Elamur Research⌕
☰
home / papers / 2026 / FLAIR: Steering LLM Mathematical Problem Solving based on A Fuzzy-Logic-AssIsted Reasoner

FLAIR: Steering LLM Mathematical Problem Solving based on A Fuzzy-Logic-AssIsted Reasoner

Paper · 2026 · Hao Wu, Hongru Sun, Wanqing Li et al.

Summary

LLMs solving math word problems usually follow one fixed reasoning style or a hard-coded pipeline. This ignores the fact that one solution attempt can contain several flaws at once, each with different severity. FLAIR (Fuzzy-Logic-AssIsted Reasoner) addresses this by treating the diagnosis of a partial solution as a fuzzy-logic problem.

After each attempt, the system scores the intermediate answer with graded membership values in four fuzzy sets, each tied to a common failure type. Parameterized Fuzzy Reasoning Rules then map these scores to a corrective action chosen from an action pool. Reinforcement Learning tunes the rule weights, using whether the final solution is correct as the reward. The loop repeats until the answer passes verification.

The authors evaluate FLAIR on five standard math benchmarks, including GSM8K and MATH, with six base LLMs. They report consistent gains over recent state-of-the-art prompting and verification methods. The approach also gives an interpretable account of what went wrong at each step. The authors describe it as the first use of fuzzy theory for LLM mathematical reasoning and argue it could transfer to agent control and tool-based planning.

Key points

  • Fuzzy sets from an error taxonomy. The four sets are Extraction Fidelity, Operational Accuracy, Step Validity and Answer Conformity. They are derived from a taxonomy by Yin et al. (2025) that lists 39 error types observed across 15 LLMs. Grouped into four categories, these errors cover 92.3% of failures.
  • Corrective actions. The action pool includes Reasoning Example Provision, Entity-Relation Extraction, Chain Decomposition and Error Step Identification.
  • Cost. FLAIR uses 3 API calls per iteration. The authors present this as comparable to or cheaper than baselines such as LBS3 and SAFE, which need more calls.
  • Case study. On GSM8K test item 855, FLAIR triggers Entity-Relation Extraction and then Chain Decomposition. This moves the answer from 48 to the correct 144.
  • Stated limitations.
    • The fuzzy sets and rules are hand-designed.
    • The inference mechanism is simple; the authors suggest type-2 or hierarchical fuzzy systems as future options.
    • Compute cost grows as the number of refinement iterations increases.

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

🎓University of Wisconsin–MadisonMadison · United States54 works · 49,089 citations🎓Florida International UniversityMiami · United States10 works · 1,508 citations🎓University of SouthamptonSouthampton · United Kingdom13 works · 2,608 citations🎓Western Sydney UniversityRichmond · Australia5 works · 63 citations🎓Central China Normal UniversityWuhan · China3 works · 1 citation🎓University of WollongongWollongong · Australia3 works · 1 citation

 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 worksComputational Theory and Mathematics · 321 works

🧭 Topics 2Research subjects and languages the work is about.Language models & pretraining · 3,894 worksReasoning · 1,685 works

Cite

Cite as Hao Wu, Hongru Sun, Wanqing Li, Xinguo Yu, Hao Ming, Xiao Luo, Wenbin Zhang, Jiahong Zhao, Yi Guo and Jie Yang. 2026. FLAIR: Steering LLM Mathematical Problem Solving based on A Fuzzy-Logic-AssIsted Reasoner. 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.1790

@inproceedings{wu2026flair,
  title = {{FLAIR: Steering LLM Mathematical Problem Solving based on A Fuzzy-Logic-AssIsted Reasoner}},
  author = {Hao Wu, Hongru Sun, Wanqing Li, Xinguo Yu, Hao Ming, Xiao Luo, Wenbin Zhang, Jiahong Zhao, Yi Guo and Jie Yang},
  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.1790},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1790}
}

References

60 references, 23 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. AI-MO Team. 2024a. AIMO Validation AIME dataset. https://huggingface.co/datasets/ AI-MO/aimo-validation-aime. ↗
  2. AI-MO Team. 2024b. AIMO Validation AMC dataset. https://huggingface.co/datasets/ AI-MO/aimo-validation-amc. ↗
  3. Lang Cao, Yingtian Zou, Chao Peng, Renhong Chen, Wu Ning, and Yitong Li. 2025. Step Guided Reasoning: Improving Mathematical Reasoning using Guidance Generation and Step Reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 21112–21129, Suzhou, C… find
  4. Ping Chen, Xiang Liu, Zhaoxiang Liu, Zezhou Chen, Xingpeng Zhang, Huan Hu, Zipeng Wang, Kai Wang, Shuming Shi, and Shiguo Lian. 2025. Fuzzy reasoning chain (FRC): An innovative reasoning framework from fuzziness to clarity. In Findings of the Association for Computational Linguistics: EMNLP 2025, pa… find
  5. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. Training verifiers to solve math word problems. Preprint, arXiv:2110.14168.
  6. Debrup Das, Debopriyo Banerjee, Somak Aditya, and Ashish Kulkarni. 2024. MATHSENSEI: A tool-augmented large language model for mathematical reasoning. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Vo… find
  7. Shizhe Diao, Pengcheng Wang, Yong Lin, Rui Pan, Xiang Liu, and Tong Zhang. 2024. Active prompting with chain-of-thought for large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1330–1350, Bangkok, Thailand. A… find
  8. Aniket Didolkar, Anirudh Goyal, Nan Rosemary Ke, Siyuan Guo, Michal Valko, Timothy Lillicrap, Danilo Rezende, Yoshua Bengio, Michael Mozer, and Sanjeev Arora. 2024. Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem Solving. In Advances in Neural Information Processing System… find
  9. Yang Fang, Cheng Xu, Shuhao Guan, Nan Yan, and Yuke Mei. 2024. Advancing Arabic sentiment analysis: ArSen benchmark and the improved fuzzy deep hybrid network. In Proceedings of the 28th Conference on Computational Natural Language Learning, pages 507–516, Miami, FL, USA. Association for Computation…
  10. Pengcheng He, Jianfeng Gao, and Weizhu Chen. 2023. DeBERTav3: Improving deBERTa using ELECTRA-style pre-training with gradient-disentangled embedding sharing. In The Eleventh International Conference on Learning Representations.
  11. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. Measuring mathematical problem solving with the math dataset. In Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks, volume 1.
  12. Cuong Hoang, Devendra Sachan, Prashant Mathur, Brian Thompson, and Marcello Federico. 2023. Improving retrieval augmented neural machine translation by controlling source and fuzzy-match interactions. In Findings of the Association for Computational Linguistics: EACL 2023, pages 289–295, Dubrovnik,… find
  13. Xijie Huang, Li Lyna Zhang, Kwang-Ting Cheng, Fan Yang, and Mao Yang. 2024. Fewer is more: Boosting math reasoning with reinforced context pruning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 13674–13695, Miami, Florida, USA. Association for Compu… find
  14. Yiming Huang, Xiao Liu, Yeyun Gong, Zhibin Gou, Yelong Shen, Nan Duan, and Weizhu Chen. 2025. Key-point-driven data synthesis with its enhancement on mathematical reasoning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 24176–24184. find
  15. J. Liu, Z. Huang, Q. Liu, Z. Ma, C. Zhai, and E. Chen. 2025. Knowledge-Centered Dual-Process Reasoning for Math Word Problems With Large Language Models. IEEE Transactions on Knowledge and Data Engineering, 37(6):3457–3471. find
  16. Feiran Jia, Tong Wu, Xin Qin, and Anna Squicciarini. 2025. The task shield: Enforcing task alignment to defend against indirect prompt injection in LLM agents. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 29680–29697, Vienn…
  17. Weisen Jiang, Han Shi, Longhui Yu, Zhengying Liu, Yu Zhang, Zhenguo Li, and James Kwok. 2024a. Forward-backward reasoning in large language models for mathematical verification. In Findings of the Association for Computational Linguistics: ACL 2024, pages 6647–6661, Bangkok, Thailand. Association fo… find
  18. Zhuoxuan Jiang, Haoyuan Peng, Shanshan Feng, Fan Li, and Dongsheng Li. 2024b. Llms can find mathematical reasoning mistakes by pedagogical chain-of-thought. In Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, IJCAI-24, pages 3439–3447. International Joint Co… find
  19. Alex Kulesza and Ben Taskar. 2012. Determinantal point processes for machine learning. Foundations and Trends in Machine Learning, 5(2–3):123–286. find
  20. Joshua Ong Jun Leang, Aryo Pradipta Gema, and Shay B Cohen. 2025. CoMAT: Chain of mathematically annotated thought improves mathematical reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 20256–20285, Suzhou, China. Association for Computation…
  21. Junzhuo Li and Deyi Xiong. 2022. KaFSP: Knowledge-aware fuzzy semantic parsing for conversational question answering over a large-scale knowledge base. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 461–473, Dublin, Ireland.… find
  22. Yiyuan Li, Shichao Sun, and Pengfei Liu. 2024. FRoG: Evaluating fuzzy reasoning of generalized quantifiers in LLMs. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 7239–7256, Miami, Florida, USA. Association for Computational Linguistics. find
  23. Zhenwen Liang, Jipeng Zhang, and Xiangliang Zhang. 2023. Don’t be blind to questions: Question-oriented math word problem solving. In Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Comput… find
  24. Yen-Ting Lin, Di Jin, Tengyu Xu, Tianhao Wu, Sainbayar Sukhbaatar, Chen Zhu, Yun He, Yun-Nung Chen, Jason E Weston, Yuandong Tian, Arash Rahnama, Sinong Wang, Hao Ma, and Han Fang. 2025. Step-KTO: Optimizing Mathematical Reasoning through Stepwise Binary Feedback. In Proceedings of The 3rd Workshop… find
  25. Chengwu Liu, Ye Yuan, Yichun Yin, Yan Xu, Xin Xu, Zaoyu Chen, Yasheng Wang, Lifeng Shang, Qun Liu, and Ming Zhang. 2025a. Safe: Enhancing Mathematical Reasoning in Large Language Models via Retrospective Step-aware Formal Verification. In Proceedings of the 63rd Annual Meeting of the Association for… find
  26. Haoxiong Liu, Yifan Zhang, Yifan Luo, and Andrew C Yao. 2025b. Augmenting math word problems via iterative question composing. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 24605–24613.
  27. Zihan Liu, Yang Chen, Mohammad Shoeybi, Bryan Catanzaro, and Wei Ping. 2025c. AceMath: Advancing Frontier Math Reasoning with Post-Training and Reward Modeling. In Findings of the Association for Computational Linguistics: ACL 2025, pages 3993–4015, Vienna, Austria. Association for Computational Lin… find
  28. Haipeng Luo, Qingfeng Sun, Can Xu, Pu Zhao, Jian-Guang Lou, Chongyang Tao, Xiubo Geng, Qingwei Lin, Shifeng Chen, Yansong Tang, and Dongmei Zhang. 2025a. WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instruct. In The Thirteenth International Conference o…
  29. Md Kamrul Hasan. 2023. Math word problem solving by generating linguistic variants of problem statements. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop), pages 362–378, Toronto, Canada. Association for Computational Li… find
  30. Kangyang Luo, Zichen Ding, Zhenmin Weng, Lingfeng Qiao, Meng Zhao, Xiang Li, Di Yin, and Jinlong Shu. 2025b. Let’s be self-generated via step by step: A curriculum learning approach to automated reasoning with large language models. In Findings of the Association for Computational Linguistics: ACL 2… find
  31. Hangliang Ren. 2025. LSRL: Process-Supervised GRPO on Latent Recurrent States Improves Mathematical Reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 12534–12545, Suzhou, China. Association for Computational Linguistics. find
  32. 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
  33. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul F Christiano, Jan Leike, and Ryan Lowe. 2022. Tra… — Training language models to follow instructions with human feedback
  34. Arkil Patel, Satwik Bhattamishra, and Navin Goyal. 2021. Are NLP models really able to solve simple math word problems? In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2080–2094, Online. Associat…
  35. Qizhi Pei, Lijun Wu, Zhuoshi Pan, Yu Li, Honglin Lin, Chenlin Ming, Xin Gao, Conghui He, and Rui Yan. 2025. MathFusion: Enhancing Mathematical Problem-solving of LLM through Instruction Fusion. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long…
  36. Tianshuo Peng, Zuchao Li, Lefei Zhang, Bo Du, and Hai Zhao. 2023. FSUIE: A novel fuzzy span mechanism for universal information extraction. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 16318–16333, Toronto, Canada. Associat… find
  37. Syed Rifat Raiyan, Md Nafis Faiyaz, Shah Md. Jawad Kabir, Mohsinul Kabir, Hasan Mahmud, and Evangelia Spiliopoulou, Riccardo Fogliato, Hanna Burnsky, Tamer Soliman, Jie Ma, Graham Horwood, and Miguel Ballesteros. 2025. Play favorites: A statistical method to measure self-bias in llm-as-a-judge. Prep… — [arXiv:2508.06709](https://arxiv.org/abs/2508.06709) find
  38. Team. 2025. Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities. Preprint, arXiv:2507.06261.
  39. Yuxuan Tong, Xiwen Zhang, Rui Wang, Ruidong Wu, and Junxian He. 2024. Dart-math: Difficulty-aware rejection tuning for mathematical problem-solving. Advances in Neural Information Processing Systems, 37:7821–7846.
  40. L. X. Wang. 1992. Fuzzy systems are universal approximators. International Journal of General Systems, 17(2-3):59–75. find
  41. Ruida Wang, Yuxin Li, Yi R. Fung, and Tong Zhang. 2025. Let’s Reason Formally: Natural-Formal Hybrid Reasoning Enhances LLM’s Math Capability. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 16794–16820, Suzhou, China. Association for Computational Li…
  42. Zhenyu Wu, Qingkai Zeng, Zhihan Zhang, Zhaoxuan Tan, Chao Shen, and Meng Jiang. 2025. Enhancing Mathematical Reasoning in LLMs by Stepwise Correction. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 21602–21623, Vienna, Austri…
  43. Fred Xu, Song Jiang, Zijie Huang, Xiao Luo, Shichang Zhang, Yuanzhou Chen, and Yizhou Sun. 2024a. FUSE: Measure-theoretic compact fuzzy set representation for taxonomy expansion. In Findings of the Association for Computational Linguistics: ACL 2024, pages 2707–2720, Bangkok, Thailand. Association f… find
  44. Jundong Xu, Hao Fei, Meng Luo, Qian Liu, Liangming Pan, William Yang Wang, Preslav Nakov, Mong-Li Lee, and Wynne Hsu. 2025. Aristotle: Mastering logical reasoning with a logic-complete decompose-search-resolve framework. In Proceedings of the 63rd Annual Meeting of the Association for Computational…
  45. Wenda Xu, Guanglei Zhu, Xuandong Zhao, Liangming Pan, Lei Li, and William Wang. 2024b. Pride and prejudice: LLM amplifies self-bias in self-refinement. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15474–15492, Bangkok, Thai…
  46. Y. Li, D. M. Ubaidali, L. Wang, and W. Zhang. 2025. Step-by-Step Correction of LLM-based Math Word Problems Solutions. In ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1–5. Journal Abbreviation: ICASSP 2025 - 2025 IEEE International Confe… find
  47. Yibo Yan, Jiamin Su, Jianxiang He, Fangteng Fu, Xu Zheng, Yuanhuiyi Lyu, Kun Wang, Shen Wang, Qingsong Wen, and Xuming Hu. 2025. A survey of mathematical reasoning in the era of multimodal large language model: Benchmark, method & challenges. In Findings of the Association for Computational Linguist… find
  48. Jie Yang, Brian Yecies, Jun Ma, and Wanqing Li. 2022. Sparse fuzzy classification for profiling online users and relevant user-generated content. Expert Systems with Applications, 194:116497. find
  49. Wen Yang, Minpeng Liao, and Kai Fan. 2025. Markov Chain of Thought for Efficient Mathematical Reasoning. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 7132– 715… find
  50. Bohan Yao and Vikas Yadav. 2025. Diverse Multi-tool Aggregation with Large Language Models for Enhanced Math Reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 25264–25282, Suzhou, China. Association for Computational Linguistics. find
  51. Zhangyue Yin, YuHong Sun, Xuanjing Huang, Xipeng Qiu, and Hui Zhao. 2025. Error Classification of Large Language Models on Math Word Problems: A Dynamically Adaptive Framework. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 338–365, Suzhou, China. Association for Com… find
  52. Erxin Yu, Jing Li, Ming Liao, Qi Zhu, Boyang Xue, Minghui Xu, Baojun Wang, Lanqing Hong, Fei Mi, and Lifeng Shang. 2025a. Self-Error-Instruct: Generalizing from Errors for LLMs Mathematical Reasoning. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume…
  53. Longhui Yu, Weisen Jiang, Han Shi, Jincheng YU, Zhengying Liu, Yu Zhang, James Kwok, Zhenguo Li, Adrian Weller, and Weiyang Liu. 2024. Metamath: Bootstrap your own mathematical questions for large language models. In The Twelfth International Conference on Learning Representations.
  54. Yiyao Yu, Yuxiang Zhang, Dongdong Zhang, Xiao Liang, Hengyuan Zhang, Xingxing Zhang, Mahmoud Khademi, Hany Hassan Awadalla, Junjie Wang, Yujiu Yang, and Furu Wei. 2025b. Chain-of-Reasoning: Towards Unified Mathematical Reasoning in Large Language Models via a Multi-Paradigm Perspective. In Proceedin…
  55. L. A. Zadeh. 1965. Fuzzy sets. Information and Control, 8(3):338–353. find
  56. Jixiao Zhang and Chunsheng Zuo. 2025. GRPO-LEAD: A Difficulty-Aware Reinforcement Learning Approach for Concise Mathematical Reasoning in Language Models. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 5642–5665, Suzhou, China. Association for Comput…
  57. Wenqi Zhang, Yongliang Shen, Guiyang Hou, Kuangyi Wang, and Weiming Lu. 2024a. Specialized mathematical solving by a step-by-step expression chain generation. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 32:3128–3140. find
  58. Wenqi Zhang, Yongliang Shen, Linjuan Wu, Qiuying Peng, Jun Wang, Yueting Zhuang, and Weiming Lu. 2024b. Self-contrast: Better reflection through inconsistent solving perspectives. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), page…
  59. Huaixiu Steven Zheng, Swaroop Mishra, Xinyun Chen, Heng-Tze Cheng, Ed H. Chi, Quoc V Le, and Denny Zhou. 2024. Take a step back: Evoking reasoning via abstraction in large language models. In The Twelfth International Conference on Learning Representations.
  60. Yue Zhou, Yada Zhu, Diego Antognini, Yoon Kim, and Yang Zhang. 2024. Paraphrase and solve: Exploring and exploiting the impact of surface form on mathematical reasoning in large language models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wu-2026-flair-steering-llm-mathematical-problem/