Elamur Research⌕
☰
home / papers / 2026 / TrustTable: A Neuro-Symbolic Auditing Framework for Faithful Table QA

TrustTable: A Neuro-Symbolic Auditing Framework for Faithful Table QA

Paper · 2026 · Guangzhen Zhao, Dechang Kong, Tongyu Wu et al.

Summary

LLM-based table question answering systems sometimes reach the right answer through faulty reasoning, such as misread columns or wrong arithmetic that happens to land on the gold value. The authors estimate this affects about 5% of correct predictions. Benchmarks that score only the final answer miss this problem. TrustTable addresses it by auditing the reasoning itself rather than relying on another LLM as judge. It treats a Chain-of-Thought as an implicit symbolic program and checks it in two independent ways. A Pandas-based FactChecker runs generated code on the table to confirm that the cited values exist. A Z3-based LogicAuditor tests whether the reasoning steps are logically satisfiable. Rejected chains go back to the generator with precise error diagnostics, forming a refinement loop that needs no human labels. The authors also release TrustTable-Bench, a diagnostic set of deliberately faithful and flawed reasoning chains. On WikiTableQuestions, TrustTable matches strong self-consistency baselines while using fewer API calls. Its verdicts also agree with human experts more closely than purely neural verifiers do. This matters for high-stakes domains such as finance and medicine.

Key points

  • TrustTable-Bench: seeds come from WTQ, PubHealthTab and SciAutomic, and GPT-5 generates the reasoning chains. Each chain falls into one of four types: Faithful, Spurious (correct answer from wrong logic), Fallacious, or Inconsistent. Human checks found 96.5% of items valid, with Fleiss' κ = 0.88.
  • Efficiency on WTQ: accuracy rises from 80.1% (Greedy CoT) to 83.4% at an average of 5.80 calls per query. Self-Consistency needs k=10 (10 calls) to reach 83.2%, so TrustTable uses about 42% fewer calls.
  • Agreement with experts: on 200 reasoning chains judged by three experts, TrustTable scores Cohen's κ = 0.86 and 93.5% agreement. The comparison verifiers score lower:
    • VeriCoT: κ = 0.79
    • LLM Self-Check (DeepSeek-V3): κ = 0.74
  • Value of structured feedback: with symbolic error messages, 48.4% of rejected chains are successfully corrected, compared with 34.1% for blind resampling.
  • Limitations: the framework handles only deterministic logic, not "soft" reasoning such as the qualitative claims in SCITAB. It works on flattened 2D DataFrames, so tables with hierarchical or merged headers can cause problems.

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

 Institutions

🎓Nanjing University of Posts and TelecommunicationsNanjing · China4 works · 22 citations

 Fields, topics and concepts

📚 Shelves 4The 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 works

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

Cite

Cite as Guangzhen Zhao, Dechang Kong, Tongyu Wu and Zhenjiang Dong. 2026. TrustTable: A Neuro-Symbolic Auditing Framework for Faithful Table QA. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1612

@inproceedings{zhao2026trusttable,
  title = {{TrustTable: A Neuro-Symbolic Auditing Framework for Faithful Table QA}},
  author = {Guangzhen Zhao, Dechang Kong, Tongyu Wu and Zhenjiang Dong},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1612},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1612}
}

References

32 references, 18 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Nikhil Abhyankar, Vivek Gupta, Dan Roth, and Chandan K Reddy. 2025. H-star: Llm-driven hybrid sql-text adaptive reasoning on tables. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL 2025)… find
  2. Mubashara Akhtar, Oana Cocarascu, and Elena Simperl. 2022. Pubhealthtab: A public health table-based dataset for evidence-based fact checking. In Findings of the Association for Computational Linguistics: NAACL 2022 (Findings of NAACL 2022), pages 1–16, Seattle, United States. find
  3. Wenhu Chen. 2023. Large language models are few (1)-shot table reasoners. In Findings of the association for computational linguistics: EACL 2023 (Findings of EACL 2023), pages 1120–1130, Dubrovnik, Croatia.
  4. Wenhu Chen, Xueguang Ma, Xinyi Wang, and William W. Cohen. 2023. Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks. Transactions on Machine Learning Research. find
  5. Zhiyu Chen, Wenhu Chen, Charese Smiley, Sameena Shah, Iana Borova, Dylan Langdon, Reema Moussa, Matt Beane, Ting-Hao Huang, Bryan Routledge, and William Yang Wang. 2021. FinQA: A dataset of numerical reasoning over financial data. In Proceedings of the 2021 Conference on Empirical Methods in Natural…
  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. Thomas M. Cover and Joy A. Thomas. 2006. Elements of Information Theory, 2nd edition. John Wiley & Sons, Hoboken, NJ. find
  8. Leonardo de Moura and Nikolaj Bjørner. 2008. Z3: An efficient SMT solver. In Proceedings of the 14th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS 2008), pages 337–340, Budapest, Hungary. find
  9. Nouha Dziri, Ximing Lu, Melanie Sclar, Xiang Lorraine Li, Liwei Jiang, Bill Yuchen Lin, Peter West, Chandra Bhagavatula, Ronan Le Bras, Jena D. Hwang, Soumya Sanyal, Sean Welleck, Xiang Ren, Allyson Ettinger, Zaid Harchaoui, and Yejin Choi. 2023. Faith and fate: limits of transformers on composition…
  10. Yu Feng, Nathaniel Weir, Kaj Bostrom, Sam Bayless, Darion Cassel, Sapana Chaudhary, Benjamin Kiesl-Reiter, and Huzefa Rangwala. 2025. Vericot: Neuro-symbolic chain-of-thought validation via logical consistency checks. arXiv preprint arXiv:2511.04662. — [arXiv:2511.04662](https://arxiv.org/abs/2511.04662) find
  11. Joseph L. Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological Bulletin, 76(5):378–382. find
  12. Heng Gong, Yawei Sun, Xiaocheng Feng, Bing Qin, Wei Bi, Xiaojiang Liu, and Ting Liu. 2020. TableGPT: Few-shot table-to-text generation with table structure reconstruction and content matching. In Proceedings of the 28th International Conference on Computational Linguistics (COLING 2020), pages 1978–… find
  13. Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. Large language models are zero-shot reasoners. In Proceedings of the 36th International Conference on Neural Information Processing Systems (NeurIPS 2022), New Orleans, LA, USA.
  14. Zhenjie Sun, Naihao Deng, Haofei Yu, and Jiaxuan You. 2025. Tables as thought: Exploring structured thoughts in LLM reasoning. In Proceedings of the 4th Table Representation Learning Workshop, pages 19–33, Vienna, Austria. find
  15. Tamera Lanham, Anna Chen, Ansh Radhakrishnan, Benoit Steiner, Carson Denison, Danny Hernandez, Dustin Li, Esin Durmus, Evan Hubinger, Jackson Kernion, Kamilė Lukošiūtė, Karina Nguyen, Newton Cheng, Nicholas Joseph, Nicholas Schiefer, Oliver Rausch, Robin Larson, Sam McCandlish, Sandipan Kundu, an… — Measuring Faithfulness in Chain-of-Thought Reasoning
  16. Hunter Lightman, Vineet Kosaraju, Yuri Burda, Harrison Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2024. Let’s verify step by step. In The Twelfth International Conference on Learning Representations (ICLR 2024), Vienna, Austria.
  17. Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, and Jian-Guang Lou. 2022. TAPEX: Table pre-training via learning a neural SQL executor. In International Conference on Learning Representations (ICLR 2022), Virtual.
  18. 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 (NAACL 2024), pages 450–482, Mexico City,… find
  19. Miles Turpin, Julian Michael, Ethan Perez, and Samuel R. Bowman. 2023. Language models don’t always say what they think: Unfaithful explanations in chain-of-thought prompting. In Proceedings of the 37th International Conference on Neural Information Processing Systems (NeurIPS 2023), pages 74952–749…
  20. Jonathan Uesato, Nate Kushman, Ramana Kumar, Francis Song, Noah Siegel, Lisa Wang, Antonia Creswell, Geoffrey Irving, and Irina Higgins. 2022. Solving math word problems with process-and outcome-based feedback. arXiv preprint arXiv:2211.14275.
  21. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is All You Need. In Advances in Neural Information Processing Systems (NIPS 2017), pages 5998–6008, Long Beach, CA, USA.
  22. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V. Le, Ed H. Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2023. Self-consistency improves chain of thought reasoning in language models. In The Eleventh International Conference on Learning Representations (ICLR 2023), Kigali, Rwanda.
  23. William Merrill and Ashish Sabharwal. 2024. The expressive power of transformers with chain of thought. In The Twelfth International Conference on Learning Representations (ICLR 2024), Vienna, Austria. find
  24. 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…
  25. 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 (NAACL 2024), pages 5725–… find
  26. Jason Wei, Yi Tay, Rishi Bommasani, Colin Raffel, Barret Zoph, Sebastian Borgeaud, Dani Yogatama, Maarten Bosma, Denny Zhou, Donald Metzler, Ed H. Chi, Tatsunori Hashimoto, Oriol Vinyals, Percy Liang, Jeff Dean, and William Fedus. 2022a. Emergent abilities of large language models. Transactions on M…
  27. Panupong Pasupat and Percy Liang. 2015. Compositional semantic parsing on semi-structured tables. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (ACL 2015), pages 1470–1480, Beijing,…
  28. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc V. Le, and Denny Zhou. 2022b. Chain-of-thought prompting elicits reasoning in large language models. In Proceedings of the 36th International Conference on Neural Information Processing Systems (NeurIPS 2022)…
  29. Xin Quan, Marco Valentino, Louise Dennis, and Andre Freitas. 2024. Enhancing ethical explanations of large language models through iterative symbolic refinement. In Proceedings of the 18th Conference Tianbao Xie, Chen Henry Wu, Peng Shi, Ruiqi Zhong, Torsten Scholak, Michihiro Yasunaga, Chien-Sheng… find
  30. Tianshu Zhang, Xiang Yue, Yifei Li, and Huan Sun. 2024. Tablellama: Towards open large generalist models for tables. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL 2024), pages 6024– 6044.
  31. Yuji Zhang, Qingyun Wang, Cheng Qian, Jiateng Liu, Chenkai Sun, Denghui Zhang, Tarek Abdelzaher, Chengxiang Zhai, Preslav Nakov, and Heng Ji. 2025. Atomic reasoning for scientific table claim verification. arXiv preprint arXiv:2506.06972. — [arXiv:2506.06972](https://arxiv.org/abs/2506.06972) find
  32. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Zhipeng Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, and 3 others. 2023. A survey of large language models. Prepri…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhao-2026-trusttable-neuro-symbolic-auditing-framework/