{"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/neural-symbolic-vqa-disentangling-reasoning","title":"Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding","arxiv_id":"1810.02338","date":"2018-10-04","proceeding":"NeurIPS 2018 12","authors":["Kexin Yi","Jiajun Wu","Chuang Gan","Antonio Torralba","Pushmeet Kohli","Joshua B. Tenenbaum"],"abstract":"We marry two powerful ideas: deep representation learning for visual\nrecognition and language understanding, and symbolic program execution for\nreasoning. Our neural-symbolic visual question answering (NS-VQA) system first\nrecovers a structural scene representation from the image and a program trace\nfrom the question. It then executes the program on the scene representation to\nobtain an answer. Incorporating symbolic structure as prior knowledge offers\nthree unique advantages. First, executing programs on a symbolic space is more\nrobust to long program traces; our model can solve complex reasoning tasks\nbetter, achieving an accuracy of 99.8% on the CLEVR dataset. Second, the model\nis more data- and memory-efficient: it performs well after learning on a small\nnumber of training data; it can also encode an image into a compact\nrepresentation, requiring less storage than existing methods for offline\nquestion answering. Third, symbolic program execution offers full transparency\nto the reasoning process; we are thus able to interpret and diagnose each\nexecution step.","url_abs":"http://arxiv.org/abs/1810.02338v2","url_pdf":"http://arxiv.org/pdf/1810.02338v2.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":"neural-symbolic-vqa-disentangling-reasoning","repo_url":"https://github.com/kexinyi/ns-vqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"neural-symbolic-vqa-disentangling-reasoning","repo_url":"https://github.com/nerdimite/neuro-symbolic-ai-soc","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-question-answering-on-clevr","task":"Visual Question Answering (VQA)","dataset":"CLEVR","model":"NS-VQA (1K programs)","rank_in_archive_order":1,"of":15,"metrics":{"Accuracy":"99.8"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-clevr-humans","task":"Visual Question Answering (VQA)","dataset":"CLEVR-Humans","model":"NS-VQA (1K programs)","rank_in_archive_order":4,"of":5,"metrics":{"Accuracy":"67.8"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1810.02338","atlas_url":"https://app.syntology.ai/?focus=1810.02338","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1810.02338"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/kexinyi/ns-vqa","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/nerdimite/neuro-symbolic-ai-soc","reach":{"status":"ok"}}],"summary":{"ran_fixture":1,"ran_draft_wrong":1},"by_repo_kind":{"listed":{"samples":2,"ran":2,"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":2,"samples":[{"code_sha256_prefix":"61279165e77b7eb5","entry":"check_program","repo":"kexinyi/ns-vqa","repo_kind":"listed","path":"reason/tools/run_test.py","file_url":"https://github.com/kexinyi/ns-vqa/blob/HEAD/reason/tools/run_test.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"61279165e77b7eb5"}},{"code_sha256_prefix":"be76eac5096d1b27","entry":"find_clevr_question_type","repo":"kexinyi/ns-vqa","repo_kind":"listed","path":"reason/tools/run_test.py","file_url":"https://github.com/kexinyi/ns-vqa/blob/HEAD/reason/tools/run_test.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"be76eac5096d1b27"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}