{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/recurrent-relational-networks","title":"Recurrent Relational Networks","arxiv_id":"1711.08028","date":"2017-11-21","proceeding":"NeurIPS 2018 12","authors":["Rasmus Berg Palm","Ulrich Paquet","Ole Winther"],"abstract":"This paper is concerned with learning to solve tasks that require a chain of\ninterdependent steps of relational inference, like answering complex questions\nabout the relationships between objects, or solving puzzles where the smaller\nelements of a solution mutually constrain each other. We introduce the\nrecurrent relational network, a general purpose module that operates on a graph\nrepresentation of objects. As a generalization of Santoro et al. [2017]'s\nrelational network, it can augment any neural network model with the capacity\nto do many-step relational reasoning. We achieve state of the art results on\nthe bAbI textual question-answering dataset with the recurrent relational\nnetwork, consistently solving 20/20 tasks. As bAbI is not particularly\nchallenging from a relational reasoning point of view, we introduce\nPretty-CLEVR, a new diagnostic dataset for relational reasoning. In the\nPretty-CLEVR set-up, we can vary the question to control for the number of\nrelational reasoning steps that are required to obtain the answer. Using\nPretty-CLEVR, we probe the limitations of multi-layer perceptrons, relational\nand recurrent relational networks. Finally, we show how recurrent relational\nnetworks can learn to solve Sudoku puzzles from supervised training data, a\nchallenging task requiring upwards of 64 steps of relational reasoning. We\nachieve state-of-the-art results amongst comparable methods by solving 96.6% of\nthe hardest Sudoku puzzles.","url_abs":"http://arxiv.org/abs/1711.08028v4","url_pdf":"http://arxiv.org/pdf/1711.08028v4.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":"recurrent-relational-networks","repo_url":"https://github.com/Kyubyong/sudoku","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"recurrent-relational-networks","repo_url":"https://github.com/rasmusbergpalm/recurrent-relational-networks","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"recurrent-relational-networks","repo_url":"https://github.com/AndreaCossu/Relation-Network-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"recurrent-relational-networks","repo_url":"https://github.com/wDaniec/pytorch-RNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"recurrent-relational-networks","repo_url":"https://github.com/wDaniec/recurrent-relational-network","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"recurrent-relational-networks","repo_url":"https://github.com/dmlc/dgl/tree/master/examples/pytorch/rrn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"diagnostic","task_name":"Diagnostic"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"relational-reasoning","task_name":"Relational Reasoning"}],"methods":[],"datasets_introduced":[{"slug":"rrn-sudoku","name":"rrn-sudoku","full_name":"RRN sudoku instances dataset"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-answering-on-babi","task":"Question Answering","dataset":"bAbi","model":"RR","rank_in_archive_order":11,"of":14,"metrics":{"Mean Error Rate":"0.46%"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1711.08028","atlas_url":"https://app.syntology.ai/?focus=1711.08028","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"1711.08028"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"1711.08028\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"1711.08028"},"call":"get_concepts_for_paper(arxiv_id=\"1711.08028\")"}],"by_repo":[{"repo":"Kyubyong/sudoku","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"rasmusbergpalm/recurrent-relational-networks","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"AndreaCossu/Relation-Network-PyTorch","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"wDaniec/pytorch-RNN","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"wDaniec/recurrent-relational-network","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0},{"repo":"dmlc/dgl","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":6,"harvested_for":["1607.00653","1612.00222","1802.04364","1806.01973","2006.05582","2010.16103"],"harvested_for_calls":[{"arxiv_id":"1607.00653","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1607.00653"},"call":"get_harvested_code_for_paper(arxiv_id=\"1607.00653\")"},{"arxiv_id":"1612.00222","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1612.00222"},"call":"get_harvested_code_for_paper(arxiv_id=\"1612.00222\")"},{"arxiv_id":"1802.04364","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1802.04364"},"call":"get_harvested_code_for_paper(arxiv_id=\"1802.04364\")"},{"arxiv_id":"1806.01973","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1806.01973"},"call":"get_harvested_code_for_paper(arxiv_id=\"1806.01973\")"},{"arxiv_id":"2006.05582","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2006.05582"},"call":"get_harvested_code_for_paper(arxiv_id=\"2006.05582\")"},{"arxiv_id":"2010.16103","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2010.16103"},"call":"get_harvested_code_for_paper(arxiv_id=\"2010.16103\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"7a707e11cb0c92cd75dddabb5a5bf9d4b4488b4810d22ae6d985a5254a4b56c3","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}