AcupunctureBench: A Multimodal Benchmark for Evaluating Generative Models in the Acupuncture Domain

Abstract

As large language models (LLMs) are increasingly applied in the field of traditional Chinese medicine (TCM), standardized benchmark evaluations for TCM have gradually emerged. Acupuncture is a crucial component of TCM. However, existing evaluation benchmarks are lacking in this domain. To address this gap, this paper proposes AcupunctureBench, the first independently constructed large-scale multimodal benchmark for the acupuncture domain, covering text, image, and video modalities. The AcupunctureBench spans multidimen sional knowledge in acupuncture, including acupuncture theory, meridians and acupoints, clinical reasoning, TCM syndromes, and practical acupoint operation. To achieve large-scale construction while ensuring medical professionalism, this work designs a human-machine collaborative data construction pipeline and proposes efficient deduplication strategies to reduce redundant samples, while extensively adopting data synthesis to mitigate data leak age risks. AcupunctureBench contains a total of 58,526 question samples and 1,216 video samples, covering multiple task types including objective questions, subjective question answering, case analysis, image understanding, image reasoning, and video understanding. Based on this benchmark, this paper evaluates 14 large language models, 14 large multimodal models, and 4 video foundation models, and further releases the first dedicated evaluation website for the acupuncture domain.

Evaluating Acupuncture for Chemotherapy-Induced Peripheral Neuropathic Pain: A Narrative Review of Clinical Evidence and Mechanistic Perspectives
Paula Sousa, Madalena Deus, Mariana Pedrosa +5
Healthcare18 Sep 2026Peer-reviewedCC BY 4.0Full textGeneral biology General medicineHAL
Sino-US-DrugQA: A benchmark for evaluating large language models in cross-jurisdictional pharmaceutical regulation
Xuejing Fu, Zhen Chen, Wentao Lu
PLOS One15 Sep 2026Peer-reviewedCC BY 4.0Full textNeuroscience Systems & synthetic biologyPLOS
Towards AI-Driven Nanomedicine Discovery: A Benchmark and Multimodal Learning Framework for Nano Self-Assembly Prediction
Quan Hao, Mengyue Fan, Zifan Dong +9
arXiv3 Sep 2026PreprintarXiv licenceFull textBioinformatics & computational biology StatisticsarXiv
Who Called? V33DA: A Physically Verified Multimodal Benchmark for Vocal Attribution in Zebra Finch Groups
Maris Basha, Yuhang Wang, Xiaoran Chen +5
arXiv28 Oct 2026PreprintCC BY 4.0Full textBioinformatics & computational biologyarXiv
Benchmark indicators for evaluating sea turtle rehabilitation outcomes at a wildlife rehabilitation center on Tenerife island, Spain (1982–2025)
Felipe Trujillo-Camejo, Natalia Montesdeoca, Sara Capelli +2
Frontiers in veterinary science19 Oct 2026Peer-reviewedCC BYFull textEcologyEurope PMCEurope PMCPMC

Cite

BibTeX RIS

BibTeX

@misc{li2026acupuncturebench,
  title = {{AcupunctureBench: A Multimodal Benchmark for Evaluating Generative Models in the Acupuncture Domain}},
  author = {Shusheng Li and Yifei Qu and Zichen Han and Xingye Zhang and Zhongxin Wei and Yicheng Qiao and Jiaxin Shen and Changshuai Zhang and Siyuan Guo and Haiyan Ren and Xingfang Pan and Yangyang Liu and Zhaoshui He and Shuguo Hu and Wenjun Tan},
  howpublished = {arXiv},
  year = {2026},
  month = nov,
  doi = {10.48550/arxiv.2611.05399},
  eprint = {2611.05399},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.05399},
}

RIS

TY  - UNPB
TI  - AcupunctureBench: A Multimodal Benchmark for Evaluating Generative Models in the Acupuncture Domain
AU  - Shusheng Li
AU  - Yifei Qu
AU  - Zichen Han
AU  - Xingye Zhang
AU  - Zhongxin Wei
AU  - Yicheng Qiao
AU  - Jiaxin Shen
AU  - Changshuai Zhang
AU  - Siyuan Guo
AU  - Haiyan Ren
AU  - Xingfang Pan
AU  - Yangyang Liu
AU  - Zhaoshui He
AU  - Shuguo Hu
AU  - Wenjun Tan
PY  - 2026
DA  - 2026/11/01
PB  - arXiv
DO  - 10.48550/arxiv.2611.05399
UR  - https://arxiv.org/abs/2611.05399
AB  - As large language models (LLMs) are increasingly applied in the field of traditional Chinese medicine (TCM), standardized benchmark evaluations for TCM have gradually emerged. Acupuncture is a crucial component of TCM. However, existing evaluation benchmarks are lacking in this domain. To address this gap, this paper proposes AcupunctureBench, the first independently constructed large-scale multimodal benchmark for the acupuncture domain, covering text, image, and video modalities. The AcupunctureBench spans multidimen sional knowledge in acupuncture, including acupuncture theory, meridians and acupoints, clinical reasoning, TCM syndromes, and practical acupoint operation. To achieve large-scale construction while ensuring medical professionalism, this work designs a human-machine collaborative data construction pipeline and proposes efficient deduplication strategies to reduce redundant samples, while extensively adopting data synthesis to mitigate data leak age risks. AcupunctureBench contains a total of 58,526 question samples and 1,216 video samples, covering multiple task types including objective questions, subjective question answering, case analysis, image understanding, image reasoning, and video understanding. Based on this benchmark, this paper evaluates 14 large language models, 14 large multimodal models, and 4 video foundation models, and further releases the first dedicated evaluation website for the acupuncture domain.
ER  -

Abstract

Related

Cite