{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/analysing-mathematical-reasoning-abilities-of-1","title":"Analysing Mathematical Reasoning Abilities of Neural Models","arxiv_id":"1904.01557","date":"2019-04-02","proceeding":"ICLR 2019 5","authors":["David Saxton","Edward Grefenstette","Felix Hill","Pushmeet Kohli"],"abstract":"Mathematical reasoning---a core ability within human intelligence---presents\nsome unique challenges as a domain: we do not come to understand and solve\nmathematical problems primarily on the back of experience and evidence, but on\nthe basis of inferring, learning, and exploiting laws, axioms, and symbol\nmanipulation rules. In this paper, we present a new challenge for the\nevaluation (and eventually the design) of neural architectures and similar\nsystem, developing a task suite of mathematics problems involving sequential\nquestions and answers in a free-form textual input/output format. The\nstructured nature of the mathematics domain, covering arithmetic, algebra,\nprobability and calculus, enables the construction of training and test splits\ndesigned to clearly illuminate the capabilities and failure-modes of different\narchitectures, as well as evaluate their ability to compose and relate\nknowledge and learned processes. Having described the data generation process\nand its potential future expansions, we conduct a comprehensive analysis of\nmodels from two broad classes of the most powerful sequence-to-sequence\narchitectures and find notable differences in their ability to resolve\nmathematical problems and generalize their knowledge.","url_abs":"http://arxiv.org/abs/1904.01557v1","url_pdf":"http://arxiv.org/pdf/1904.01557v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/deepmind/mathematics_dataset","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/andrewschreiber/hs-math-nlp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/berniwal/DeepLearningProject","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/jlrussin/interpret-math-transformer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/mandubian/pytorch_math_dataset","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/mesotron/teaching_transformers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"analysing-mathematical-reasoning-abilities-of-1","repo_url":"https://github.com/r-bakes/math_language_processing","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"math-word-problem-solving","task_name":"Math Word Problem Solving"},{"task_slug":"mathematical-question-answering","task_name":"Mathematical Question Answering"},{"task_slug":"mathematical-reasoning","task_name":"Mathematical Reasoning"},{"task_slug":"question-answering","task_name":"Question Answering"}],"methods":[],"datasets_introduced":[{"slug":"mathematics","name":"Mathematics Dataset","full_name":""}],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-answering-on-mathematics-dataset","task":"Question Answering","dataset":"Mathematics Dataset","model":"Transformer","rank_in_archive_order":2,"of":3,"metrics":{"Accuracy":"0.76"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-mathematics-dataset","task":"Question Answering","dataset":"Mathematics Dataset","model":"LSTM","rank_in_archive_order":3,"of":3,"metrics":{"Accuracy":"0.57"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.01557","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1904.01557"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mandubian/pytorch_math_dataset","reach":{"status":"ok","spdx":"NOASSERTION"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/deepmind/mathematics_dataset","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/andrewschreiber/hs-math-nlp","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jlrussin/interpret-math-transformer","reach":{"status":"ok","spdx":"GPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mesotron/teaching_transformers","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/r-bakes/math_language_processing","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/berniwal/DeepLearningProject","reach":{"status":"gone","observed_at":"2026-09-18","how":"tree_404+repo_404"}}],"summary":{"unverified":1},"by_repo_kind":{"named_in_paper":{"samples":1,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"7e475e235904012f","entry":"pluralize","repo":"deepmind/mathematics_dataset","repo_kind":"named_in_paper","path":"mathematics_dataset/modules/measurement.py","file_url":"https://github.com/deepmind/mathematics_dataset/blob/HEAD/mathematics_dataset/modules/measurement.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"7e475e235904012f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}