%0 Conference Paper %1 conf/naacl/GulordavaBGLB18 %A Gulordava, Kristina %A Bojanowski, Piotr %A Grave, Edouard %A Linzen, Tal %A Baroni, Marco %B NAACL-HLT %D 2018 %E Walker, Marilyn A. %E Ji, Heng %E Stent, Amanda %I Association for Computational Linguistics %K deep_learning nlp rnn %P 1195-1205 %T Colorless Green Recurrent Networks Dream Hierarchically. %U http://dblp.uni-trier.de/db/conf/naacl/naacl2018-1.html#GulordavaBGLB18 %X Recurrent neural networks (RNNs) have achieved impressive results in a variety of lin- guistic processing tasks, suggesting that they can induce non-trivial properties of language. We investigate here to what extent RNNs learn to track abstract hierarchical syntactic struc- ture. We test whether RNNs trained with a generic language modeling objective in four languages (Italian, English, Hebrew, Russian) can predict long-distance number agreement in various constructions. We include in our evaluation nonsensical sentences where RNNs cannot rely on semantic or lexical cues (“The colorless green ideas ideas I ate with the chair sleep sleep furiously”), and, for Italian, we com- pare model performance to human intuitions. Our language-model-trained RNNs make re- liable predictions about long-distance agree- ment, and do not lag much behind human performance. We thus bring support to the hypothesis that RNNs are not just shallow- pattern extractors, but they also acquire deeper grammatical competence. %@ 978-1-948087-27-8

@inproceedings{conf/naacl/GulordavaBGLB18, abstract = {Recurrent neural networks (RNNs) have achieved impressive results in a variety of lin- guistic processing tasks, suggesting that they can induce non-trivial properties of language. We investigate here to what extent RNNs learn to track abstract hierarchical syntactic struc- ture. We test whether RNNs trained with a generic language modeling objective in four languages (Italian, English, Hebrew, Russian) can predict long-distance number agreement in various constructions. We include in our evaluation nonsensical sentences where RNNs cannot rely on semantic or lexical cues (“The colorless green ideas ideas I ate with the chair sleep sleep furiously”), and, for Italian, we com- pare model performance to human intuitions. Our language-model-trained RNNs make re- liable predictions about long-distance agree- ment, and do not lag much behind human performance. We thus bring support to the hypothesis that RNNs are not just shallow- pattern extractors, but they also acquire deeper grammatical competence.}, added-at = {2018-11-07T10:41:05.000+0100}, author = {Gulordava, Kristina and Bojanowski, Piotr and Grave, Edouard and Linzen, Tal and Baroni, Marco}, biburl = {https://www.bibsonomy.org/bibtex/237fe1a115bda87566917006c3969dce9/dallmann}, booktitle = {NAACL-HLT}, crossref = {conf/naacl/2018-1}, editor = {Walker, Marilyn A. and Ji, Heng and Stent, Amanda}, ee = {https://aclanthology.info/papers/N18-1108/n18-1108}, interhash = {5ca571e6d6d3ead85fe1fe484fcab6b8}, intrahash = {37fe1a115bda87566917006c3969dce9}, isbn = {978-1-948087-27-8}, keywords = {deep_learning nlp rnn}, pages = {1195-1205}, publisher = {Association for Computational Linguistics}, timestamp = {2018-11-07T10:41:05.000+0100}, title = {Colorless Green Recurrent Networks Dream Hierarchically.}, url = {http://dblp.uni-trier.de/db/conf/naacl/naacl2018-1.html#GulordavaBGLB18}, year = 2018 }