Elamur Research⌕
☰
home / papers / 2026 / SMART: Semantic Header Flattening and Pseudo-Code-Style Reasoning for LLM-based Complex Table Question Answering

SMART: Semantic Header Flattening and Pseudo-Code-Style Reasoning for LLM-based Complex Table Question Answering

Paper · 2026 · Yongshuo Zhang, Jixiong Chen, Kaihe xu et al.

Summary

Real-world tables often have multi-level, merged headers built for human readers. LLMs struggle to link conditions, attributes and values in such tables. Earlier methods encode the hierarchy as tuples, HTML or cell graphs, which leaves the model to infer header relations on its own. They also rely on CoT (unverifiable), PoT (fragile code) or costly hybrids. SMART splits structure understanding from reasoning in three stages:

  1. Semantic Header Flattening: an LLM rewrites each hierarchical header path as one natural-language label. Headers spanning the whole table become part of the caption.
  2. Global Understanding: the model writes a short guide covering the table overview, header roles, question analysis and expected answer type.
  3. Pseudo-Code-Style Reasoning: the model chains a small set of atomic operations (SELECT, SORT, INFER, COMPUTE, ANSWER). An external calculator and sorter check each step and trigger a revision when they find errors.

The authors report state-of-the-art accuracy on HiTab, AIT-QA and WikiTableQA. SMART also needs far fewer reasoning iterations than competing methods. This makes it a practical, interpretable option for complex table QA.

Key points

  • Evaluated on HiTab, AIT-QA and WikiTableQA. Baselines include TableParser, E5, GraphOTTER, MIX-SC, TIDE and Chain-of-Table.
  • Efficiency on HiTab with Qwen2: SMART averages about 1.03 iterations. The baselines need roughly 3.2 to 6.7, with GraphOTTER at 3.2 and E5 at 6.7.
  • A correction round runs only when verification flags a numeric or sorting mistake. The iteration limit is 3.
  • Stated limitations: verification covers only numeric and sorting errors, not logical or textual ones. There is also no filter for irrelevant table content.
  • Published in Findings of ACL 2026 by authors from HUST together with Ping An and Brilliance Technology.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksComputer Vision and Pattern Recognition · 3,486 works

🧭 Topics 2Research subjects and languages the work is about.Reasoning · 1,685 worksQuestion answering · 487 works

Cite

Cite as Yongshuo Zhang, Jixiong Chen, Kaihe xu, Wei Wei and Shihao Zou. 2026. SMART: Semantic Header Flattening and Pseudo-Code-Style Reasoning for LLM-based Complex Table Question Answering. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.1474

@inproceedings{zhang2026smart,
  title = {{SMART: Semantic Header Flattening and Pseudo-Code-Style Reasoning for LLM-based Complex Table Question Answering}},
  author = {Yongshuo Zhang, Jixiong Chen, Kaihe xu, Wei Wei and Shihao Zou},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1474},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1474}
}

References

27 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774.
  2. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, and 1 others. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  3. Wenhu Chen. 2022. Large language models are few (1)-shot table reasoners. arXiv preprint arXiv:2210.06710.
  4. Wenhu Chen, Xueguang Ma, Xinyi Wang, and William W Cohen. 2022. Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks. arXiv preprint arXiv:2211.12588. find
  5. Zhoujun Cheng, Haoyu Dong, Zhiruo Wang, Ran Jia, Jiaqi Guo, Yan Gao, Shi Han, Jian-Guang Lou, and Dongmei Zhang. 2022. HiTab: A hierarchical table dataset for question answering and natural language generation. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistic…
  6. Zhoujun Cheng, Tianbao Xie, Peng Shi, Chengzu Li, Rahul Nadkarni, Yushi Hu, Caiming Xiong, Dragomir Radev, Mari Ostendorf, Luke Zettlemoyer, Noah A. Smith, and Tao Yu. 2023. Binding language models in symbolic languages. In The Eleventh International Conference on Learning Representations, ICLR 2023…
  7. Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Yunxuan Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, and 1 others. 2024. Scaling instruction-finetuned language models. Journal of Machine Learning Research, 25(70):1–53. find
  8. Yannis Katsis, Saneem Chemmengath, Vishwajeet Kumar, Samarth Bharadwaj, Mustafa Canim, Michael Glass, Alfio Gliozzo, Feifei Pan, Jaydeep Sen, Karthik Sankaranarayanan, and 1 others. 2021. Ait-qa: Question answering dataset over complex tables in the airline industry. arXiv preprint arXiv:2106.12944. find
  9. Qianlong Li, Chen Huang, Shuai Li, Yuanxin Xiang, Deng Xiong, and Wenqiang Lei. 2025. Graphotter: Evolving llm-based graph reasoning for complex table question answering. In Proceedings of the 31st International Conference on Computational Linguistics, pages 5486–5506. find
  10. Tianyang Liu, Fei Wang, and Muhao Chen. 2024. Rethinking tabular data understanding with large language models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), NAACL 2024, Mexic… find
  11. Qingyang Mao, Qi Liu, Zhi Li, Mingyue Cheng, Zheng Zhang, and Rui Li. 2024. Potable: Programming standardly on table-based reasoning like a human analyst. CoRR, abs/2412.04272. — [arXiv:2412.04272](https://arxiv.org/abs/2412.04272) find
  12. Md Mahadi Hasan Nahid and Davood Rafiei. 2024. Tab-SQLify: Enhancing reasoning capabilities of LLMs through table decomposition. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers),… find
  13. Panupong Pasupat and Percy Liang. 2015. Compositional semantic parsing on semi-structured tables. arXiv preprint arXiv:1508.00305.
  14. Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton-Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, and 6 o… — Code Llama: Open Foundation Models for Code
  15. Pranab Sahoo, Ayush Kumar Singh, Sriparna Saha, Vinija Jain, Samrat Mondal, and Aman Chadha. 2024. A systematic survey of prompt engineering in large language models: Techniques and applications. arXiv preprint arXiv:2402.07927.
  16. Zhiruo Wang, Haoyu Dong, Ran Jia, Jia Li, Zhiyi Fu, Shi Han, and Dongmei Zhang. 2021. Tuta: Tree-based transformers for generally structured table pre-training. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 1780–1790. find
  17. Zilong Wang, Hao Zhang, Chun-Liang Li, Julian Martin Eisenschlos, Vincent Perot, Zifeng Wang, Lesly Miculicich, Yasuhisa Fujii, Jingbo Shang, Chen-Yu Lee, and Tomas Pfister. 2024. Chain-of-table: Evolving tables in the reasoning chain for table understanding. In The Twelfth International Conference…
  18. August 1, 2025, pages 17432–17451. Association for Computational Linguistics. find
  19. Yadong Zhang, Shaoguang Mao, Tao Ge, Xun Wang, Adrian de Wynter, Yan Xia, Wenshan Wu, Ting Song, Man Lan, and Furu Wei. 2024a. Llm as a mastermind: A survey of strategic reasoning with large language models. arXiv preprint arXiv:2404.01230. find
  20. Yunjia Zhang, Jordan Henkel, Avrilia Floratou, Joyce Cahoon, Shaleen Deep, and Jignesh M. Patel. 2024b. Reactable: Enhancing react for table question answering. Proc. VLDB Endow., 17(8):1981–1994. find
  21. Zhehao Zhang, Yan Gao, and Jian-Guang Lou. 2024c. e5: Zero-shot hierarchical table analysis using augmented llms via explain, extract, execute, exhibit and extrapolate. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Languag… find
  22. Bowen Zhao, Changkai Ji, Yuejie Zhang, Wen He, Yingwen Wang, Qing Wang, Rui Feng, and Xiaobo Zhang. 2023. Large language models are complex table parsers. arXiv preprint arXiv:2312.11521. find
  23. Mingyu Zheng, Yang Hao, Wenbin Jiang, Zheng Lin, Yajuan Lyu, QiaoQiao She, and Weiping Wang. 2023. IM-TQA: A Chinese table question answering dataset with implicit and multi-type table structures. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: L… find
  24. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, and 1 others. 2022. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824– 24837.
  25. Zhen Yang, Ziwei Du, Minghan Zhang, Wei Du, Jie Chen, Zhen Duan, and Shu Zhao. 2025. Triples as the key: Structuring makes decomposition and verification easier in llm-based tableqa. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025. Open… find
  26. Yunhu Ye, Binyuan Hui, Min Yang, Binhua Li, Fei Huang, and Yongbin Li. 2023. Large language models are versatile decomposers: Decomposing evidence and questions for table-based reasoning. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retriev… find
  27. Peiying Yu, Guoxin Chen, and Jingjing Wang. 2025. Table-critic: A multi-agent framework for collaborative criticism and refinement in table reasoning. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienna, Austria, July 2…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhang-2026-smart-semantic-header-flattening-pseudo/