mlsanityData to 7 Oct 2026

Learning to Reason Deductively: Math Word Problem Solving as Complex Relation Extraction

arXiv 2203.10316 ↗First posted 19 Mar 2022cs.CL cs.LGcited by 44

On the map: LLM reasoning → Mathematical reasoning

Abstract (arXiv)

Solving math word problems requires deductive reasoning over the quantities in the text. Various recent research efforts mostly relied on sequence-to-sequence or sequence-to-tree models to generate mathematical expressions without explicitly performing relational reasoning between quantities in the given context. While empirically effective, such approaches typically do not provide explanations for the generated expressions. In this work, we view the task as a complex relation extraction problem, proposing a novel approach that presents explainable deductive reasoning steps to iteratively construct target expressions, where each step involves a primitive operation over two quantities defining their relation. Through extensive experiments on four benchmark datasets, we show that the proposed model significantly outperforms existing strong baselines. We further demonstrate that the deductive procedure not only presents more explainable steps but also enables us to make more accurate predictions on questions that require more complex reasoning.

Key facts. Cited by 44 papers in the map (data through 7 Oct 2026). Placed in Mathematical reasoning, one of 17 directions in LLM reasoning. Builds on 29 of its 75 references that are in the map. First posted on arXiv 19 Mar 2022.

Builds on

29 of 75 references.

References · 29

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  4. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  5. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,907
  6. 2020Transformers are RNNs: Fast Autoregressive Transformers with Linear Attentioncited by 1,979
  7. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,303
  8. 2015Effective Approaches to Attention-based Neural Machine Translationcited by 1,024
  9. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 861
  10. 2019MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalismscited by 653
  11. 2017Program Induction by Rationale Generation : Learning to Solve and Explain Algebraic Word Problemscited by 642
  12. 2015Neural Module Networkscited by 375
17 more
  1. 2016Solving General Arithmetic Word Problemscited by 353
  2. 2017Six Challenges for Neural Machine Translationcited by 334
  3. 2016Rationalizing Neural Predictionscited by 281
  4. 2017End-to-end Neural Coreference Resolutioncited by 240
  5. 2017Lexically Constrained Decoding for Sequence Generation Using Grid Beam Searchcited by 234
  6. 2017Neural-Symbolic Learning and Reasoning: A Survey and Interpretationcited by 195
  7. 2018Higher-order Coreference Resolution with Coarse-to-fine Inferencecited by 150
  8. 2021Generate & Rank: A Multi-task Framework for Math Word Problemscited by 82
  9. 2019Neural Module Networks for Reasoning over Textcited by 65
  10. 2018Semantically-Aligned Equation Generation for Solving and Reasoning Math Word Problemscited by 49
  11. 2021MWPToolkit: An Open-Source Framework for Deep Learning-Based Math Word Problem Solverscited by 39
  12. 2021Neural-Symbolic Solver for Math Word Problems with Auxiliary Taskscited by 38
  13. 2017Mapping to Declarative Knowledge for Word Problem Solvingcited by 36
  14. 2020Graph-to-Tree Neural Networks for Learning Structured Input-Output Translation with Applications to Semantic Parsing and Math Word Problemcited by 31
  15. 2021Seeking Patterns, Not just Memorizing Procedures: Contrastive Learning for Solving Math Word Problemscited by 30
  16. 2021Investigating Math Word Problems using Pretrained Multilingual Language Modelscited by 21
  17. 2018A Meaning-based Statistical English Math Word Problem Solvercited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 8; 2023: 21; 2024: 9; 2025: 4; 2026: 2.

44 papers.

Sort:

By area:

  1. 2023WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instructcited by 618
  2. 2023Scaling Relationship on Learning Mathematical Reasoning with Large Language Modelscited by 417
  3. 2023REFINER: Reasoning Feedback on Intermediate Representationscited by 203
  4. 2022A Survey of Deep Learning for Mathematical Reasoningcited by 124
  5. 2024A Comprehensive Survey of Scientific Large Language Models and Their Applications in Scientific Discoverycited by 96
  6. 2024A Review of Large Language Models and Autonomous Agents in Chemistrycited by 96
  7. 2024Self-Contrast: Better Reflection Through Inconsistent Solving Perspectivescited by 71
  8. 2024Interpreting and Improving Large Language Models in Arithmetic Calculationcited by 43
  9. 2023Let GPT be a Math Tutor: Teaching Math Word Problem Solvers with Customized Exercise Generationcited by 27
  10. 2023Mathematical Language Models: A Surveycited by 24
  11. 2023A Symbolic Character-Aware Model for Solving Geometry Problemscited by 20
  12. 2023Interpretable Math Word Problem Solution Generation Via Step-by-step Planningcited by 17
32 more
  1. 2023MinT: Boosting Generalization in Mathematical Reasoning via Multi-View Fine-Tuningcited by 16
  2. 2022Chaining Simultaneous Thoughts for Numerical Reasoningcited by 14
  3. 2022Multi-View Reasoning: Consistent Contrastive Learning for Math Word Problemcited by 13
  4. 2023Learning Multi-Step Reasoning by Solving Arithmetic Taskscited by 12
  5. 2023Design of Chain-of-Thought in Math Problem Solvingcited by 12
  6. 2025A Survey on Mathematical Reasoning and Optimization with Large Language Modelscited by 12
  7. 2022Self-consistent Reasoning For Solving Math Word Problemscited by 10
  8. 2022DyRRen: A Dynamic Retriever-Reranker-Generator Model for Numerical Reasoning over Tabular and Textual Datacited by 10
  9. 2022A Survey in Mathematical Language Processingcited by 9
  10. 2023Fill in the Blank: Exploring and Enhancing LLM Capabilities for Backward Reasoning in Math Word Problemscited by 9
  11. 2023Three Questions Concerning the Use of Large Language Models to Facilitate Mathematics Learningcited by 9
  12. 2024Exploring the Limitations of Large Language Models in Compositional Relation Reasoningcited by 9
  13. 2022APOLLO: An Optimized Training Approach for Long-form Numerical Reasoningcited by 8
  14. 2023Non-Autoregressive Math Word Problem Solver with Unified Tree Structurecited by 7
  15. 2023Math Word Problem Solving by Generating Linguistic Variants of Problem Statementscited by 7
  16. 2023Learning by Applying: A General Framework for Mathematical Reasoning via Enhancing Explicit Knowledge Learningcited by 5
  17. 2024Chain-of-Translation Prompting (CoTR): A Novel Prompting Technique for Low Resource Languagescited by 5
  18. 2022Analogical Math Word Problems Solving with Enhanced Problem-Solution Associationcited by 4
  19. 2023An Expression Tree Decoding Strategy for Mathematical Equation Generationcited by 4
  20. 2023Solving Math Word Problems with Reexaminationcited by 2
  21. 2025MathEDU: Feedback Generation on Problem-Solving Processes for Mathematical Learning Supportcited by 2
  22. 2026Transparent AI for Mathematics: Transformer-Based Large Language Models for Mathematical Entity Relationship Extraction with XAIcited by 2
  23. 2023A Framework For Refining Text Classification and Object Recognition from Academic Articlescited by 1
  24. 2023Object Recognition from Scientific Document based on Compartment Refinement Frameworkcited by 1
  25. 2024Mathematical Derivation Graphs: A Relation Extraction Task in STEM Manuscriptscited by 1
  26. 2025Foundation of Intelligence: Review of Math Word Problems from Human Cognition Perspectivecited by 1
  27. 2023Techniques to Improve Neural Math Word Problem Solverscited by 0
  28. 2023Multi-Method Self-Training: Improving Code Generation With Text, And Vice Versacited by 0
  29. 2024HiLight: A Hierarchy-aware Light Global Model with Hierarchical Local ConTrastive Learningcited by 0
  30. 2024A Theoretical Review on Solving Algebra Problemscited by 0
  31. 2025A Diversity-Enhanced Knowledge Distillation Model for Practical Math Word Problem Solvingcited by 0
  32. 2026Generative Optimization for Incentivized Advertising with Global Level Constraintscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022Learning to Reason With Relational Abstractions
  2. 2025DeduCE: Deductive Consistency as a Framework to Evaluate LLM Reasoning
  3. 2025MathFimer: Enhancing Mathematical Reasoning by Expanding Reasoning Steps through Fill-in-the-Middle Task
  4. 2024Logic Contrastive Reasoning with Lightweight Large Language Model for Math Word Problems
  5. 2026Constraint-First Reasoning: A Training-Free Protocol for Exploiting Answer-Space Constraints in Mathematical Problem Solving
  6. 2023Reasoning in Large Language Models Through Symbolic Math Word Problems
  7. 2025MetaLadder: Ascending Mathematical Solution Quality via Analogical-Problem Reasoning Transfer
  8. 2024ReasonAgain: Using Extractable Symbolic Programs to Evaluate Mathematical Reasoning