{"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/simple-recurrent-units-for-highly","title":"Simple Recurrent Units for Highly Parallelizable Recurrence","arxiv_id":"1709.02755","date":"2017-09-08","proceeding":"EMNLP 2018 10","authors":["Tao Lei","Yu Zhang","Sida I. Wang","Hui Dai","Yoav Artzi"],"abstract":"Common recurrent neural architectures scale poorly due to the intrinsic\ndifficulty in parallelizing their state computations. In this work, we propose\nthe Simple Recurrent Unit (SRU), a light recurrent unit that balances model\ncapacity and scalability. SRU is designed to provide expressive recurrence,\nenable highly parallelized implementation, and comes with careful\ninitialization to facilitate training of deep models. We demonstrate the\neffectiveness of SRU on multiple NLP tasks. SRU achieves 5--9x speed-up over\ncuDNN-optimized LSTM on classification and question answering datasets, and\ndelivers stronger results than LSTM and convolutional models. We also obtain an\naverage of 0.7 BLEU improvement over the Transformer model on translation by\nincorporating SRU into the architecture.","url_abs":"http://arxiv.org/abs/1709.02755v5","url_pdf":"http://arxiv.org/pdf/1709.02755v5.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":"simple-recurrent-units-for-highly","repo_url":"https://github.com/asappresearch/sru","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/Aspire-Mayank/Top-Deep-Learning-Stuffs","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/Helsinki-NLP/OpenNMT-py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/Lukasz-G/Hydra","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/aymericdamien/TopDeepLearning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/butsugiri/shape","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/bzhangGo/lrn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/memray/OpenNMT-kpg-release","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/midobal/OpenNMT-py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/ra1nty/sru-naive","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"simple-recurrent-units-for-highly","repo_url":"https://github.com/suzhiba/SimpleRecurrentUnits-SRU-","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"text-classification","task_name":"Text Classification"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"highway-layer","method_name":"Highway Layer"},{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sru","method_name":"SRU"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[{"slug":"sru","name":"SRU","full_name":"SRU"}],"results":[{"leaderboard":"/sota/machine-translation-on-wmt2014-english-german","task":"Machine Translation","dataset":"WMT2014 English-German","model":"Transformer + SRU","rank_in_archive_order":45,"of":91,"metrics":{"BLEU score":"28.4","Hardware Burden":"34G"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11","task":"Question Answering","dataset":"SQuAD1.1","model":"SRU","rank_in_archive_order":150,"of":213,"metrics":{"EM":"71.4","F1":"80.2","Hardware Burden":"4G"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11-dev","task":"Question Answering","dataset":"SQuAD1.1 dev","model":"SRU","rank_in_archive_order":32,"of":55,"metrics":{"EM":"71.4","F1":"80.2"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1709.02755","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1709.02755"}},"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/ra1nty/sru-naive","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/bzhangGo/lrn","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/butsugiri/shape","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Aspire-Mayank/Top-Deep-Learning-Stuffs","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/suzhiba/SimpleRecurrentUnits-SRU-","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Lukasz-G/Hydra","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/midobal/OpenNMT-py","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Helsinki-NLP/OpenNMT-py","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/aymericdamien/TopDeepLearning","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/memray/OpenNMT-kpg-release","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/asappresearch/sru","reach":{"status":"unanswered"}}],"summary":{"ran_violates":1},"by_repo_kind":{"listed":{"samples":1,"ran":1,"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":"0c93830510f645a3","entry":"check_sru_requirement","repo":"butsugiri/shape","repo_kind":"listed","path":"onmt/models/sru.py","file_url":"https://github.com/butsugiri/shape/blob/HEAD/onmt/models/sru.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0c93830510f645a3"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}