Pith. sign in

REVIEW 4 major objections 6 minor 52 references

PerMedCQA: Benchmarking Large Language Models on Medical Consumer Question Answering in Persian Language

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read PerMedCQA introduces the first large-scale Persian benchmark for consumer medical question answering, with 68,138 QA pairs and an LLM-based judge.

desk verdict The PerMedCQA dataset is a genuine new resource, but the paper as submitted has no results section, so the benchmark claims are unsupported. read the letter →

arxiv 2505.18331 v1 pith:ZNPNTZ5C submitted 2025-05-23 cs.CL cs.AI

Naghmeh Jamali , Milad Mohammadi , Danial Baledi , Zahra Rezvani , Hesham Faili This is my paper · ORCID
classification cs.CLcs.AI
keywords PerMedCQAPersianmedicalQAconsumerhealthquestionsbenchmarkdatasetLLM-as-a-judgeICD-11low-resourceNLPquestionanswering
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

PerMedCQA aims to fill a gap in medical question answering: almost all consumer-oriented benchmarks are in English, and none target Persian. The paper builds a dataset of 68,138 question-answer pairs drawn from four Persian forums where verified physicians answer real patient questions, cleaned from 87,780 raw entries and annotated with ICD-11 disease categories and 25 question types. To score open-ended answers, it introduces MedJudge, an LLM grader that compares model responses with the physician answers using a rubric for correctness, coverage, and clinical impact, and reports 75% agreement with physicians on a 100-item sample. The paper also benchmarks 16 LLMs and tests prompting and fine-tuning strategies, though the results section is currently marked as under development. If the resource holds up, it would give Persian-speaking users a way to measure and improve medical AI systems in their own language.

What carries the argument

The load-bearing mechanism is MedJudge, a large-language-model grader prompted with a three-part rubric: Correctness (correct, partially correct, incorrect, contradictory), Coverage (equal, model subset, expert subset, no overlap), and Clinical Impact (negligible, moderate, significant, critical). MedJudge is given the patient question, the verified-physician answer as gold, and the model answer, and is explicitly forbidden to use its own medical knowledge or outside sources. The dataset pipeline is the other half of the machinery: rule-based filters remove short, duplicate, or non-textual entries; GPT-4o-mini flags personally identifiable information; the same LLM assigns ICD-11 categories; and the benchmark is split into train, evaluation, and test sets stratified by ICD-11 category.

What would settle it

Take a random sample of PerMedCQA gold answers and have independent board-certified physicians, blinded to the forum posts, rate each answer's medical correctness against current guidelines; if a substantial fraction are judged inaccurate, incomplete, or outdated, the reference standard that MedJudge scores against is not reliable.

Watch

Extended reading notes

Core claim

The central claim is that PerMedCQA is the first Persian-language benchmark for consumer medical question answering, and that it is large and realistic enough to support meaningful evaluation. The dataset contains 68,138 QA pairs from four public Persian medical forums, restricted to questions from real patients and answers from verified physicians, and enriched with ICD-11 labels, 25 standardized question types, patient age and sex, physician specialty, and source metadata. The paper further claims that open-ended medical answers can be reliably scored by MedJudge, an LLM-based rubric grader that is instructed to judge only against the expert answer and that reached 75% agreement with board-certified physicians on the correctness dimension of a 100-item subset. On this basis the authors argue that multilingual and instruction-tuned models vary substantially on Persian consumer health questions and that prompt-based techniques and fine-tuning can change model output quality, with the detailed results deferred to a future revision.

Load-bearing premise

The benchmark treats answers posted by verified physicians on four public forums as the gold standard, and MedJudge is told to judge only against those answers; if those posts are medically wrong or incomplete, every model score inherits that error.

Editorial extensions

If this is right

  • Persian-speaking patients can have medical QA systems evaluated on the kinds of questions they actually ask, rather than on translated exam questions.
  • Researchers get a public, de-identified resource for fine-tuning and for comparing models in a low-resource language.
  • The ICD-11 and question-type annotations allow analysis of which medical topics and question intents are hardest for current LLMs.
  • The reported gender and topic distributions suggest that any deployed system must handle sexual health, digestive, and skin questions well to serve Persian forum users.
  • MedJudge, if its agreement with physicians generalizes, offers an alternative to BLEU and ROUGE for open-ended medical evaluation in other languages.

Reading between the lines

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

  • The reported 75% correctness agreement and quadratic Cohen's kappa of 0.42 between MedJudge and physicians are moderate; if a different judge model or prompt changes rankings, part of what is being measured is the judge, not the medical QA systems.
  • Because the gold answers come from forum posts and MedJudge is barred from external knowledge, any systematic error in those posts, such as outdated or overly cautious advice, becomes part of the benchmark's definition of correctness.
  • The text-only scope leaves out skin and visual-system questions, which are among the most common categories in the dataset; a multimodal extension would likely change model rankings.
  • The translation-pivot experiments imply that English-centric models may lose information in Persian; testing the same models on native Persian outputs after fine-tuning would isolate how much of the gap is language coverage rather than medical knowledge.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper introduces PerMedCQA, a Persian-language corpus of 68,138 consumer health question-answer pairs collected from four public forums, cleaned with rule-based filtering and LLM-based PII detection, annotated with ICD-11 categories and question types, and split into stratified train/dev/test sets. It also proposes MedJudge, an LLM-based rubric grader that compares model answers against forum expert answers on Correctness, Coverage, and Clinical Impact, with a 100-item human validation study. The paper claims to benchmark 16 LLMs using zero-shot prompting, role-based prompting, pivot translation, and LoRA fine-tuning; however, Section 5, which would contain all results, is explicitly marked as under development, and no model scores or comparisons are reported anywhere in the manuscript.

Significance. If completed, PerMedCQA would be a valuable resource for Persian medical NLP: it is large, publicly released, derived from real consumer questions, and includes structured metadata and stratified splits. The MedJudge design is a sensible approach for open-ended QA, and the authors are transparent about the limitations of their human validation. Nevertheless, the current submission does not deliver its central benchmarking claim: the complete absence of results means the reader cannot assess whether the benchmark distinguishes models, whether MedJudge scores are stable, or whether the proposed prompting and fine-tuning methods have any measurable effect. As submitted, the paper is better described as a dataset resource than as a benchmark evaluation, and the abstract's claims about results are unsupported.

major comments (4)
  1. [5] Section 5 is empty, stating 'This section is currently under development and will be included in a future revision.' The abstract and Section 1 assert that extensive evaluations were performed and that the results highlight key challenges, but no model scores, tables, or comparisons appear in the main text or appendix; Figures 15-17 are only captions with no accompanying data. This is not a presentation issue: the central claim of the paper is benchmarking, and the results are the evidence for that claim. The manuscript cannot be accepted until the full Results section is included (baseline scores, prompt-method comparisons, fine-tuning comparisons, and MedJudge reliability analyses) and the abstract is made consistent with the content actually reported.
  2. [A.2] The human validation of MedJudge is too weak to support the claim of a clinically informed evaluation framework validated by expert reviews. The primary correctness dimension has 75% collapsed agreement with quadratic Cohen's kappa = 0.42 (95% CI 0.19-0.58), and the Clinical Impact dimension has kappa = 0.07, detecting only one-third of high-impact discrepancies. Since no downstream result shows that MedJudge separates strong from weak models or correlates with human judgments at scale, the evaluation framework's reliability is not established. Please report a confusion matrix, per-label precision and recall, and a sensitivity analysis of model rankings when the grader model or rubric is changed.
  3. [3.1] The gold standard is assumed to be the answers posted by verified physicians on four forums, and MedJudge is explicitly instructed to judge exclusively against these answers and ignore external knowledge (A.1). If the forum answers are incomplete or contain systematic errors, every model score inherits those errors. The paper provides no evidence that the gold answers are clinically acceptable. The authors should validate a random sample of gold answers with independent expert review, or construct references through consensus, and report the resulting agreement.
  4. [3.3] ICD-11 and question-type labels are generated by GPT-4o-mini without any reported accuracy or human agreement. These labels are load-bearing: the test split is stratified by ICD-11 category, and the role-based prompting in Section 4.3 conditions on them. Label noise could distort both the evaluation split and the prompt conditions. Please report annotation accuracy on a human-annotated sample and, ideally, a sensitivity analysis showing that key conclusions are robust to label noise.
minor comments (6)
  1. [3.3] The text says 'Figure 6 shows the distribution of ICD-11 categories,' but Figure 6 is titled 'Distribution of Question Type' and Figure 5 already shows the ICD-11 distribution; the cross-references and captions should be corrected.
  2. [Abstract] The abstract states PerMedCQA is 'the first Persian-language benchmark,' while Section 1 uses the more careful phrase 'To the best of our knowledge.' The abstract should either match that qualification or report a systematic search for prior Persian medical QA resources.
  3. [4.1] MedJudge is described in the contributions as a novel evaluation framework, but Section 4.1 states it is 'based on the criteria (Hosseini et al., 2024)'; please clarify which components are new to this work.
  4. [Figure 1] Figure 1 contains the typo 'DIFFRENTTECHNIQUES,' and Figure 13 uses 'MedJude' while the rest of the paper uses 'MedJudge'; these should be harmonized.
  5. [References] The reference for PaLM (Chowdhery et al., 2023) is malformed, containing an unrelated string of author names, and several URL-only entries (e.g., Claude3.5, GetZoop) lack access dates; the bibliography needs a careful cleanup.
  6. [3.2] The phrasing 'resulting the final numbers QA pairs' should be revised to 'resulting in the final number of QA pairs,' and in Section 3.3 'PerMedCQA were split' should be 'PerMedCQA was split.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the dataset, MedJudge rubric, and human validation are externally grounded; the missing Results section is an incompleteness issue, not an input–output loop.

full rationale

The paper's central resource claim is that PerMedCQA is a Persian consumer medical QA dataset built from public forum question–answer pairs and annotated with ICD-11 categories and question types. The evaluation chain uses MedJudge, an LLM-as-a-judge grader that compares model answers against the verified-physician expert answers using a rubric, and MedJudge is independently checked on a 100-item subset against board-certified physician ratings (75% collapsed agreement, quadratic Cohen's κ = 0.42, τ = 0.41, p < 0.001). This human validation is external to the evaluated models: the grader's reliability is not established by the models it later scores, and the human ratings do not depend on any fitted parameter from the benchmark. The gold answers themselves come from forum physicians rather than from the evaluated systems, so using them as the reference standard is a dataset-construction choice, not a self-referential derivation. The ICD-11 and question-type labels were produced by GPT-4o-mini and later used to condition role-based prompts, but these labels are data annotations rather than predictions of the benchmark outcome, and the final evaluation still runs through MedJudge against the expert answers rather than through the annotations. There are no equations in the paper that define the claimed result in terms of its inputs, and there is no fitted parameter that is renamed as a prediction. The most serious flaw in the manuscript is that Section 5 is explicitly marked “under development and will be included in a future revision,” so the abstract's claim that the results highlight key challenges is unsupported by any presented scores. That is a severe completeness and correctness risk, but it is not circularity. For the circularity dimension, the derivation chain is self-contained and externally grounded, so the appropriate score is 0.

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

The paper's central claims rest on the reliability of forum gold answers, the accuracy of LLM-based annotation, and the validity of MedJudge as a proxy for human clinical judgment. No free parameters are fitted to data.

assumptions (3)
  • domain assumption Forum expert answers are medically accurate gold references.
    Section 3.1 describes data sources as verified specialists, and MedJudge is instructed to evaluate strictly against these expert answers (A.1). If these answers are wrong or incomplete, every evaluation based on them is invalid.
  • domain assumption GPT-4o-mini PII detection and ICD-11 tagging are sufficiently accurate for the full dataset.
    Sections 3.2 and 3.3 use GPT-4o-mini without human validation on the full 68k instances; human validation is only reported for the 100-item MedJudge subset.
  • domain assumption The 25 question-type taxonomy (Abacha et al., 2019) and 28 ICD-11 categories are appropriate for Persian consumer questions.
    These taxonomies are borrowed from prior literature and applied without adaptation studies for Persian cultural and linguistic contexts.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PerMedCQA: Benchmarking Large Language Models on Medical Consumer Question Answering in Persian Language." pith.science (2026). https://pith.science/paper/ZNPNTZ5C

@misc{pith2026250518331,
  author       = {Pith},
  title        = {Pith review of: PerMedCQA: Benchmarking Large Language Models on Medical Consumer Question Answering in Persian Language},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZNPNTZ5C}},
  note         = {Machine review of arXiv:2505.18331}
}
read the original abstract

Medical consumer question answering (CQA) is crucial for empowering patients by providing personalized and reliable health information. Despite recent advances in large language models (LLMs) for medical QA, consumer-oriented and multilingual resources, particularly in low-resource languages like Persian, remain sparse. To bridge this gap, we present PerMedCQA, the first Persian-language benchmark for evaluating LLMs on real-world, consumer-generated medical questions. Curated from a large medical QA forum, PerMedCQA contains 68,138 question-answer pairs, refined through careful data cleaning from an initial set of 87,780 raw entries. We evaluate several state-of-the-art multilingual and instruction-tuned LLMs, utilizing MedJudge, a novel rubric-based evaluation framework driven by an LLM grader, validated against expert human annotators. Our results highlight key challenges in multilingual medical QA and provide valuable insights for developing more accurate and context-aware medical assistance systems. The data is publicly available on https://huggingface.co/datasets/NaghmehAI/PerMedCQA

Figures

Figures reproduced from arXiv: 2505.18331 by the authors.

Figure 3
Figure 3. Task instructions for ICD-11 classification and PII tagging. [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figure 4
Figure 4. Standardized ICD-11 classification codes used for QA annotation. [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Distribution of ICD-11 Categories in PerMedCQA [PITH_FULL_IMAGE:figures/full_fig_p015_5.png] view at source ↗
Figures from the paper (12 more)
Figure 6
Figure 6. Figure 6: Distribution of Question Type in PerMedCQA [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Prompt description for the Question Type Tagging Task. [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Baseline system prompt used for all models in the default setting. [PITH_FULL_IMAGE:figures/full_fig_p016_8.png]
Figure 9
Figure 9. Figure 9: Role-based prompting strategy using ICD-11-conditioned specialist titles (e.g., [PITH_FULL_IMAGE:figures/full_fig_p017_9.png]
Figure 10
Figure 10. Figure 10: Translation pipeline (separate LLM inference) for leveraging English LLMs with Persian input and [PITH_FULL_IMAGE:figures/full_fig_p017_10.png]
Figure 11
Figure 11. Figure 11: MedJudge system prompt with structured grading rubric for model evaluation. [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]
Figure 12
Figure 12. Figure 12: Template used for submitting evaluation tasks to MedJudge. Placeholders are replaced at runtime. [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: MedJude Structured Output Schema [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]
Figure 14
Figure 14. Figure 14: Human Evaluation 19 [PITH_FULL_IMAGE:figures/full_fig_p019_14.png]
Figure 15
Figure 15. Figure 15: Correctness for baselines and advanced Prompting [PITH_FULL_IMAGE:figures/full_fig_p020_15.png]
Figure 16
Figure 16. Figure 16: Coverage for baselines and advanced Prompting [PITH_FULL_IMAGE:figures/full_fig_p020_16.png]
Figure 17
Figure 17. Figure 17: Critical impact for baselines and advanced Prompting [PITH_FULL_IMAGE:figures/full_fig_p020_17.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 45 canonical work pages

  1. [1]

    Correct: Clinically equivalent; no mean- ingful differences

    work page

  2. [2]

    Partially_correct: Minor deviations, no significant clinical impact

    work page

  3. [3]

    InProceedings of the 23rd Workshop on Biomedical Natural Language Processing, pages 277–294

    K-qa: A real-world medical q&a benchmark. InProceedings of the 23rd Workshop on Biomedical Natural Language Processing, pages 277–294. Mavara-e-Teb. 2025.https://mavarateb.com. Xiangbin Meng, Xiangyu Yan, Kuo Zhang, Da Liu, Xi- aojuan Cui, Yaodong Yang, Muhan Zhang, Chunxia Cao, Jingjia Wang, Xuliang Wang, and 1 others. 2024. The application of large lang...

    arXiv 2025

  4. [4]

    arXiv preprint arXiv:2404.18416

    Capabilities of gemini models in medicine. arXiv preprint arXiv:2404.18416. Xiaoming Shi, Zeming Liu, Li Du, Yuxuan Wang, Hon- gru Wang, Yuhang Guo, Tong Ruan, Jie Xu, Xiaofan Zhang, and Shaoting Zhang. 2024. Medical dialogue system: A survey of categories, methods, evaluation and challenges.Findings of the Association for Com- putational Linguistics ACL ...

    arXiv 2024

  5. [5]

    (5A00–5D46) Endocrine, nutritional or metabolic dis- eases

    work page

  6. [6]

    (6A00–6E8Z) Mental, behavioural or neurodevelop- mental disorders

    work page

  7. [7]

    InPro- ceedings of the AAAI conference on artificial intelli- gence, volume 38, pages 19368–19376

    Zhongjing: Enhancing the chinese medical 10 capabilities of large language model through expert feedback and real-world multi-turn dialogue. InPro- ceedings of the AAAI conference on artificial intelli- gence, volume 38, pages 19368–19376. Guangtao Zeng, Wenmian Yang, Zeqian Ju, Yue Yang, Sicheng Wang, Ruisi Zhang, Meng Zhou, Jiaqi Zeng, Xiangyu Dong, Ruo...

    work page 2020

  8. [8]

    (8A00–8E7Z) Diseases of the nervous system

    work page

Show all 52 references
  1. [9]

    InProceedings of the 62nd Annual Meeting of the Association for Compu- tational Linguistics (Volume 3: System Demonstra- tions), Bangkok, Thailand

    Llamafactory: Unified efficient fine-tuning of 100+ language models. InProceedings of the 62nd Annual Meeting of the Association for Compu- tational Linguistics (Volume 3: System Demonstra- tions), Bangkok, Thailand. Association for Computa- tional Linguistics. A Appendix A.1 ...

  2. [10]

    (AA00–AC0Z) Diseases of the ear or mastoid process

  3. [11]

    (BA00–BE2Z) Diseases of the circulatory system

  4. [12]

    Incorrect: Substantial differences that affect accuracy or completeness

  5. [13]

    •Coverage:

    Contradictory: Model advice directly conflicts with the expert reference. •Coverage:

  6. [14]

    Equal: Both answers contain the same key facts

  7. [15]

    Model_subset: Model omits critical facts present in the expert answer

  8. [16]

    Expert_subset: Model introduces rele- vant facts not found in the expert answer

  9. [17]

    •Clinical_impact: 11

    Overlap_none: No substantial factual overlap. •Clinical_impact: 11

  10. [18]

    Negligible: No effect on care or under- standing

  11. [19]

    Moderate: Slight effect on treatment or comprehension

  12. [20]

    Significant: Likely to affect recommen- dations or outcomes

  13. [21]

    ac- ceptable

    Critical: May result in unsafe or harmful guidance. ReliabilityTo assess the reliability of Med- Judge, we compared its labels against blinded ratings from board-certified physicians on a 100- item subset. Agreement on the primary dimen- sion—correctness—was 75% (collapsed to ...

  14. [22]

    Return only the integer (1–28) that best corresponds to the core subject matter

    ICD-11 Classification Classify the content based on ICD-11 categories using the catalogue provided in the next page. Return only the integer (1–28) that best corresponds to the core subject matter

  15. [23]

    identity

    PII Detection Check if any personal information of the patient or expert is exposed in the messages. Personal information includes examples such as real name, address, phone number, email, etc. If such information exists, set "identity" to true; otherwise, set it tofalse. Figu...

  16. [24]

    (1A00–1H0Z) Certain infectious or parasitic diseases

  17. [25]

    (2A00–2F9Z) Neoplasms

  18. [26]

    (3A00–3C0Z) Diseases of the blood or blood-forming organs

  19. [27]

    (4A00–4B4Z) Diseases of the immune system

  20. [28]

    Med-Judge

    (XA0060–XY9U) Extension Codes 13 Table 3: Question Type Categories. Question Type Definition Example Information Asks for general identification or classifica- tion of a drug. What type of drug is am- phetamine? Dose Queries recommended or safe dosage. What is a daily amount o...

  21. [30]

    (7A00–7B2Z) Sleep-wake disorders

  22. [32]

    (9A00–9E1Z) Diseases of the visual system

  23. [35]

    (CA00–CB7Z) Diseases of the respiratory system

  24. [36]

    (DA00–DE2Z) Diseases of the digestive system

  25. [37]

    (EA00–EM0Z) Diseases of the skin

  26. [38]

    (FA00–FC0Z) Diseases of the musculoskeletal system or connective tissue

  27. [39]

    (GA00–GC8Z) Diseases of the genitourinary system

  28. [40]

    (HA00–HA8Z) Conditions related to sexual health

  29. [41]

    (JA00–JB6Z) Pregnancy, childbirth or the puerperium

  30. [42]

    (KA00–KD5Z) Certain conditions originating in the perinatal period

  31. [43]

    (LA00–LD9Z) Developmental anomalies

  32. [44]

    (MA00–MH2Y) Symptoms, signs or clinical findings, not elsewhere classified

  33. [45]

    (NA00–NF2Z) Injury, poisoning or other conse- quences of external causes

  34. [46]

    (PA00–PL2Z) External causes of morbidity or mortal- ity

  35. [47]

    (QA00–QF4Z) Factors influencing health status or con- tact with health services

  36. [48]

    (RA00–RA26) Codes for special purposes

  37. [49]

    (SA00–SJ3Z) Traditional Medicine Conditions – Mod- ule I

  38. [50]

    (V A00–VC50) Functioning assessment

  39. [52]

    2.Step-by-step and Explainable Evaluation: • Provide a brief and concise comparison analysis (brief_analysis), clearly explaining your evaluation

    Exclusivity:Base your judgmentonly on the Expert’s provided answer, and never rely on your own medical knowledge or external resources. 2.Step-by-step and Explainable Evaluation: • Provide a brief and concise comparison analysis (brief_analysis), clearly explaining your evalua...

  40. [2019]

    InProceedings of the 18th BioNLP Workshop and Shared Task, pages 250–260, Florence, Italy

    ChiMed: A Chinese medical corpus for ques- tion answering. InProceedings of the 18th BioNLP Workshop and Shared Task, pages 250–260, Florence, Italy. Association for Computational Linguistics. Augustin Toma, Patrick R Lawler, Jimmy Ba, Rahul G Krishnan, Barry B Rubin, and Bo W...

    2023 arXiv

  41. [2020]

    HiSalamat

    Measuring massive multitask language under- standing.arXiv preprint arXiv:2009.03300. HiSalamat. 2025.https://www.hisalamat.com. Pedram Hosseini, Jessica M Sin, Bing Ren, Bryce- ton G Thomas, Elnaz Nouri, Ali Farahanchi, and Saeed Hassanpour. 2024. A benchmark for long- form m...

    2009 arXiv

  42. [2023]

    Baize: An open-source chat model with parameter-efficient tuning on self-chat data.arXiv preprint arXiv:2304.01196. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Day- iheng Liu, Fan Zhou, Fe...

    2025 arXiv

  43. [2024]

    LLaMA4Scout

    Deepseek-v3 technical report.arXiv preprint arXiv:2412.19437. LLaMA4Scout. 2025. https://ai.meta.com/blog/ llama-4-multimodal-intelligence/. Jenish Maharjan, Anurag Garikipati, Navan Preet Singh, Leo Cyrus, Mayank Sharma, Madalina Ciobanu, Gina Barnes, Rahul Thapa, Qingqing Ma...

    2025 arXiv

  44. [2025]

    InProceedings of the AAAI Conference on Artificial Intelligence, vol- ume 39, pages 26039–26047

    Carebot: A pioneering full-process open- source medical language model. InProceedings of the AAAI Conference on Artificial Intelligence, vol- ume 39, pages 26039–26047. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, D...

    2023 arXiv

Pith tools

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