{"url":"/sota/math-word-problem-solving-on-math23k","task":{"name":"Math Word Problem Solving","url":"/task/math-word-problem-solving","note":null},"dataset":{"name":"Math23K","url":"/dataset/math23k"},"category":"Reasoning","categories":["Reasoning"],"category_note":null,"description":"A math word problem is a mathematical exercise (such as in a textbook, worksheet, or exam) where significant background information on the problem is presented in ordinary language rather than in mathematical notation. As most word problems involve a narrative of some sort, they are sometimes referred to as story problems and may vary in the amount of technical language used.","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Accuracy (5-fold)","Accuracy (training-test)","weakly-supervised"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy (5-fold)":"higher","Accuracy (training-test)":"higher","weakly-supervised":null}},"counts":{"rows":19,"rows_with_code":14,"rows_with_paper_page":16,"rows_dated":16,"rows_using_additional_data":1},"rows":[{"rank_in_archive_order":1,"model":"GPT-4 (Teaching-Inspired)","metrics":{"Accuracy (5-fold)":"94.3"},"uses_additional_data":false,"paper_date":"2024-10-10","paper":"/paper/teaching-inspired-integrated-prompting","paper_url":"https://arxiv.org/abs/2410.08068v1","paper_title":"Teaching-Inspired Integrated Prompting Framework: A Novel Approach for Enhancing Reasoning in Large Language Models","code":"https://github.com/sallytan13/teaching-inspired-prompting","n_code_links":1,"syntology":{"n_ran":8,"n_unverified":0,"n_samples":8,"n_pointer_only_licence":8}},{"rank_in_archive_order":2,"model":"Multi-view* (ours)","metrics":{"Accuracy (5-fold)":"85.2","Accuracy (training-test)":"87.1"},"uses_additional_data":true,"paper_date":"2022-10-21","paper":"/paper/multi-view-reasoning-consistent-contrastive","paper_url":"https://arxiv.org/abs/2210.11694v2","paper_title":"Multi-View Reasoning: Consistent Contrastive Learning for Math Word Problem","code":"https://github.com/zwq2018/multi-view-consistency-for-mwp","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":12,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"Generate and Rank","metrics":{"Accuracy (5-fold)":"84.3","Accuracy (training-test)":"85.4"},"uses_additional_data":false,"paper_date":"2021-09-07","paper":"/paper/generate-rank-a-multi-task-framework-for-math","paper_url":"https://arxiv.org/abs/2109.03034v1","paper_title":"Generate & Rank: A Multi-task Framework for Math Word Problems","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"Exp-Tree","metrics":{"Accuracy (5-fold)":"84.1","Accuracy (training-test)":"86.2"},"uses_additional_data":false,"paper_date":"2023-10-14","paper":"/paper/an-expression-tree-decoding-strategy-for","paper_url":"https://arxiv.org/abs/2310.09619v3","paper_title":"An Expression Tree Decoding Strategy for Mathematical Equation Generation","code":"https://github.com/zwq2018/multi-view-consistency-for-mwp","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"REAL2: Memory-augmented Solver","metrics":{"Accuracy (5-fold)":"83.18","Accuracy (training-test)":"85.2"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"Roberta-DeductReasoner","metrics":{"Accuracy (5-fold)":"83"},"uses_additional_data":false,"paper_date":"2022-03-19","paper":"/paper/learning-to-reason-deductively-math-word","paper_url":"https://arxiv.org/abs/2203.10316v4","paper_title":"Learning to Reason Deductively: Math Word Problem Solving as Complex Relation Extraction","code":"https://github.com/allanj/deductive-mwp","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":7,"model":"MWP-BERT","metrics":{"Accuracy (5-fold)":"82.4","Accuracy (training-test)":"84.7"},"uses_additional_data":false,"paper_date":"2021-07-28","paper":"/paper/mwp-bert-a-strong-baseline-for-math-word","paper_url":"https://arxiv.org/abs/2107.13435v2","paper_title":"MWP-BERT: Numeracy-Augmented Pre-training for Math Word Problem Solving","code":"https://github.com/lzhenwen/mwp-bert","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":4}},{"rank_in_archive_order":8,"model":"Recall and Learn","metrics":{"Accuracy (5-fold)":"80.8","Accuracy (training-test)":"82.3"},"uses_additional_data":false,"paper_date":"2021-09-27","paper":"/paper/recall-and-learn-a-memory-augmented-solver","paper_url":"https://arxiv.org/abs/2109.13112v1","paper_title":"Recall and Learn: A Memory-augmented Solver for Math Word Problems","code":"https://github.com/sfeng-m/real4mwp","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"RoBERTaGen","metrics":{"Accuracy (5-fold)":"76.6"},"uses_additional_data":false,"paper_date":"2021-09-02","paper":"/paper/mwptoolkit-an-open-source-framework-for-deep","paper_url":"https://arxiv.org/abs/2109.00799v2","paper_title":"MWPToolkit: An Open-Source Framework for Deep Learning-Based Math Word Problem Solvers","code":"https://github.com/lyh-yf/mwptoolkit","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"GTS w/ Data Augmentation","metrics":{"Accuracy (5-fold)":"75.9"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"Graph2Tree","metrics":{"Accuracy (5-fold)":"75.5","Accuracy (training-test)":"77.4"},"uses_additional_data":false,"paper_date":"2020-07-01","paper":"/paper/graph-to-tree-learning-for-solving-math-word","paper_url":"https://aclanthology.org/2020.acl-main.362","paper_title":"Graph-to-Tree Learning for Solving Math Word Problems","code":"https://github.com/2003pro/Graph2Tree","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"SAU-Solver","metrics":{"Accuracy (5-fold)":"74.84"},"uses_additional_data":false,"paper_date":"2020-10-14","paper":"/paper/semantically-aligned-universal-tree","paper_url":"https://arxiv.org/abs/2010.06823v1","paper_title":"Semantically-Aligned Universal Tree-Structured Solver for Math Word Problems","code":"https://github.com/QinJinghui/SAU-Solver","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":13,"model":"GTS","metrics":{"Accuracy (5-fold)":"74.3"},"uses_additional_data":false,"paper_date":"2019-08-10","paper":"/paper/a-goal-driven-tree-structured-neural-model","paper_url":"https://www.ijcai.org/Proceedings/2019/736","paper_title":"A Goal-Driven Tree-Structured Neural Model for Math Word Problems","code":"https://github.com/ShichaoSun/math_seq2tree","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"GROUP-ATT","metrics":{"Accuracy (5-fold)":"66.9","Accuracy (training-test)":"69.5"},"uses_additional_data":false,"paper_date":"2019-07-01","paper":"/paper/modeling-intra-relation-in-math-word-problems","paper_url":"https://aclanthology.org/P19-1619","paper_title":"Modeling Intra-Relation in Math Word Problems with Different Functional Multi-Head Attentions","code":"https://github.com/lijierui/group-attention","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"Hybrid model w/ SNI","metrics":{"Accuracy (5-fold)":"64.7"},"uses_additional_data":false,"paper_date":"2017-09-01","paper":"/paper/deep-neural-solver-for-math-word-problems","paper_url":"https://aclanthology.org/D17-1088","paper_title":"Deep Neural Solver for Math Word Problems","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":16,"model":"ATHENA (roberta-large)","metrics":{"Accuracy (training-test)":"86.5"},"uses_additional_data":false,"paper_date":"2023-11-02","paper":"/paper/athena-mathematical-reasoning-with-thought","paper_url":"https://arxiv.org/abs/2311.01036v1","paper_title":"ATHENA: Mathematical Reasoning with Thought Expansion","code":"https://github.com/the-jb/athena-math","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"ATHENA (roberta-base)","metrics":{"Accuracy (training-test)":"84.4"},"uses_additional_data":false,"paper_date":"2023-11-02","paper":"/paper/athena-mathematical-reasoning-with-thought","paper_url":"https://arxiv.org/abs/2311.01036v1","paper_title":"ATHENA: Mathematical Reasoning with Thought Expansion","code":"https://github.com/the-jb/athena-math","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"T-RNN","metrics":{"Accuracy (training-test)":"66.9"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"LBF","metrics":{"weakly-supervised":"59.8"},"uses_additional_data":false,"paper_date":"2020-12-19","paper":"/paper/learning-by-fixing-solving-math-word-problems","paper_url":"https://arxiv.org/abs/2012.10582v2","paper_title":"Learning by Fixing: Solving Math Word Problems with Weak Supervision","code":"https://github.com/evelinehong/LBF","n_code_links":1,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":6,"rows_with_any_sample_ran":6,"distinct_papers_with_graph_line":6,"distinct_papers_with_any_sample_ran":6,"samples_over_distinct_papers":{"n_ran":15,"n_unverified":17,"n_samples":32,"n_pointer_only_licence":17,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":15,"n_unverified":17,"n_samples":32,"n_pointer_only_licence":17,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}