Pith. sign in

REVIEW 4 major objections 8 minor 52 references

WenyanGPT: A Large Language Model for Classical Chinese Tasks

T0 review · 4 major / 8 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 WenyanGPT, an 8B model adapted to Classical Chinese by continued pre-training and instruction tuning, reports the best score on all six tasks of the new WenyanBENCH benchmark, beating GPT-4o and DeepSeek-V3.

desk verdict Useful open resources for Classical Chinese NLP, but the headline benchmark is contaminated by training data overlap, so the performance claim doesn't stand without an independent test set. read the letter →

arxiv 2504.20609 v1 pith:6HLBF5JH submitted 2025-04-29 cs.CL

Xinyu Yao , Mengdi Wang , Bo Chen , Xiaobing Zhao This is my paper · ORCID
classification cs.CL
keywords ClassicalChineselargelanguagemodelcontinuedpre-traininginstructionfine-tuningbenchmarknamedentityrecognitionpunctuationpart-of-speechtagging
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

This paper tries to establish that a modest open-weight language model can be turned into the best available system for Classical Chinese by domain-specific continued pre-training and instruction tuning. It introduces WenyanGPT, built from LLaMA3-8B-Chinese, and WenyanBENCH, a six-task benchmark covering punctuation, part-of-speech tagging, named entity recognition, translation, word explanation, and reverse dictionary. On WenyanBENCH, WenyanGPT outperforms strong general-purpose models including GPT-4o and DeepSeek-V3 on every task, with the largest gaps in NER and punctuation. The authors also release their pre-training corpus, instruction data, and benchmark so the result can be reproduced and extended.

What carries the argument

The engine of the method is a two-stage training pipeline: continued pre-training of LLaMA3-8B-Chinese on a cleaned ~16GB Classical Chinese corpus, followed by supervised instruction tuning on ~1.85M examples. The instruction data are built by a semi-automatic protocol in which human-written templates are expanded by LLMs, reverse reasoning generates new prompts from existing pairs, and manual plus automatic filtering keeps high-quality samples. WenyanBENCH provides the measuring rod, with six tasks scored by precision, recall, and F1 for understanding tasks and by BLEU and BERT-Score for generation tasks.

What would settle it

Construct a held-out Classical Chinese test set from sources that are absent from both the ~16GB pre-training corpus and the ~1.85M instruction set, for example newly published transcriptions of excavated manuscripts, and compare WenyanGPT against GPT-4o and DeepSeek-V3; the central claim fails if the performance margins shrink to noise. A complementary check is to measure n-gram overlap between WenyanBENCH items and the released training data.

Watch

Extended reading notes

Core claim

The paper's central claim is that a domain-adapted 8B model can outperform far larger general models on Classical Chinese understanding and generation. WenyanGPT achieves F1 of 91.16 in NER versus 68.42 for GPT-4o, 75.66 in punctuation versus 59.01 for DeepSeek-V3, and 89.1 in part-of-speech tagging versus 81.75 for GPT-4o; in generation, its translation BLEU1 is 0.47 versus 0.41 for GPT-4o, word-explanation BLEU1 is 0.35 versus 0.19 for GPT-4o, and reverse-dictionary BERT-Score F1 is 75.39 versus 71.88 for DeepSeek-V3.

Load-bearing premise

The comparison assumes WenyanBENCH measures genuine Classical Chinese ability, but because the benchmark shares its data sources with the model's instruction-tuning set, the scores may partly reflect overlap with training rather than general skill.

Editorial extensions

If this is right

  • If the results hold, a specialized 8B-parameter model can outperform much larger general-purpose models on Classical Chinese understanding and generation, making domain adaptation a viable path for low-resource historical languages.
  • WenyanBENCH gives researchers a shared six-task test with fixed metrics, so future Classical Chinese models can be compared directly instead of across inconsistent private datasets.
  • Releasing the pre-training corpus, instruction data, and benchmark makes the entire pipeline reproducible and lets others test the same recipe on adjacent tasks such as poetry appreciation or ancient-text question answering.
  • Multitask instruction tuning appears to stabilize fine-grained predictions, with WenyanGPT staying above 80% F1 across NER entity categories, suggesting the model internalizes shared linguistic structure across tasks.

Reading between the lines

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

  • Because WenyanBENCH is drawn from the same sources as the instruction-tuning data, the reported margins are best read as evidence of strong domain adaptation rather than proof of general Classical Chinese competence; an independent benchmark built from disjoint sources would be needed to separate the two.
  • The same data-construction pipeline—manual templates, LLM expansion, reverse reasoning, filtering—could be ported to other historical or low-resource languages, such as Literary Japanese or Medieval Latin, where parallel and tagged corpora are scarce.
  • The authors' planned move to multimodal input could serve as a discriminating test: if WenyanGPT's gains are mostly memorization of training text, adding image-based manuscript input should not transfer; if they reflect learned syntax and semantics, multimodal extension should succeed.
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 / 8 minor

Summary. The paper presents WenyanGPT, an 8B-parameter LLM for Classical Chinese, built by continued pre-training and instruction fine-tuning on a LLaMA3-8B-Chinese base. The authors contribute a 16GB Classical Chinese pre-training corpus, a 1.85M-example instruction tuning dataset, and WenyanBENCH, a six-task benchmark. They report that WenyanGPT outperforms general-purpose LLMs and one Classical Chinese model on all six tasks, with particularly large margins in NER and punctuation. The paper also describes a pipeline for generating domain-specific instruction data.

Significance. If the empirical claims were sound, this would be a useful contribution: it addresses an under-served language domain, releases a model and datasets, and proposes a generalizable instruction-data construction method. The transparency about data sources and the decision to release the resources are commendable. However, the central evaluation is compromised by the overlap between the benchmark and training data, and by the lack of statistical rigor, so the headline claim of 'significantly outperforms' is not currently supported. The resources themselves may be valuable to the community, but the paper's main empirical conclusion requires a fundamentally more trustworthy evaluation.

major comments (4)
  1. [§4 (Benchmarking), Tables 3 and 5] The benchmark WenyanBENCH is explicitly stated to share the same data sources as the instruction fine-tuning data, with only 'duplicate data removal' as a safeguard. The fine-tuning set (Table 3) and the benchmark (Table 5) draw from the same corpora (e.g., Daizhige, classical-modern, wenyanguji.com, and for NER a 'Self-built' source). Because WenyanGPT was fine-tuned on 1.85M examples from these sources, its high scores on WenyanBENCH (e.g., NER F1 91.16 vs. 68.42 for GPT-4o) may reflect distribution overlap rather than general Classical Chinese competence. The paper provides no details on how duplicates were detected, no overlap statistics, and no evidence that the advantage persists on a test set from sources not used in training. This is a load-bearing flaw: without an independent test set, the abstract and §5.2 claim of 'significantly outperforms' is not supported.
  2. [§5.2, Tables 6 and 7] All reported results are single-point estimates with no error bars, confidence intervals, or significance tests. Several tasks have small evaluation sets (e.g., part-of-speech tagging has 1,247 examples; Table 5), so the observed gaps (e.g., punctuation F1 75.66 vs. 59.01 for DeepSeek-V3) could be unstable. To substantiate the claim of 'significant' superiority, the paper should report variance across multiple evaluation runs or bootstrap resampling, and perform a significance test such as paired bootstrap or McNemar's test for the categorical tasks.
  3. [§5.1 Baselines] The comparison includes only one Classical Chinese-specific model, Xunzi-Qwen1.5-7B-Chat. TongGu (Cao et al., 2024) is described in the related work as a Classical Chinese LLM capable of punctuation, translation, and appreciation, but it is not evaluated. Given that the paper claims to surpass 'current advanced LLMs in various Classical Chinese tasks' and that TongGu is a direct competing model, its absence makes the headline comparison incomplete. The authors should include existing Classical Chinese models as baselines, or justify their exclusion.
  4. [§4 Evaluation Method and §5.1 Data and Evaluation] The evaluation protocol is not described with enough detail to ensure fair comparison. The paper does not specify the prompt format used for each baseline, whether any few-shot examples were provided, how structured outputs (e.g., NER JSON) were parsed and scored, or how BLEU and BERTScore were computed (tokenization, reference selection). These details are necessary for reproducibility and for assessing whether the reported margins could be due to prompt or parsing artifacts rather than model capability.
minor comments (8)
  1. [Table 3] The entry '107,3017' appears to be a formatting error; the intended number is likely '1,073,017' given the stated total of 1,847,922.
  2. [Table 5] The caption says 'Data Sources and Detailed Statistics of Wenyan-Bench' but the table only lists tasks and numbers; either add a source column or revise the caption.
  3. [Figure 4] The x-axis label 'characte rs' contains an unwanted space, and the NER subcategory names are not defined in the main text.
  4. [Throughout] There are several typos and minor language errors: 'outpermforms' in the Conclusion, 'ccessing' in §5.3, 'We attend to explore' (likely 'intend') in the Conclusion, 'the forth year' in §5.3, and inconsistent capitalization of 'WenyanBench'/'WenyanBENCH'.
  5. [Abstract] The abstract contains LaTeX artifacts ('\footnote') and footnote markers that do not render; these should be removed or resolved.
  6. [References] The BERT citation is given as 'Kenton and Toutanova, 2019' but the actual paper is Devlin et al., 2019; the citation format should be corrected.
  7. [Table 8] In the DeepSeek-V3 output, there is a spurious '/ns' token (an empty field) which appears to be a formatting error.
  8. [§5.2, Figure 5] The radar plot normalizes values to 0-1 but the normalization method is not specified; please clarify how the scaling is performed.

Circularity Check

1 steps flagged · score 8.0 of 10

WenyanBENCH is drawn from the same data sources as WenyanGPT's instruction fine-tuning data, so the headline 'significantly outperforms' claim is an in-distribution evaluation rather than an independent test.

  1. fitted input called prediction [Section 4 (Benchmarking Classical Chinese Tasks), WenyanBench paragraph; cf. Table 3 (instruction fine-tuning data sources) and Table 5 (WenyanBench statistics).]
    "WenyanBench shares the same data sources as the instruction fine-tuning data and has undergone duplicate data removal, as well as validation by both manual and LLMs. For quality control, we sample a subset of the data."

    WenyanGPT is fine-tuned on about 1.85M instruction examples (Table 3) built from Daizhige, evahan, classical-modern, wenyanguji.com, gushiwen.com, chinese-dictionary, and chinese-xinhua. WenyanBENCH (Table 5) is sampled from the same sources. The model is therefore evaluated on the same distribution it was optimized on, while all baselines see this distribution for the first time. The large reported gaps (NER F1 91.16 vs 68.42 for GPT-4o; punctuation F1 75.66 vs 59.01 for DeepSeek-V3) are not independent evidence of general Classical Chinese ability; they measure how well the model fits the training corpora. Exact-duplicate removal does not remove near-duplicates or source-level overlap, and no overlap statistics are reported.

full rationale

The paper's central empirical claim is the comparison on WenyanBENCH. Section 4 states directly that WenyanBENCH 'shares the same data sources as the instruction fine-tuning data,' and Tables 3 and 5 show the same corpora feeding both SFT and evaluation. Because WenyanGPT was instruction-tuned on those corpora, its WenyanBENCH scores largely reflect in-distribution performance rather than generalization to unseen Classical Chinese; the baselines were never exposed to this distribution, making the comparison unbalanced. This is a benchmark-design circularity rather than a derivation error: the 'prediction' of superiority is built, by construction, on training/test source overlap. The released resources may still be useful, but the abstract and Section 5.2 headline claim is not supported as an independent evaluation. No other load-bearing circularity (e.g., self-citation chains or uniqueness arguments) was found; the problem is concentrated in the benchmark construction.

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

The paper does not introduce new theoretical entities or forces. The free parameters are standard training hyperparameters, not fitted to the benchmark. The critical assumptions are about the validity of the benchmark and the quality of the training data, which are the main uncertain premises.

free parameters (5)
  • learning rate (pre-training) = 1.0e-4
    Chosen hyperparameter for continued pre-training; not fitted to WenyanBENCH, but it affects model quality.
  • learning rate (fine-tuning) = 1.0e-4
    Chosen hyperparameter for SFT; a standard setting, not tuned on the benchmark.
  • training epochs (both phases) = 1
    Single epoch for both continued pre-training and fine-tuning; chosen to control overfitting, but interacts with data size.
  • per-device batch size (pre-training / fine-tuning) = 16 / 8
    Computational choices that influence training dynamics, not scientific parameters.
  • warmup ratio = 0.1
    Standard cosine schedule setting for both training phases.
assumptions (5)
  • ad hoc to paper WenyanBENCH, drawn from the same sources as the training data with only duplicate removal, is an unbiased estimate of Classical Chinese task ability.
    This assumption is load-bearing for the reported performance comparison and is stated in Section 4. If false, the evaluation is circular.
  • domain assumption LLaMA3-8B-Chinese is a suitable base model for Classical Chinese adaptation.
    The model starts from LLaMA3-8B-Chinese; the choice is not experimentally justified within the paper.
  • domain assumption The assembled 16GB Classical Chinese corpus is clean and high-quality enough for domain adaptation.
    The paper states the corpus is cleaned and deduplicated but provides no quantitative quality checks or details on the cleaning process.
  • domain assumption LLM-generated instruction data (from GPT and Qwen models) is accurate and useful for SFT.
    The instruction data construction relies on LLMs to generate answers and instructions, with manual filtering described only at a high level.
  • domain assumption Manual and LLM-based filtering ensures the quality of the final instruction dataset.
    The quality control process is described qualitatively; no inter-annotator agreement or systematic validation is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of WenyanGPT: A Large Language Model for Classical Chinese Tasks." pith.science (2026). https://pith.science/paper/6HLBF5JH

@misc{pith2026250420609,
  author       = {Pith},
  title        = {Pith review of: WenyanGPT: A Large Language Model for Classical Chinese Tasks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6HLBF5JH}},
  note         = {Machine review of arXiv:2504.20609}
}
read the original abstract

Classical Chinese, as the core carrier of Chinese culture, plays a crucial role in the inheritance and study of ancient literature. However, existing natural language processing models primarily optimize for Modern Chinese, resulting in inadequate performance on Classical Chinese. This paper presents a comprehensive solution for Classical Chinese language processing. By continuing pre-training and instruction fine-tuning on the LLaMA3-8B-Chinese model, we construct a large language model, WenyanGPT, which is specifically designed for Classical Chinese tasks. Additionally, we develop an evaluation benchmark dataset, WenyanBENCH. Experimental results on WenyanBENCH demonstrate that WenyanGPT significantly outperforms current advanced LLMs in various Classical Chinese tasks. We make the model's training data, instruction fine-tuning data\footnote, and evaluation benchmark dataset publicly available to promote further research and development in the field of Classical Chinese processing.

Figures

Figures reproduced from arXiv: 2504.20609 by the authors.

Figure 1
Figure 1. Examples of tasks from WenyanGPT. The model demonstrates advanced knowledge in Classical Chinese and shows strong performance in both Classical Chinese understanding and generation tasks. urgent need for cultural inheritance and academic development. Early research in Classical Chinese language processing focused on tasks such as punctuation, word segmentation, part-of-speech tagging, named entity recognition and tr… view at source ↗
Figure 2
Figure 2. Overall Training Framework of WenyanGPT. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Instruction Fine-Tuning Data Construction Process. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The F1 scores of WenYanGPT for the subcategories of understanding tasks (including Punctuation, [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Radar plot showing model performance on WenyanBench, with values normalized to a 0-1 scale. WenyanGPT’s precision, recall, and F1 score all exceed 90%, while the second-best model, GPT￾4o, fails to surpass 80% in any of these metrics. In the punctuation task, WenyanGPT…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 27 canonical work pages

  1. [1]

    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 INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, and 1 others. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774

    arXiv 2023

  4. [4]

    Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, and 1 others. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33:1877--1901

    work page 2020

  5. [5]

    Jiahuan Cao, Dezhi Peng, Yongxin Shi, Zongyuan Jiang, and Lianwen Jin. 2023. https://api.semanticscholar.org/CorpusID:265158334 Translating ancient chinese to modern chinese at scale: A large language model-based approach . In International Conference on Algorithmic Learning Theory

    work page 2023

  6. [6]

    Jiahuan Cao, Dezhi Peng, Peirong Zhang, Yongxin Shi, and 1 others. 2024. https://api.semanticscholar.org/CorpusID:271039137 Tonggu: Mastering classical chinese understanding with knowledge-grounded large language models . In Conference on Empirical Methods in Natural Language Processing

    work page 2024

  7. [7]

    Bolin Chang, Yiguo Yuan, Bin Li, Zhixing Xu, and 1 others. 2024. Automatic word segmentation and part-of-speech tagging for classical chinese based on radicals. Data Analysis and Knowledge Discovery, 8(11):102--113

    work page 2024

  8. [8]

    Ning Cheng, Bin Li, Liming Xiao, Changwei Xu, and 1 others. 2020. Integration of automatic sentence segmentation and lexical analysis of ancient chinese based on bilstm-crf model. In Proceedings of LT4HALA 2020-1st Workshop on Language Technologies for Historical and Ancient Languages, pages 52--58

    work page 2020

Show all 52 references
  1. [9]

    Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, and 1 others. 2023. Palm: Scaling language modeling with pathways. Journal of Machine Learning Research, 24(240):1--113

    2023

  2. [10]

    Jiaxi Cui, Zongjian Li, Yang Yan, Bohua Chen, and Li Yuan. 2023. Chatlaw: Open-source legal large language model with integrated external knowledge bases. CoRR

    2023

  3. [11]

    Yuhao Dan, Zhikai Lei, Yiyang Gu, Yong Li, and 1 others. 2023. https://arxiv.org/abs/2308.02773 Educhat: A large-scale language model-based chatbot system for intelligent education . Preprint, arXiv:2308.02773

    2023 arXiv

  4. [12]

    DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, and 181 others. 2025. https://arxiv.org/abs/24...

    2025 arXiv

  5. [13]

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, and 1 others. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783

    2024 arXiv

  6. [14]

    Richter, Quentin G

    Kshitij Gupta, Benjamin Th'erien, Adam Ibrahim, Mats L. Richter, Quentin G. Anthony, Eugene Belilovsky, Irina Rish, and Timoth \'e e Lesort. 2023. https://api.semanticscholar.org/CorpusID:260704601 Continual pre-training of large language models: How to (re)warm your model? Ar...

    2023 arXiv

  7. [15]

    Chuen-Min Huang, Kuo-Lin Lu, Yi-Ying Cheng, and Yu-Chen Peng. 2020. Generating chinese classical poetry with quatrain generation model (qgm) using encoder-decoder lstm. In 2020 IEEE International Conference on Big Data (Big Data), pages 5700--5702. IEEE

    2020

  8. [16]

    Hen - Hsen Huang, Chuen - Tsai Sun, and Hsin - Hsi Chen. 2010. https://aclanthology.org/W10-4103/ Classical chinese sentence segmentation . In CIPS-SIGHAN Joint Conference on Chinese Language Processing, Beijing, China, August 28-29, 2010

    2010

  9. [17]

    Liang Huang, Yinan Peng, Huan Wang, and Zhenyu Wu. 2002. Statistical part-of-speech tagging for classical chinese. In Proceedings of the 5th International Conference on Text, Speech and Dialogue, TSD '02, page 115–122, Berlin, Heidelberg. Springer-Verlag

    2002

  10. [18]

    Quzhe Huang, Mingxu Tao, Zhenwei An, Chen Zhang, Cong Jiang, Zhibin Chen, Zirui Wu, and Yansong Feng. 2023. https://api.semanticscholar.org/CorpusID:258865862 Lawyer llama technical report

    2023

  11. [19]

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, and 1 others. 2024. Gpt-4o system card. arXiv preprint arXiv:2410.21276

    2024 arXiv

  12. [20]

    Richter, Quentin Anthony, Timoth \'e e Lesort, Eugene Belilovsky, and Irina Rish

    Adam Ibrahim, Benjamin Th'erien, Kshitij Gupta, Mats L. Richter, Quentin Anthony, Timoth \'e e Lesort, Eugene Belilovsky, and Irina Rish. 2024. https://api.semanticscholar.org/CorpusID:268379604 Simple and scalable strategies to continually pre-train large language models . Ar...

    2024 arXiv

  13. [21]

    Zixuan Ke, Yijia Shao, Haowei Lin, Tatsuya Konishi, Gyuhak Kim, and Bin Liu. 2023. https://api.semanticscholar.org/CorpusID:258079422 Continual pre-training of language models . In International Conference on Learning Representations

    2023

  14. [22]

    Jacob Devlin Ming-Wei Chang Kenton and Lee Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of naacL-HLT, volume 1, page 2. Minneapolis, Minnesota

    2019

  15. [23]

    Eric Lehman, Evan Hernandez, Diwakar Mahajan, Jonas Wulff, Micah J Smith, Zachary Ziegler, Daniel Nadler, Peter Szolovits, Alistair Johnson, and Emily Alsentzer. 2023. https://api.semanticscholar.org/CorpusID:256900662 Do we still need clinical language models? ArXiv, abs/2302.08091

    2023 arXiv

  16. [24]

    N Li. 2018. Automatic extraction of alias in ancient local chronicles based on conditional random fields. J. Chin. Inf. Process, 32:41

    2018

  17. [25]

    Chang Liu, Dongbo Wang, Zhixiao Zhao, Die Hu, Mengcheng Wu, Litao Lin, Si Shen, Bin Li, Jiangfeng Liu, Hai Zhang, and Lianzheng Zhao. 2023 a . https://api.semanticscholar.org/CorpusID:258179714 Sikugpt: A generative pre-trained model for intelligent information processing of a...

    2023 arXiv

  18. [26]

    Dayiheng Liu, Jiancheng Lv, Kexin Yang, and Qian Qu. 2018. https://api.semanticscholar.org/CorpusID:51972516 Ancient–modern chinese translation with a new large training dataset . ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP), 19:1 -- 13

    2018

  19. [27]

    Liu, Donghao Li, He Cao, Tianhe Ren, Zeyi Liao, and Jiamin Wu

    June M. Liu, Donghao Li, He Cao, Tianhe Ren, Zeyi Liao, and Jiamin Wu. 2023 b . https://api.semanticscholar.org/CorpusID:262943261 Chatcounselor: A large language models for mental health support . ArXiv, abs/2309.15461

    2023 arXiv

  20. [28]

    Zhuang Liu, Degen Huang, Kaiyu Huang, Zhuang Li, and Jun Zhao. 2020. https://api.semanticscholar.org/CorpusID:265038146 Finbert: A pre-trained financial language representation model for financial text mining . In International Joint Conference on Artificial Intelligence

    2020

  21. [29]

    Alec Radford and Karthik Narasimhan. 2018. https://api.semanticscholar.org/CorpusID:49313245 Improving language understanding by generative pre-training

    2018

  22. [30]

    Alec Radford, Jeff Wu, Rewon Child, David Luan, and 1 others. 2019. https://api.semanticscholar.org/CorpusID:160025533 Language models are unsupervised multitask learners

    2019

  23. [31]

    Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao, Abu Awal Md Shoeb, and 1 others. 2023. https://api.semanticscholar.org/CorpusID:271601672 Beyond the imitation game: Quantifying and extrapolating the capabilities of language models . Trans. Mach. Learn. Res., 2023

    2023

  24. [32]

    Ross Taylor, Marcin Kardas, Guillem Cucurull, Thomas Scialom, and 1 others. 2022. https://api.semanticscholar.org/CorpusID:253553203 Galactica: A large language model for science . ArXiv, abs/2211.09085

    2022 arXiv

  25. [33]

    Huishuang Tian, Kexin Yang, Dayiheng Liu, and Jiancheng Lv. 2020. https://api.semanticscholar.org/CorpusID:221879140 Anchibert: A pre-trained model for ancient chinese language understanding and generation . 2021 International Joint Conference on Neural Networks (IJCNN), pages 1--8

    2020

  26. [34]

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, and 1 others. 2023. https://api.semanticscholar.org/CorpusID:257219404 Llama: Open and efficient foundation language models . ArXiv, abs/2302.13971

    2023 arXiv

  27. [35]

    Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N

    Ashish Vaswani, Noam M. Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. https://api.semanticscholar.org/CorpusID:13756489 Attention is all you need . In Neural Information Processing Systems

    2017

  28. [36]

    Dongbo Wang, Chang Liu, Zhixiao Zhao, Si Shen, and 1 others. 2023 a . Gujibert and gujigpt: Construction of intelligent information processing foundation language models for ancient texts. arXiv preprint arXiv:2307.05354

    2023 arXiv

  29. [37]

    Dongbo Wang, Chang Liu, Zihe Zhu, Jiangfeng Liu, and 1 others. 2022. Sikubert and sikuroberta: Construction and application of pre-trained models for the siku quanshu in the field of digital humanities. Library Tribune, 42(06):31--43

    2022

  30. [38]

    Hao Wang, Chi-Liang Liu, Nuwa Xi, Zewen Qiang, Sendong Zhao, Bing Qin, and Ting Liu. 2023 b . https://api.semanticscholar.org/CorpusID:258170497 Huatuo: Tuning llama model with chinese medical knowledge . ArXiv, abs/2304.06975

    2023 arXiv

  31. [39]

    Hongbin Wang, Haibing Wei, Jianyi Guo, and Liang Cheng. 2019. Ancient chinese sentence segmentation based on bidirectional lstm+ crf model. Journal of advanced computational intelligence and intelligent informatics, 23(4):719--725

    2019

  32. [40]

    Honglin Xiong, Sheng Wang, Yitao Zhu, Zihao Zhao, Yuxiao Liu, Linlin Huang, Qian Wang, and Dinggang Shen. 2023. https://api.semanticscholar.org/CorpusID:257912795 Doctorglm: Fine-tuning your chinese doctor is not a herculean task . ArXiv, abs/2304.01097

    2023 arXiv

  33. [41]

    Rui Yan, Cheng-Te Li, Xiaohua Hu, and Ming Zhang. 2016. Chinese couplet generation with neural network structures. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2347--2357

    2016

  34. [42]

    Aiyuan Yang, Bin Xiao, Bingning Wang, Borong Zhang, and 1 others. 2023 a . Baichuan 2: Open large-scale language models. arXiv preprint arXiv:2309.10305

    2023 arXiv

  35. [43]

    An Yang, Baosong Yang, Beichen Zhang, and 1 others. 2024 a . Qwen2. 5 technical report. arXiv preprint arXiv:2412.15115

    2024 arXiv

  36. [44]

    Songhua Yang, Hanjie Zhao, Senbin Zhu, Guangyu Zhou, and 1 others. 2024 b . Zhongjing: Enhancing the chinese medical capabilities of large language model through expert feedback and real-world multi-turn dialogue. In Proceedings of the AAAI Conference on Artificial Intelligenc...

    2024

  37. [45]

    Yi Yang, Yixuan Tang, and Kar Yan Tam. 2023 b . https://arxiv.org/abs/2309.13064 Investlm: A large language model for investment using financial domain instruction tuning . Preprint, arXiv:2309.13064

    2023 arXiv

  38. [46]

    Xiaoyuan Yi, Ruoyu Li, and Maosong Sun. 2017. Generating chinese classical poems with rnn encoder-decoder. In Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data: 16th China National Conference, CCL 2017, and 5th Internationa...

    2017

  39. [47]

    Y Yuan, D Wang, S Huang, and B Li. 2019. The comparative study of different tagging sets on entity extraction of classical books. Data Analysis and Knowledge Discovery, 3(03):57--65

    2019

  40. [48]

    Shengbin Yue, Wei Chen, Siyuan Wang, Bingxuan Li, and 1 others. 2023. https://api.semanticscholar.org/CorpusID:262064568 Disc-lawllm: Fine-tuning large language models for intelligent legal services . ArXiv, abs/2309.11325

    2023 arXiv

  41. [49]

    Team Glm Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Diego Rojas, and 1 others. 2024. https://api.semanticscholar.org/CorpusID:270562306 Chatglm: A family of large language models from glm-130b to glm-4 all tools . ArXiv, abs/2406.12793

    2024 arXiv

  42. [50]

    Jundong Zhang, Songhua Yang, Jiangfeng Liu, and Qi Huang. 2024. Aigc empowering the revitalization of ancient books on traditional chinese medicine:building the huang-di large language model. Library Tribune, 44(10):103--112

    2024

  43. [51]

    Xuanyu Zhang, Qing Yang, and Dongliang Xu. 2023 a . https://api.semanticscholar.org/CorpusID:258833440 Xuanyuan 2.0: A large chinese financial chat model with hundreds of billions parameters . Proceedings of the 32nd ACM International Conference on Information and Knowledge Management

    2023

  44. [52]

    Yiqin Zhang, Sanhong Deng, Qi Zhang, Dongbo Wang, and Hongcun Gong. 2023 b . Comparative analysis of language models for linguistic examination of ancient chinese classics: A case study of zuozhuan corpus. In 2023 International Conference on Asian Language Processing (IALP), p...

    2023

Pith tools

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