{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T05:43:23Z","timestamp":1782279803558,"version":"3.54.5"},"reference-count":129,"publisher":"Association for Computing Machinery (ACM)","issue":"5","license":[{"start":{"date-parts":[[2022,12,3]],"date-time":"2022-12-03T00:00:00Z","timestamp":1670025600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Comput. Surv."],"published-print":{"date-parts":[[2023,5,31]]},"abstract":"<jats:p>Despite their success, deep networks are used as black-box models with outputs that are not easily explainable during the learning and the prediction phases. This lack of interpretability is significantly limiting the adoption of such models in domains where decisions are critical such as the medical and legal fields. Recently, researchers have been interested in developing methods that help explain individual decisions and decipher the hidden representations of machine learning models in general and deep networks specifically. While there has been a recent explosion of work on Explainable Artificial Intelligence (ExAI) on deep models that operate on imagery and tabular data, textual datasets present new challenges to the ExAI community. Such challenges can be attributed to the lack of input structure in textual data, the use of word embeddings that add to the opacity of the models and the difficulty of the visualization of the inner workings of deep models when they are trained on textual data.<\/jats:p>\n          <jats:p>Lately, methods have been developed to address the aforementioned challenges and present satisfactory explanations on Natural Language Processing (NLP) models. However, such methods are yet to be studied in a comprehensive framework where common challenges are properly stated and rigorous evaluation practices and metrics are proposed.<\/jats:p>\n          <jats:p>\n            Motivated to democratize ExAI methods in the NLP field, we present in this work a survey that studies\n            <jats:italic>model-agnostic<\/jats:italic>\n            as well as\n            <jats:italic>model-specific<\/jats:italic>\n            explainability methods on NLP models. Such methods can either develop\n            <jats:italic>inherently<\/jats:italic>\n            interpretable NLP models or operate on pre-trained models in a\n            <jats:italic>post hoc<\/jats:italic>\n            manner. We make this distinction and we further decompose the methods into three categories according to what they explain: (1) word embeddings (input level), (2) inner workings of NLP models (processing level), and (3) models\u2019 decisions (output level). We also detail the different evaluation approaches interpretability methods in the NLP field. Finally, we present a case-study on the well-known neural machine translation in an appendix, and we propose promising future research directions for ExAI in the NLP field.\n          <\/jats:p>","DOI":"10.1145\/3529755","type":"journal-article","created":{"date-parts":[[2022,7,19]],"date-time":"2022-07-19T12:11:17Z","timestamp":1658232677000},"page":"1-31","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":96,"title":["On the Explainability of Natural Language Processing Deep Models"],"prefix":"10.1145","volume":"55","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7499-8668","authenticated-orcid":false,"given":"Julia El","family":"Zini","sequence":"first","affiliation":[{"name":"Department of Electrical and Computer Engineering, American University of Beirut, Beirut, Lebanon"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4815-6894","authenticated-orcid":false,"given":"Mariette","family":"Awad","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, American University of Beirut, Beirut, Lebanon"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2022,12,3]]},"reference":[{"key":"e_1_3_2_2_2","first-page":"223","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Allen Carl","year":"2019","unstructured":"Carl Allen and Timothy Hospedales. 2019. Analogies explained: Towards understanding word embeddings. In Proceedings of the International Conference on Machine Learning. 223\u2013231."},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D17-1042"},{"key":"e_1_3_2_4_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N16-1181"},{"key":"e_1_3_2_5_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.12"},{"key":"e_1_3_2_6_2","doi-asserted-by":"crossref","first-page":"211","DOI":"10.1007\/978-3-030-28954-6_11","volume-title":"Explainable ai: Interpreting, Explaining and Visualizing Deep Learning","author":"Arras Leila","year":"2019","unstructured":"Leila Arras, Jos\u00e9 Arjona-Medina, Michael Widrich, Gr\u00e9goire Montavon, Michael Gillhofer, Klaus-Robert M\u00fcller, Sepp Hochreiter, and Wojciech Samek. 2019. Explaining and interpreting LSTMs. In Explainable ai: Interpreting, Explaining and Visualizing Deep Learning. Springer, 211\u2013238."},{"key":"e_1_3_2_7_2","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0130140"},{"key":"e_1_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.5555\/1756006.1859912"},{"key":"e_1_3_2_9_2","article-title":"Neural machine translation by jointly learning to align and translate","author":"Bahdanau Dzmitry","year":"2014","unstructured":"Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. arXiv:1409.0473. Retrieved from https:\/\/arxiv.org\/abs\/1409.0473.","journal-title":"arXiv:1409.0473"},{"key":"e_1_3_2_10_2","unstructured":"Gino Brunner Yang Liu Damian Pascual Oliver Richter Massimiliano Ciaramita and Roger Wattenhofer. 2019. On identifiability in transformers. In International Conference on Learning Representations ."},{"key":"e_1_3_2_11_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1284"},{"key":"e_1_3_2_12_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-2003"},{"key":"e_1_3_2_13_2","first-page":"4349","volume-title":"Advances in Neural Information Processing Systems","author":"Bolukbasi Tolga","year":"2016","unstructured":"Tolga Bolukbasi, Kai-Wei Chang, James Y. Zou, Venkatesh Saligrama, and Adam T. Kalai. 2016. Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. In Advances in Neural Information Processing Systems. 4349\u20134357."},{"key":"e_1_3_2_14_2","unstructured":"Gino Brunner Yang Liu Dami\u00e1n Pascual Oliver Richter Massimiliano Ciaramita and Roger Wattenhofer. [n.d.]. On Identifiiability in Transformers."},{"key":"e_1_3_2_15_2","first-page":"1448","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Chang Shiyu","year":"2020","unstructured":"Shiyu Chang, Yang Zhang, Mo Yu, and Tommi Jaakkola. 2020. Invariant rationalization. In Proceedings of the International Conference on Machine Learning. PMLR, 1448\u20131458."},{"key":"e_1_3_2_16_2","first-page":"883","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Chen Jianbo","year":"2018","unstructured":"Jianbo Chen, Le Song, Martin Wainwright, and Michael Jordan. 2018. Learning to explain: An information-theoretic perspective on model interpretation. In Proceedings of the International Conference on Machine Learning. PMLR, 883\u2013892."},{"key":"e_1_3_2_17_2","article-title":"L-shapley and c-shapley: Efficient model interpretation for structured data","author":"Chen Jianbo","year":"2018","unstructured":"Jianbo Chen, Le Song, Martin J. Wainwright, and Michael I. Jordan. 2018. L-shapley and c-shapley: Efficient model interpretation for structured data. arXiv:1808.02610. Retrieved from https:\/\/arxiv.org\/abs\/1808.02610.","journal-title":"arXiv:1808.02610"},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1165"},{"key":"e_1_3_2_19_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2019.112948"},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W19-4828"},{"key":"e_1_3_2_21_2","doi-asserted-by":"publisher","DOI":"10.1007\/BF02310792"},{"key":"e_1_3_2_22_2","first-page":"6967","volume-title":"Advances in Neural Information Processing Systems","author":"Dabkowski Piotr","year":"2017","unstructured":"Piotr Dabkowski and Yarin Gal. 2017. Real time image saliency for black box classifiers. In Advances in Neural Information Processing Systems. 6967\u20136976."},{"key":"e_1_3_2_23_2","article-title":"A survey of the state of explainable AI for natural language processing","author":"Danilevsky Marina","year":"2020","unstructured":"Marina Danilevsky, Kun Qian, Ranit Aharonov, Yannis Katsis, Ban Kawas, and Prithviraj Sen. 2020. A survey of the state of explainable AI for natural language processing. arXiv:2010.00711. Retrieved from https:\/\/arxiv.org\/abs\/2010.00711.","journal-title":"arXiv:2010.00711"},{"key":"e_1_3_2_24_2","article-title":"Opportunities and challenges in explainable artificial intelligence (xai): A survey","author":"Das Arun","year":"2020","unstructured":"Arun Das and Paul Rad. 2020. Opportunities and challenges in explainable artificial intelligence (xai): A survey. arXiv:2006.11371. Retrieved from https:\/\/arxiv.org\/abs\/2006.11371.","journal-title":"arXiv:2006.11371"},{"key":"e_1_3_2_25_2","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Das Rajarshi","year":"2018","unstructured":"Rajarshi Das, Shehzaad Dhuliawala, Manzil Zaheer, Luke Vilnis, Ishan Durugkar, Akshay Krishnamurthy, Alex Smola, and Andrew McCallum. 2018. Go for a walk and arrive at the answer: Reasoning over paths in knowledge bases using reinforcement learning. In Proceedings of the International Conference on Learning Representations."},{"key":"e_1_3_2_26_2","first-page":"4171","volume-title":"Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)","author":"Devlin Jacob","year":"2019","unstructured":"Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 4171\u20134186."},{"key":"e_1_3_2_27_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.408"},{"key":"e_1_3_2_28_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P17-1106"},{"key":"e_1_3_2_29_2","article-title":"Towards a rigorous science of interpretable machine learning","author":"Doshi-Velez Finale","year":"2017","unstructured":"Finale Doshi-Velez and Been Kim. 2017. Towards a rigorous science of interpretable machine learning. arXiv:1702.08608. Retrieved from https:\/\/arxiv.org\/abs\/1702.08608.","journal-title":"arXiv:1702.08608"},{"key":"e_1_3_2_30_2","doi-asserted-by":"publisher","DOI":"10.23919\/MIPRO.2018.8400040"},{"key":"e_1_3_2_31_2","article-title":"Inverting convolutional networks with convolutional networks","author":"Dosovitskiy Alexey","year":"2015","unstructured":"Alexey Dosovitskiy and Thomas Brox. 2015. Inverting convolutional networks with convolutional networks. arXiv:1506.02753. Retrieved from https:\/\/arxiv.org\/abs\/1506.02753.","journal-title":"arXiv:1506.02753"},{"key":"e_1_3_2_32_2","article-title":"Analytical methods for interpretable ultradense word embeddings","author":"Dufter Philipp","year":"2019","unstructured":"Philipp Dufter and Hinrich Sch\u00fctze. 2019. Analytical methods for interpretable ultradense word embeddings. arXiv:1904.08654. Retrieved from https:\/\/arxiv.org\/abs\/1904.08654.","journal-title":"arXiv:1904.08654"},{"key":"e_1_3_2_33_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1354"},{"key":"e_1_3_2_34_2","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/N15-1184"},{"key":"e_1_3_2_35_2","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/P15-1144"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISIT.2004.1365067"},{"key":"e_1_3_2_37_2","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/D14-1044"},{"key":"e_1_3_2_38_2","doi-asserted-by":"publisher","DOI":"10.1109\/DSAA.2018.00018"},{"key":"e_1_3_2_39_2","volume-title":"Google Engineer Apologizes After Photos App Tags Two Black People as Gorillas","author":"Grush Loren","year":"2015","unstructured":"Loren Grush. 2015. Google Engineer Apologizes After Photos App Tags Two Black People as Gorillas. Retrieved September 28, 2020 from https:\/\/www.theverge.com\/2015\/7\/1\/8880363\/google-apologizes-photos-app-tags-two-black-people-gorillas."},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.1145\/3236009"},{"key":"e_1_3_2_41_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1108"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1024"},{"key":"e_1_3_2_43_2","article-title":"Neural module networks for reasoning over text","author":"Gupta Nitish","year":"2019","unstructured":"Nitish Gupta, Kevin Lin, Dan Roth, Sameer Singh, and Matt Gardner. 2019. Neural module networks for reasoning over text. arXiv:1912.04971. Retrieved from https:\/\/arxiv.org\/abs\/1812.04971.","journal-title":"arXiv:1912.04971"},{"key":"e_1_3_2_44_2","article-title":"Deep unfolding: Model-based inspiration of novel deep architectures","author":"Hershey John R.","year":"2014","unstructured":"John R. Hershey, Jonathan Le Roux, and Felix Weninger. 2014. Deep unfolding: Model-based inspiration of novel deep architectures. arXiv:1409.2574. Retrieved from https:\/\/arxiv.org\/abs\/1409.2574.","journal-title":"arXiv:1409.2574"},{"key":"e_1_3_2_45_2","article-title":"exbert: A visual analysis tool to explore learned representations in transformers models","author":"Hoover Benjamin","year":"2019","unstructured":"Benjamin Hoover, Hendrik Strobelt, and Sebastian Gehrmann. 2019. exbert: A visual analysis tool to explore learned representations in transformers models. arXiv:1910.05276. Retrieved from https:\/\/arxiv.org\/abs\/1910.05276.","journal-title":"arXiv:1910.05276"},{"key":"e_1_3_2_46_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2020.2967051"},{"key":"e_1_3_2_47_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.93"},{"key":"e_1_3_2_48_2","first-page":"3543","volume-title":"Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)","author":"Jain Sarthak","year":"2019","unstructured":"Sarthak Jain and Byron C. Wallace. 2019. Attention is not explanation. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 3543\u20133556."},{"key":"e_1_3_2_49_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1356"},{"key":"e_1_3_2_50_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM.2018.00135"},{"key":"e_1_3_2_51_2","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00324"},{"key":"e_1_3_2_52_2","doi-asserted-by":"publisher","DOI":"10.1162\/COLI_a_00300"},{"key":"e_1_3_2_53_2","article-title":"Visualizing and understanding recurrent networks","author":"Karpathy Andrej","year":"2015","unstructured":"Andrej Karpathy, Justin Johnson, and Li Fei-Fei. 2015. Visualizing and understanding recurrent networks. arXiv:1506.02078. Retrieved from https:\/\/arxiv.org\/abs\/1503.02078.","journal-title":"arXiv:1506.02078"},{"key":"e_1_3_2_54_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-1027"},{"key":"e_1_3_2_55_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D17-2021"},{"key":"e_1_3_2_56_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1011"},{"key":"e_1_3_2_57_2","article-title":"Explanation-based human debugging of NLP models: A survey","author":"Lertvittayakumjorn Piyawat","year":"2021","unstructured":"Piyawat Lertvittayakumjorn and Francesca Toni. 2021. Explanation-based human debugging of NLP models: A survey. arXiv:2104.15135. Retrieved from https:\/\/arxiv.org\/abs\/2104.15135.","journal-title":"arXiv:2104.15135"},{"key":"e_1_3_2_58_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N16-1082"},{"key":"e_1_3_2_59_2","article-title":"Understanding neural networks through representation erasure","author":"Li Jiwei","year":"2016","unstructured":"Jiwei Li, Will Monroe, and Dan Jurafsky. 2016. Understanding neural networks through representation erasure. arXiv:1612.08220. Retrieved from https:\/\/arxiv.org\/abs\/1312.08220.","journal-title":"arXiv:1612.08220"},{"key":"e_1_3_2_60_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.234"},{"key":"e_1_3_2_61_2","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00115"},{"key":"e_1_3_2_62_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1560"},{"key":"e_1_3_2_63_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-2007"},{"key":"e_1_3_2_64_2","first-page":"4765","volume-title":"Advances in Neural Information Processing Systems","author":"Lundberg Scott M.","year":"2017","unstructured":"Scott M. Lundberg and Su-In Lee. 2017. A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems. 4765\u20134774."},{"key":"e_1_3_2_65_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D15-1196"},{"key":"e_1_3_2_66_2","first-page":"2579","article-title":"Visualizing data using t-SNE","volume":"9","author":"Maaten Laurens van der","year":"2008","unstructured":"Laurens van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-SNE. J. Mach. Learn. Res. 9 (November 2008), 2579\u20132605.","journal-title":"J. Mach. Learn. Res."},{"key":"e_1_3_2_67_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7299155"},{"key":"e_1_3_2_68_2","article-title":"Efficient estimation of word representations in vector space","author":"Mikolov Tomas","year":"2013","unstructured":"Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. 2013. Efficient estimation of word representations in vector space. arXiv:1301.3781. Retrieved from https:\/\/arxiv.org\/abs\/1301.3781.","journal-title":"arXiv:1301.3781"},{"key":"e_1_3_2_69_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.artint.2018.07.007"},{"key":"e_1_3_2_70_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.387"},{"key":"e_1_3_2_71_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.dsp.2017.10.011"},{"key":"e_1_3_2_72_2","volume-title":"Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP\u201996)","author":"Mooney Raymond","year":"1996","unstructured":"Raymond Mooney. 1996. Comparative experiments on disambiguating word senses: An illustration of the role of bias in machine learning. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP\u201996)."},{"key":"e_1_3_2_73_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1100"},{"key":"e_1_3_2_74_2","first-page":"1933","volume-title":"Proceedings of the International Conference on Computational Linguistics (COLING\u201912)","author":"Murphy Brian","year":"2012","unstructured":"Brian Murphy, Partha Talukdar, and Tom Mitchell. 2012. Learning effective and interpretable semantic models using non-negative sparse embedding. In Proceedings of the International Conference on Computational Linguistics (COLING\u201912). 1933\u20131950."},{"key":"e_1_3_2_75_2","doi-asserted-by":"crossref","first-page":"55","DOI":"10.1007\/978-3-030-28954-6_4","volume-title":"Explainable AI: Interpreting, Explaining and Visualizing Deep Learning","author":"Nguyen Anh","year":"2019","unstructured":"Anh Nguyen, Jason Yosinski, and Jeff Clune. 2019. Understanding neural networks via feature visualization: A survey. In Explainable AI: Interpreting, Explaining and Visualizing Deep Learning. Springer, 55\u201376."},{"key":"e_1_3_2_76_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1570"},{"key":"e_1_3_2_77_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D17-1041"},{"key":"e_1_3_2_78_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W16-1620"},{"key":"e_1_3_2_79_2","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00049"},{"key":"e_1_3_2_80_2","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/D14-1162"},{"key":"e_1_3_2_81_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-1179"},{"key":"e_1_3_2_82_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N18-1202"},{"key":"e_1_3_2_83_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1250"},{"key":"e_1_3_2_84_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.432"},{"key":"e_1_3_2_85_2","article-title":"Explainable reinforcement learning: A survey","author":"Puiutta Erika","year":"2020","unstructured":"Erika Puiutta and Eric Veith. 2020. Explainable reinforcement learning: A survey. arXiv:2005.06247. Retrieved from https:\/\/arxiv.org\/abs\/2005.06247.","journal-title":"arXiv:2005.06247"},{"key":"e_1_3_2_86_2","doi-asserted-by":"publisher","DOI":"10.3934\/mfc.2018008"},{"issue":"8","key":"e_1_3_2_87_2","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford Alec","year":"2019","unstructured":"Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019. Language models are unsupervised multitask learners. OpenAI Blog 1, 8 (2019), 9.","journal-title":"OpenAI Blog"},{"key":"e_1_3_2_88_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5431"},{"key":"e_1_3_2_89_2","doi-asserted-by":"publisher","DOI":"10.1145\/2939672.2939778"},{"key":"e_1_3_2_90_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.11491"},{"key":"e_1_3_2_91_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.437"},{"key":"e_1_3_2_92_2","article-title":"Reasoning about entailment with neural attention","author":"Rockt\u00e4schel Tim","year":"2015","unstructured":"Tim Rockt\u00e4schel, Edward Grefenstette, Karl Moritz Hermann, Tom\u00e1\u0161 Ko\u010disk\u1ef3, and Phil Blunsom. 2015. Reasoning about entailment with neural attention. arXiv:1509.06664. Retrieved from https:\/\/arxiv.org\/abs\/1509.06664.","journal-title":"arXiv:1509.06664"},{"key":"e_1_3_2_93_2","first-page":"2690","volume-title":"Proceedings of the 27th International Conference on Computational Linguistics","author":"Rogers Anna","year":"2018","unstructured":"Anna Rogers, Shashwath Hosur Ananthakrishna, and Anna Rumshisky. 2018. What\u2019s in your embedding, and how it predicts task performance. In Proceedings of the 27th International Conference on Computational Linguistics. 2690\u20132703."},{"key":"e_1_3_2_94_2","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00349"},{"key":"e_1_3_2_95_2","article-title":"Explaining nlp models via minimal contrastive editing (mice)","author":"Ross Alexis","year":"2020","unstructured":"Alexis Ross, Ana Marasovi\u0107, and Matthew E. Peters. 2020. Explaining nlp models via minimal contrastive editing (mice). arXiv:2012.13985. Retrieved from https:\/\/arxiv.org\/abs\/2012.13985.","journal-title":"arXiv:2012.13985"},{"key":"e_1_3_2_96_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P16-2083"},{"key":"e_1_3_2_97_2","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-019-0048-x"},{"key":"e_1_3_2_98_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33013027"},{"key":"e_1_3_2_99_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.74"},{"key":"e_1_3_2_100_2","doi-asserted-by":"publisher","DOI":"10.5555\/3305890.3306006"},{"key":"e_1_3_2_101_2","doi-asserted-by":"publisher","DOI":"10.1038\/nature16961"},{"key":"e_1_3_2_102_2","doi-asserted-by":"publisher","DOI":"10.1109\/TVCG.2018.2865044"},{"key":"e_1_3_2_103_2","article-title":"Energy and policy considerations for deep learning in NLP","author":"Strubell Emma","year":"2019","unstructured":"Emma Strubell, Ananya Ganesh, and Andrew McCallum. 2019. Energy and policy considerations for deep learning in NLP. arXiv:1906.02243. Retrieved from https:\/\/arxiv.org\/abs\/1906.02243.","journal-title":"arXiv:1906.02243"},{"key":"e_1_3_2_104_2","doi-asserted-by":"publisher","DOI":"10.5555\/1756006.1756007"},{"key":"e_1_3_2_105_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.11935"},{"key":"e_1_3_2_106_2","first-page":"2915","volume-title":"Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI\u201916)","author":"Sun Fei","year":"2016","unstructured":"Fei Sun, Jiafeng Guo, Yanyan Lan, Jun Xu, and Xueqi Cheng. 2016. Sparse word embeddings using l1 regularized online learning. In Proceedings of the 25th International Joint Conference on Artificial Intelligence (IJCAI\u201916). AAAI Press, 2915\u20132921."},{"key":"e_1_3_2_107_2","doi-asserted-by":"publisher","DOI":"10.3115\/v1\/P15-1150"},{"key":"e_1_3_2_108_2","doi-asserted-by":"publisher","DOI":"10.17261\/Pressacademia.2017.591"},{"key":"e_1_3_2_109_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1452"},{"key":"e_1_3_2_110_2","article-title":"What do you learn from context? Probing for sentence structure in contextualized word representations","author":"Tenney Ian","year":"2019","unstructured":"Ian Tenney, Patrick Xia, Berlin Chen, Alex Wang, Adam Poliak, R. Thomas McCoy, Najoung Kim, Benjamin Van Durme, Samuel R. Bowman, Dipanjan Das, et\u00a0al. 2019. What do you learn from context? Probing for sentence structure in contextualized word representations. arXiv:1905.06316. Retrieved from https:\/\/arxiv.org\/abs\/1905.06313.","journal-title":"arXiv:1905.06316"},{"key":"e_1_3_2_111_2","unstructured":"Erico Tjoa and Cuntai Guan. 2019. A survey on explainable artificial intelligence (XAI): Towards medical XAI. CoRR abs\/1907.07374 (2019)."},{"key":"e_1_3_2_112_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-1503"},{"key":"e_1_3_2_113_2","article-title":"Attention interpretability across nlp tasks","author":"Vashishth Shikhar","year":"2019","unstructured":"Shikhar Vashishth, Shyam Upadhyay, Gaurav Singh Tomar, and Manaal Faruqui. 2019. Attention interpretability across nlp tasks. arXiv:1909.11218. Retrieved from https:\/\/arxiv.org\/abs\/1909.11218.","journal-title":"arXiv:1909.11218"},{"key":"e_1_3_2_114_2","first-page":"193","volume-title":"Proceedings of the 13th Conference of the Association for Machine Translation in the Americas (Volume 1: Research Track)","author":"Vaswani Ashish","year":"2018","unstructured":"Ashish Vaswani, Samy Bengio, Eugene Brevdo, Francois Chollet, Aidan Gomez, Stephan Gouws, Llion Jones, \u0141ukasz Kaiser, Nal Kalchbrenner, Niki Parmar, et\u00a0al. 2018. Tensor2Tensor for neural machine translation. In Proceedings of the 13th Conference of the Association for Machine Translation in the Americas (Volume 1: Research Track). 193\u2013199."},{"key":"e_1_3_2_115_2","first-page":"5998","volume-title":"Advances in Neural Information Processing Systems","author":"Vaswani Ashish","year":"2017","unstructured":"Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, \u0141ukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in Neural Information Processing Systems. 5998\u20136008."},{"key":"e_1_3_2_116_2","article-title":"Visualizing attention in transformer-based language representation models","author":"Vig Jesse","year":"2019","unstructured":"Jesse Vig. 2019. Visualizing attention in transformer-based language representation models. arXiv:1904.02679. Retrieved from https:\/\/arxiv.org\/abs\/1904.02679.","journal-title":"arXiv:1904.02679"},{"key":"e_1_3_2_117_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W19-4808"},{"key":"e_1_3_2_118_2","article-title":"Commonsense knowledge graph reasoning by selection or generation? Why?","author":"Wang Cunxiang","year":"2020","unstructured":"Cunxiang Wang, Jinhang Wu, Luxin Liu, and Yue Zhang. 2020. Commonsense knowledge graph reasoning by selection or generation? Why? arXiv:2008.05925. Retrieved from https:\/\/arxiv.org\/abs\/2008.05925.","journal-title":"arXiv:2008.05925"},{"key":"e_1_3_2_119_2","article-title":"Interpretable recurrent neural networks using sequential sparse recovery","author":"Wisdom Scott","year":"2016","unstructured":"Scott Wisdom, Thomas Powers, James Pitton, and Les Atlas. 2016. Interpretable recurrent neural networks using sequential sparse recovery. arXiv:1611.07252. Retrieved from https:\/\/arxiv.org\/abs\/1611.07252.","journal-title":"arXiv:1611.07252"},{"key":"e_1_3_2_120_2","article-title":"Google\u2019s neural machine translation system: Bridging the gap between human and machine translation","author":"Wu Yonghui","year":"2016","unstructured":"Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V. Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, et\u00a0al. 2016. Google\u2019s neural machine translation system: Bridging the gap between human and machine translation. arXiv:1609.08144. Retrieved from https:\/\/arxiv.org\/abs\/1609.08144.","journal-title":"arXiv:1609.08144"},{"key":"e_1_3_2_121_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D17-1060"},{"key":"e_1_3_2_122_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46478-7_28"},{"key":"e_1_3_2_123_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.10"},{"key":"e_1_3_2_124_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1420"},{"key":"e_1_3_2_125_2","first-page":"260","volume-title":"Proceedings of the Human Language Technologies: The Conference of the North American Chapter of the Association for Computational Linguistics","author":"Zaidan Omar","year":"2007","unstructured":"Omar Zaidan, Jason Eisner, and Christine Piatko. 2007. Using \u201cannotator rationales\u201d to improve machine learning for text categorization. In Proceedings of the Human Language Technologies: The Conference of the North American Chapter of the Association for Computational Linguistics. 260\u2013267."},{"key":"e_1_3_2_126_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D16-1076"},{"key":"e_1_3_2_127_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-1144"},{"key":"e_1_3_2_128_2","doi-asserted-by":"crossref","unstructured":"Jieyu Zhao Tianlu Wang Mark Yatskar Ryan Cotterell Vicente Ordonez and Kai-Wei Chang. 2019. Gender bias in contextualized word embeddings. arxiv:1904.03310 [cs.CL]. Retrieved from https:\/\/arxiv.org\/abs\/1904.03310.","DOI":"10.18653\/v1\/N19-1064"},{"key":"e_1_3_2_129_2","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-1521"},{"key":"e_1_3_2_130_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.aiopen.2021.01.001"}],"container-title":["ACM Computing Surveys"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3529755","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3529755","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,17]],"date-time":"2025-06-17T19:31:24Z","timestamp":1750188684000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3529755"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,12,3]]},"references-count":129,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2023,5,31]]}},"alternative-id":["10.1145\/3529755"],"URL":"https:\/\/doi.org\/10.1145\/3529755","relation":{},"ISSN":["0360-0300","1557-7341"],"issn-type":[{"value":"0360-0300","type":"print"},{"value":"1557-7341","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,12,3]]},"assertion":[{"value":"2021-04-02","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-03-30","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-12-03","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}