{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,17]],"date-time":"2026-04-17T16:48:03Z","timestamp":1776444483376,"version":"3.51.2"},"reference-count":74,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"6","license":[{"start":{"date-parts":[[2023,6,1]],"date-time":"2023-06-01T00:00:00Z","timestamp":1685577600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2023,6,1]],"date-time":"2023-06-01T00:00:00Z","timestamp":1685577600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2023,6,1]],"date-time":"2023-06-01T00:00:00Z","timestamp":1685577600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2023,6,1]]},"DOI":"10.1109\/tpami.2022.3225334","type":"journal-article","created":{"date-parts":[[2022,11,29]],"date-time":"2022-11-29T19:24:42Z","timestamp":1669749882000},"page":"7739-7750","source":"Crossref","is-referenced-by-count":2,"title":["State-Regularized Recurrent Neural Networks to Extract Automata and Explain Predictions"],"prefix":"10.1109","volume":"45","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9177-9563","authenticated-orcid":false,"given":"Cheng","family":"Wang","sequence":"first","affiliation":[{"name":"Amazon, Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Carolin","family":"Lawrence","sequence":"additional","affiliation":[{"name":"NEC Laboratories Europe, Heidelberg, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mathias","family":"Niepert","sequence":"additional","affiliation":[{"name":"NEC Laboratories Europe, Heidelberg, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref13","article-title":"Frustratingly short attention spans in neural language modeling","author":"daniluk","year":"2017"},{"key":"ref57","first-page":"5998","article-title":"Attention is all you need","author":"vaswani","year":"2017","journal-title":"Proc 31st Int Conf Neural Inf Process Syst"},{"key":"ref12","first-page":"1986","article-title":"Associative long short-term memory","author":"danihelka","year":"2016","journal-title":"Proc 33rd Int Conf Mach Learn"},{"key":"ref56","first-page":"4965","article-title":"Learning longer-term dependencies in RNNs with auxiliary losses","author":"trinh","year":"2018","journal-title":"Proc 35th Int Conf Mach Learn"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1016\/0364-0213(90)90002-E"},{"key":"ref59","article-title":"A comparison of rule extraction for different recurrent neural network models and grammatical complexity","author":"wang","year":"2018"},{"key":"ref14","first-page":"1252","article-title":"Noisin: Unbiased regularization for recurrent neural networks","author":"dieng","year":"2018","journal-title":"Proc 35th Int Conf Mach Learn"},{"key":"ref58","article-title":"Uncertainty estimation and calibration with finite-state probabilistic RNNs","author":"wang","year":"2021","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref53","article-title":"Theano: A Python framework for fast computation of mathematical expressions","year":"2016"},{"key":"ref52","article-title":"Visual analysis of hidden state dynamics in recurrent neural networks","author":"strobelt","year":"2016"},{"key":"ref11","first-page":"3079","article-title":"Semi-supervised sequence learning","author":"dai","year":"2015","journal-title":"Proc 28th Int Conf Neural Inf Process Syst"},{"key":"ref55","first-page":"105","article-title":"Dynamic construction of finite automata from examples using hill-climbing","author":"tomita","year":"1982","journal-title":"Proc 4th Annu Conf Cogn Sci Soc"},{"key":"ref10","first-page":"1163","article-title":"Training restricted boltzmann machines on word observations","author":"dahl","year":"2012","journal-title":"Proc 29th Int Conf Mach Learn"},{"key":"ref54","first-page":"26","article-title":"Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude","volume":"4","author":"tieleman","year":"2012","journal-title":"COURSERA Neural Netw Mach Learn"},{"key":"ref17","first-page":"2199","article-title":"Sequential neural models with stochastic layers","author":"fraccaro","year":"2016","journal-title":"Proc 30th Int Conf Neural Inf Process Syst"},{"key":"ref16","first-page":"1136","article-title":"Input switched affine networks: An RNN architecture designed for interpretability","author":"foerster","year":"2017","journal-title":"Proc 34th Int Conf Mach Learn"},{"key":"ref19","first-page":"1019","article-title":"A theoretically grounded application of dropout in recurrent neural networks","author":"gal","year":"2016","journal-title":"Proc 30th Int Conf Neural Inf Process Syst"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/ICPR.1994.576922"},{"key":"ref51","article-title":"Deep inside convolutional networks: Visualising image classification models and saliency maps","author":"simonyan","year":"2013"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1016\/0304-3975(94)90178-3"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.3115\/1603899.1603912"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v36i10.21364"},{"key":"ref48","author":"siegelmann","year":"2012","journal-title":"Neural Networks and Analog Computation Beyond the Turing Limit"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1162\/089976602320263980"},{"key":"ref42","article-title":"When recurrent models don't need to be recurrent","author":"miller","year":"2018"},{"key":"ref41","article-title":"Regularizing and optimizing LSTM language models","author":"merity","year":"2018","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref44","article-title":"Automatic rule extraction from long short term memory networks","author":"murdoch","year":"2017","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1016\/j.dsp.2017.10.011"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1145\/130385.130432"},{"key":"ref8","article-title":"Empirical evaluation of gated recurrent neural networks on sequence modeling","author":"chung","year":"2014"},{"key":"ref7","first-page":"77","article-title":"Dilated recurrent neural networks","author":"chang","year":"2017","journal-title":"Proc 31st Int Conf Neural Inf Process Syst"},{"key":"ref9","article-title":"Recurrent batch normalization","author":"cooijmans","year":"2017","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref4","article-title":"Learning stochastic recurrent networks","author":"bayer","year":"2014"},{"key":"ref3","article-title":"An empirical evaluation of generic convolutional and recurrent networks for sequence modeling","author":"bai","year":"2018"},{"key":"ref6","article-title":"Skip RNN: Learning to skip state updates in recurrent neural networks","author":"campos","year":"2018","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref5","first-page":"993","article-title":"Latent dirichlet allocation","volume":"3","author":"blei","year":"2003","journal-title":"J Mach Learn Res"},{"key":"ref40","first-page":"2579","article-title":"Visualizing data using t-SNE","volume":"9","author":"van der maaten","year":"2008","journal-title":"J Mach Learn Res"},{"key":"ref35","article-title":"Zoneout: Regularizing RNNs by randomly preserving hidden activations","author":"krueger","year":"2017","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-97610-0_6"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"ref36","article-title":"A simple way to initialize recurrent networks of rectified linear units","author":"le","year":"2015"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/N15-1011"},{"key":"ref30","article-title":"Categorical reparameterization with gumbel-softmax","author":"jang","year":"2017","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref74","first-page":"6596","article-title":"State-regularized recurrent neural networks","author":"wang","year":"2019","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref33","first-page":"5580","article-title":"What uncertainties do we need in Bayesian deep learning for computer vision?","author":"kendall","year":"2017","journal-title":"Proc 31st Int Conf Neural Inf Process Syst"},{"key":"ref32","article-title":"Visualizing and understanding recurrent networks","author":"karpathy","year":"2016","journal-title":"Proc Int Conf Learn Representations Workshop"},{"key":"ref2","article-title":"Neural machine translation by jointly learning to align and translate","author":"bahdanau","year":"2015","journal-title":"Proc Int Conf Learn Representations"},{"key":"ref1","first-page":"1120","article-title":"Unitary evolution recurrent neural networks","author":"arjovsky","year":"2016","journal-title":"Proc 33rd Int Conf Mach Learn"},{"key":"ref39","first-page":"142","article-title":"Learning word vectors for sentiment analysis","author":"maas","year":"2011","journal-title":"Proc 49th Annu Meeting Assoc Comput Linguistics Hum Lang Technol"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N16-1082"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00920"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1993.5.6.976"},{"key":"ref73","first-page":"4189","article-title":"Recurrent highway networks","author":"zilly","year":"2017","journal-title":"Proc 34th Int Conf Mach Learn"},{"key":"ref72","first-page":"1822","article-title":"Architectural complexity measures of recurrent neural networks","author":"zhang","year":"2016","journal-title":"Proc 30th Int Conf Neural Inf Process Syst"},{"key":"ref24","article-title":"Neural turing machines","author":"graves","year":"2014"},{"key":"ref68","article-title":"ADADELTA: An adaptive learning rate method","author":"zeiler","year":"2012"},{"key":"ref23","first-page":"6713","article-title":"Z-forcing: Training stochastic recurrent networks","author":"goyal","year":"2017","journal-title":"Proc 31st Int Conf Neural Inf Process Syst"},{"key":"ref67","article-title":"Recurrent neural network regularization","author":"zaremba","year":"2014"},{"key":"ref26","first-page":"1828","article-title":"Learning to transduce with unbounded memory","author":"grefenstette","year":"2015","journal-title":"Proc 28th Int Conf Neural Inf Process Syst"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1038\/nature20101"},{"key":"ref69","first-page":"818","article-title":"Visualizing and understanding convolutional networks","author":"zeiler","year":"2014","journal-title":"Proc Eur Conf Comput Vis"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/72.963769"},{"key":"ref64","first-page":"4880","article-title":"Full-capacity unitary recurrent neural networks","author":"wisdom","year":"2016","journal-title":"Proc 30th Int Conf Neural Inf Process Syst"},{"key":"ref63","article-title":"Memory networks","author":"weston","year":"2015"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.1991.155350"},{"key":"ref66","article-title":"Learning to execute","author":"zaremba","year":"2014"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.2000.861302"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P17-1172"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5433"},{"key":"ref27","article-title":"Statistical theory of extreme values and some practical applications. A series of lectures","author":"gumbel","year":"1954","journal-title":"Number 33 US Govt Print Office"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1162\/neco_a_01111"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-2117"},{"key":"ref61","first-page":"5247","article-title":"Extracting automata from recurrent neural networks using queries and counterexamples","author":"weiss","year":"2018","journal-title":"Proc 35th Int Conf Mach Learn"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/34\/10120646\/09965745.pdf?arnumber=9965745","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,5,29]],"date-time":"2023-05-29T17:34:21Z","timestamp":1685381661000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9965745\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,1]]},"references-count":74,"journal-issue":{"issue":"6"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2022.3225334","relation":{},"ISSN":["0162-8828","2160-9292","1939-3539"],"issn-type":[{"value":"0162-8828","type":"print"},{"value":"2160-9292","type":"electronic"},{"value":"1939-3539","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,6,1]]}}}