. Debating with more persuasive llms leads to more truthful answers.arXiv preprint arXiv:2402.06782, 2024

    2024 arXiv

  19. [27]

    H. J. Kim, H. Cho, J. Kim, T. Kim, K. M. Yoo, and S.-g. Lee. Self-generated in-context learning: Leveraging auto-regressive language models as a demonstration generator.arXiv preprint arXiv:2206.08082, 2022

    2022 arXiv

  20. [28]

    Kojima, S

    T. Kojima, S. S. Gu, M. Reid, Y . Matsuo, and Y . Iwasawa. Large language models are zero-shot reasoners.Advances in neural information processing systems, 35:22199–22213, 2022

    2022

  21. [29]

    H. Li, Y . Chen, Q. Ai, Y . Wu, R. Zhang, and Y . Liu. Lexeval: A comprehensive chinese legal benchmark for evaluating large language models, 2024

    2024

  22. [30]

    Q. Li, X. Liu, Z. Tang, P. Dong, Z. Li, X. Pan, and X. Chu. Should we really edit language models? on the evaluation of edited language models.arXiv preprint arXiv:2410.18785, 2024

    2024 arXiv

  23. [31]

    Li et al

    X. Li et al. Pmet: Precise model editing in a transformer.arXiv preprint arXiv:2308.08742,

    arXiv

  24. [32]

    Y . Li, Z. Lin, S. Zhang, Q. Fu, B. Chen, J.-G. Lou, and W. Chen. Making large language models better reasoners with step-aware verifier.arXiv preprint arXiv:2206.02336, 2022

    2022 arXiv

  25. [33]

    H. Lu, H. Yang, H. Huang, D. Zhang, W. Lam, and F. Wei. Chain-of-dictionary prompting elicits translation in large language models.arXiv preprint arXiv:2305.06575, 2023

    2023 arXiv

  26. [34]

    Mehrotra, M

    A. Mehrotra, M. Zampetakis, P. Kassianik, B. Nelson, H. Anderson, Y . Singer, and A. Karbasi. Tree of attacks: Jailbreaking black-box llms automatically.arXiv preprint arXiv:2312.02119, 2023

    2023 arXiv

  27. [35]

    K. Meng, D. Bau, A. Andonian, and Y . Belinkov. Locating and editing factual associations in gpt. InNeural Information Processing Systems, 2022

    2022

  28. [36]

    K. Meng, A. S. Sharma, A. Andonian, Y . Belinkov, and D. Bau. Mass editing memory in a trans- former. InProceedings of the Eleventh International Conference on Learning Representations (ICLR), 2023. 11

    2023

  29. [37]

    Michael, S

    J. Michael, S. Mahdi, D. Rein, J. Petty, J. Dirani, V . Padmakumar, and S. R. Bowman. Debate helps supervise unreliable experts.arXiv preprint arXiv:2311.08702, 2023

    2023 arXiv

  30. [38]

    Mitchell, C

    E. Mitchell, C. Lin, A. Bosselut, C. D. Manning, and C. Finn. Memory-based model editing at scale. InInternational Conference on Machine Learning, pages 15817–15831. PMLR, 2022

    2022

  31. [39]

    T. D. Nguyen, L. H. Pham, and J. Sun. Uniadapt: A universal adapter for knowledge calibration. arXiv preprint arXiv:2410.00454, 2024

    2024 arXiv

  32. [40]

    Ouyang, J

    L. Ouyang, J. Wu, X. Jiang, D. Almeida, C. Wainwright, P. Mishkin, C. Zhang, S. Agarwal, K. Slama, A. Ray, et al. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35:27730–27744, 2022

    2022

  33. [41]

    Perez, S

    E. Perez, S. Huang, F. Song, T. Cai, R. Ring, J. Aslanides, A. Glaese, N. McAleese, and G. Irving. Red teaming language models with language models.arXiv preprint arXiv:2202.03286, 2022

    2022 arXiv

  34. [42]

    Press, M

    O. Press, M. Zhang, S. Min, L. Schmidt, N. A. Smith, and M. Lewis. Measuring and narrowing the compositionality gap in language models.arXiv preprint arXiv:2210.03350, 2022

    2022 arXiv

  35. [43]

    Rafailov, A

    R. Rafailov, A. Sharma, E. Mitchell, C. D. Manning, S. Ermon, and C. Finn. Direct preference optimization: Your language model is secretly a reward model. InThirty-seventh Conference on Neural Information Processing Systems, 2023

    2023

  36. [44]

    Rafailov, A

    R. Rafailov, A. Sharma, E. Mitchell, C. D. Manning, S. Ermon, and C. Finn. Direct preference optimization: Your language model is secretly a reward model.Advances in Neural Information Processing Systems, 36, 2024

    2024

  37. [45]

    Z. Shao, P. Wang, Q. Zhu, R. Xu, J. Song, X. Bi, H. Zhang, M. Zhang, Y . Li, Y . Wu, et al. Deepseekmath: Pushing the limits of mathematical reasoning in open language models.arXiv preprint arXiv:2402.03300, 2024

    2024 arXiv

  38. [46]

    H. Su, J. Kasai, C. H. Wu, W. Shi, T. Wang, J. Xin, R. Zhang, M. Ostendorf, L. Zettlemoyer, N. A. Smith, et al. Selective annotation makes language models better few-shot learners.arXiv preprint arXiv:2209.01975, 2022

    2022 arXiv

  39. [47]

    Z. Sun, Y . Shen, Q. Zhou, H. Zhang, Z. Chen, D. Cox, Y . Yang, and C. Gan. Principle-driven self-alignment of language models from scratch with minimal human supervision.Advances in Neural Information Processing Systems, 36, 2024

    2024

  40. [48]

    Tuggener, P

    D. Tuggener, P. V on Däniken, T. Peetz, and M. Cieliebak. Ledgar: a large-scale multi-label corpus for text classification of legal provisions in contracts. InProceedings of the twelfth language resources and evaluation conference, pages 1235–1241, 2020

    2020

  41. [49]

    H. Wang, A. Prasad, E. Stengel-Eskin, and M. Bansal. Soft self-consistency improves language model agents.arXiv preprint arXiv:2402.13212, 2024

    2024 arXiv

  42. [50]

    H. Wang, R. Wang, F. Mi, Y . Deng, Z. Wang, B. Liang, R. Xu, and K.-F. Wong. Cue-cot: Chain-of-thought prompting for responding to in-depth dialogue questions with llms.arXiv preprint arXiv:2305.11792, 2023

    2023 arXiv

  43. [51]

    J. Wang, Q. Sun, X. Li, and M. Gao. Boosting language models reasoning with chain-of- knowledge prompting.arXiv preprint arXiv:2306.06427, 2023

    2023 arXiv

  44. [52]

    K. Wang, G. Zhang, Z. Zhou, J. Wu, M. Yu, S. Zhao, C. Yin, J. Fu, Y . Yan, H. Luo, et al. A comprehensive survey in llm (-agent) full stack safety: Data, training and deployment.arXiv preprint arXiv:2504.15585, 2025

    2025 arXiv

  45. [53]

    P. Wang, Z. Li, N. Zhang, Z. Xu, Y . Yao, Y . Jiang, P. Xie, F. Huang, and H. Chen. Wise: Rethinking the knowledge memory for lifelong model editing of large language models.arXiv preprint arXiv:2405.14768, 2024

    2024 arXiv

  46. [54]

    Wang and P

    R. Wang and P. Li. Lemoe: Advanced mixture of experts adaptor for lifelong model editing of large language models.arXiv preprint arXiv:2406.20030, 2024. 12

    2024 arXiv

  47. [55]

    Wang and P

    R. Wang and P. Li. Memoe: Enhancing model editing with mixture of experts adaptors.arXiv preprint arXiv:2405.19086, 2024

    2024 arXiv

  48. [56]

    X. Wang, J. Wei, D. Schuurmans, Q. Le, E. Chi, S. Narang, A. Chowdhery, and D. Zhou. Self-consistency improves chain of thought reasoning in language models.arXiv preprint arXiv:2203.11171, 2022

    2022 arXiv

  49. [57]

    Q. Wu, G. Bansal, J. Zhang, Y . Wu, B. Li, E. Zhu, L. Jiang, X. Zhang, S. Zhang, J. Liu, et al. Autogen: Enabling next-gen llm applications via multi-agent conversation.arXiv preprint arXiv:2308.08155, 2023

    2023 arXiv

  50. [58]

    H. Xu, A. Sharaf, Y . Chen, W. Tan, L. Shen, B. Van Durme, K. Murray, and Y . J. Kim. Contrastive preference optimization: Pushing the boundaries of llm performance in machine translation.arXiv preprint arXiv:2401.08417, 2024

    2024 arXiv

  51. [59]

    M. Xue, D. Liu, W. Lei, X. Ren, B. Yang, J. Xie, Y . Zhang, D. Peng, and J. Lv. Dynamic voting for efficient reasoning in large language models. InFindings of the Association for Computational Linguistics: EMNLP 2023, pages 3085–3104, 2023

    2023

  52. [60]

    J. Yu, X. Lin, Z. Yu, and X. Xing. Gptfuzzer: Red teaming large language models with auto-generated jailbreak prompts.arXiv preprint arXiv:2309.10253, 2023

    2023 arXiv

  53. [61]

    L. Yu, Q. Chen, J. Zhou, and L. He. Melo: Enhancing model editing with neuron-indexed dynamic lora. InProceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19449–19457, 2024

    2024

  54. [62]

    Zhang, A

    Z. Zhang, A. Zhang, M. Li, and A. Smola. Automatic chain of thought prompting in large language models.arXiv preprint arXiv:2210.03493, 2022

    2022 arXiv

  55. [63]

    Zheng, H

    J. Zheng, H. Wang, A. Zhang, T. D. Nguyen, J. Sun, and T.-S. Chua. Ali-agent: Assessing llms’ alignment with human values via agent-based evaluation.arXiv preprint arXiv:2405.14125, 2024

    2024 arXiv

  56. [64]

    Jason eats a burger on train

    L. Zheng, N. Guha, B. R. Anderson, P. Henderson, and D. E. Ho. When does pretraining help? assessing self-supervised learning for law and the casehold dataset of 53,000+ legal holdings. InProceedings of the eighteenth international conference on artificial intelligence and law...

    2021

  57. [2023]

    Submitted on 17 Aug 2023, last revised on 11 Mar 2024

    2023

Pith tools

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