Elamur Research⌕
☰
home / papers / 2025 / Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table Reasoning

Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table Reasoning

Paper · 2025 · Peiying Yu, Guoxin Chen and Jingjing Wang

Summary

Large language models often fail at multi-step reasoning over tables. A wrong choice early in the chain, such as picking the wrong row or column, tends to carry through to the final answer. Earlier decomposition methods like Binder, Dater and Chain-of-Table have no reliable way to catch and fix these intermediate mistakes, and plain self-reflection can make things worse. The model may defend its errors or attack steps that were correct. Table-Critic tackles this with four cooperating LLM agents:

  • a Judge that checks each step and classifies the error type,
  • a Critic that finds the first faulty step and writes a targeted critique,
  • a Refiner that rewrites the reasoning, and
  • a Curator that turns experience into reusable critique templates.

The templates sit in a self-evolving tree organised by error type. The Judge navigates this tree to pick guidance for the Critic. The loop runs over several rounds until the Judge accepts the solution. Starting from Chain-of-Table outputs, the framework clearly improves accuracy on WikiTQ and TabFact. It also beats majority voting at comparable or greater compute. This suggests that structured critique is a better use of extra inference budget than simply sampling more answers.

Key points

  • Results: Table-Critic reaches 77.2% accuracy on WikiTQ and 92.6% on TabFact.
  • Comparison with majority voting: Self-consistency over Chain-of-Table (temperature 0.8) reaches only 70.0% on WikiTQ and 90.1% on TabFact, even with 15 sampled solutions.
  • Cost: Table-Critic uses roughly 1.8–2.2× the compute of basic Chain-of-Table.
  • Ablation: Freezing the template tree at its two initial templates lowers accuracy to 76.1% on WikiTQ (−1.1) and 90.8% on TabFact (−1.8).
  • Design and availability: Initial reasoning chains come from Chain-of-Table, but the authors say the framework can work with other methods. Code is released at github.com/Peiying-Yu/Table-Critic.

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

🏛️Chinese Academy of SciencesBeijing · China194 works · 23,906 citations🎓Soochow UniversitySuzhou · China63 works · 4,507 citations🔬Institute of Computing TechnologyBeijing · China51 works · 3,411 citations🎓Soochow UniversityTaipei · Taiwan2 works · 26 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksComputer Vision and Pattern Recognition · 2,869 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 1,426 works

Cite

Cite as Peiying Yu, Guoxin Chen and Jingjing Wang. 2025. Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table Reasoning. Proceedings of the 63rd 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/v1/2025.acl-long.853

@inproceedings{yu2025table,
  title = {{Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table Reasoning}},
  author = {Peiying Yu, Guoxin Chen and Jingjing Wang},
  booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2025},
  doi = {10.18653/v1/2025.acl-long.853},
  url = {https://doi.org/10.18653/v1/2025.acl-long.853}
}

References

34 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Anthropic. 2024. Introducing computer use, a new claude 3.5 sonnet, and claude 3.5 haiku. Anthropic Blog. find
  2. Guoxin Chen, Minpeng Liao, Chengxi Li, and Kai Fan. 2024a. Alphamath almost zero: Process supervision without process. In Advances in Neural Information Processing Systems, volume 37, pages 27689–27724. Curran Associates, Inc.
  3. Guoxin Chen, Minpeng Liao, Chengxi Li, and Kai Fan. 2024b. Step-level value preference optimization for mathematical reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 7889–7903, Miami, Florida, USA. Association for Computational Linguistics. find
  4. Guoxin Chen, Minpeng Liao, Peiying Yu, Dingmin Wang, Zile Qiao, Chao Yang, Xin Zhao, and Kai Fan. 2025a. C-3PO: Compact plug-and-play proxy optimization to achieve human-like retrieval-augmented generation. In Forty-second International Conference on Machine Learning. find
  5. Guoxin Chen, Zhong Zhang, Xin Cong, Fangda Guo, Yesai Wu, Yankai Lin, Wenzheng Feng, and Yasheng Wang. 2025b. Learning evolving tools for large language models. In The Thirteenth International Conference on Learning Representations. find
  6. Si-An Chen, Lesly Miculicich, Julian Eisenschlos, Zifeng Wang, Zilong Wang, Yanfei Chen, Yasuhisa Fujii, Hsuan-Tien Lin, Chen-Yu Lee, and Tomas Pfister. 2024c. Tablerag: Million-token table understanding with language models. In Advances in Neural Information Processing Systems 38: Annual Conference… find
  7. Wenhu Chen, Hongmin Wang, Jianshu Chen, Yunkai Zhang, Hong Wang, Shiyang Li, Xiyou Zhou, and William Yang Wang. 2020. Tabfact: A large-scale dataset for table-based fact verification. In 8th International Conference on Learning Representations, ICLR 2020, Addis Ababa, Ethiopia, April 26-30, 2020. Op…
  8. Zhoujun Cheng, Tianbao Xie, Peng Shi, Chengzu Li, Rahul Nadkarni, Yushi Hu, Caiming Xiong, Dragomir Radev, Mari Ostendorf, Luke Zettlemoyer, et al. 2022. Binding language models in symbolic languages. arXiv preprint arXiv:2210.02875.
  9. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, Arun Rao, Aston Zhang, Aurélien Rodriguez,… — The Llama 3 Herd of Models
  10. Andrew Codispoti, Andrew Galu, Andrew Kondrich, Andrew Tulloch, Andrey Mishchenko, Angela Baek, Angela Jiang, Antoine Pelisse, Antonia Woodford, Anuj Gosalia, Arka Dhar, Ashley Pantuliano, Avi Nayak, Avital Oliver, Barret Zoph, Behrooz Ghorbani, Ben Leimberger, Ben Rossen, Ben Sokolowsky, Ben Wang,… — GPT-4o System Card
  11. Zihui Gu, Ju Fan, Nan Tang, Preslav Nakov, Xiaoman Zhao, and Xiaoyong Du. 2022. PASTA: table-operations aware fact verification via sentence-table cloze pre-training. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022, Abu Dhabi, United Arab Emirate… find
  12. Shuhao Guan and Derek Greene. 2024. Advancing post-OCR correction: A comparative study of synthetic data. In Findings of the Association for Computational Linguistics: ACL 2024, pages 6036– 6047, Bangkok, Thailand. Association for Computational Linguistics. find
  13. Shuhao Guan, Moule Lin, Cheng Xu, Xinyi Liu, Jinman Zhao, Jiexin Fan, Qi Xu, and Derek Greene. 2025. Prep-OCR: A complete pipeline for document image restoration and enhanced OCR accuracy. In The 63rd Annual Meeting of the Association for Computational Linguistics. find
  14. Shuhao Guan, Cheng Xu, Moule Lin, and Derek Greene. 2024. Effective synthetic data and test-time adaptation for OCR correction. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 15412–15425, Miami, Florida, USA. Association for Computational Linguistics… find
  15. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948.
  16. Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V. Chawla, Olaf Wiest, and Xiangliang Zhang. 2024. Large language model based multi-agents: A survey of progress and challenges. In Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, IJCAI…
  17. Aaron Hurst, Adam Lerer, Adam P. Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, Aleksander Madry, Alex Baker-Whitcomb, Alex Beutel, Alex Borzunov, Alex Carney, Alex Chow, Alex Kirillov, Alex Nichol, Alex Paino, Alex Renzin, Alex Tachard Pass… — TAPEX: Table Pre-training via Learning a Neural SQL Executor
  18. Aman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan, Luyu Gao, Sarah Wiegreffe, Uri Alon, Nouha Dziri, Shrimai Prabhumoye, Yiming Yang, Shashank Gupta, Bodhisattwa Prasad Majumder, Katherine Hermann, Sean Welleck, Amir Yazdanbakhsh, and Peter Clark. 2023. Self-refine: Iterative refinement with… find
  19. Thomas Mesnard, Cassidy Hardin, Robert Dadashi, Surya Bhupatiraju, Shreya Pathak, Laurent Sifre, Morgane Rivière, Mihir Sanjay Kale, Juliette Love, Pouya Tafti, Léonard Hussenot, Aakanksha Chowdhery, Adam Roberts, Aditya Barua, Alex Botev, Alex Castro-Ros, Ambrose Slone, Amélie Héliou, Andrea Tacche… find
  20. Panupong Pasupat and Percy Liang. 2015. Compositional semantic parsing on semi-structured tables. arXiv preprint arXiv:1508.00305.
  21. Aske Plaat, Annie Wong, Suzan Verberne, Joost Broekens, Niki van Stein, and Thomas Back. 2024. find
  22. Reasoning with large language models, a survey. arXiv preprint arXiv:2407.11511. find
  23. 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.
  24. Zilong Wang, Hao Zhang, Chun-Liang Li, Julian Martin Eisenschlos, Vincent Perot, Zifeng Wang, Lesly Miculicich, Yasuhisa Fujii, Jingbo Shang, Chen-Yu Lee, et al. 2024. Chain-of-table: Evolving tables in the reasoning chain for table understanding. Proceedings of the 12th International Conference on… find
  25. Yilun Zhao, Lyuhao Chen, Arman Cohan, and Chen Zhao. 2024. Tapera: Enhancing faithfulness and interpretability in long-form table QA by content planning and execution-based reasoning. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),… find
  26. Xin Zheng, Jie Lou, Boxi Cao, Xueru Wen, Yuqiu Ji, Hongyu Lin, Yaojie Lu, Xianpei Han, Debing Zhang, and Le Sun. 2024. Critic-cot: Boosting the reasoning abilities of large language model via chain-of-thoughts critic. CoRR, abs/2408.16326. — [arXiv:2408.16326](https://arxiv.org/abs/2408.16326) find
  27. Zhenhua Xu, Zhebo Wang, Maike Li, Wenpeng Xing, Chunqiang Hu, Chen Zhi, and Meng Han. 2025. Rap-sm: Robust adversarial prompt via shadow models for copyright verification of large language models. arXiv preprint arXiv:2505.06304. — [arXiv:2505.06304](https://arxiv.org/abs/2505.06304) find
  28. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, et al. 2024a. Qwen2. 5 technical report. arXiv preprint arXiv:2412.15115. find
  29. Zhe Yang, Yichang Zhang, Yudong Wang, Ziyao Xu, Junyang Lin, and Zhifang Sui. 2024b. Confidence v.s. critique: A decomposition of self-correction capability for llms. CoRR, abs/2412.19513.
  30. 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
  31. Pengcheng Yin, Graham Neubig, Wen-tau Yih, and Sebastian Riedel. 2020. Tabert: Pretraining for joint understanding of textual and tabular data. arXiv preprint arXiv:2005.08314.
  32. Fei Yu, Hongbo Zhang, Prayag Tiwari, and Benyou Wang. 2024. Natural language reasoning, a survey. ACM Computing Surveys, 56(12):1–39. find
  33. Bin Zhang, Hangyu Mao, Lijuan Li, Zhiwei Xu, Dapeng Li, Rui Zhao, and Guoliang Fan. 2024. Sequential asynchronous action coordination in multi-agent systems: A stackelberg decision transformer approach. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceeding… find
  34. Xuanliang Zhang, Dingzirui Wang, Longxu Dou, Qingfu Zhu, and Wanxiang Che. 2025. A survey of table reasoning with large language models. Frontiers of Computer Science, 19(9):199348. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yu-2025-table-critic-multi-agent-framework/