Filters








9,384 Hits in 6.3 sec

New Protocols and Negative Results for Textual Entailment Data Collection [article]

Samuel R. Bowman, Jennimaria Palomaki, Livio Baldini Soares, Emily Pitler
<span title="2020-09-29">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Using these alternatives and a fifth baseline protocol, we collect and compare five new 8.5k-example training sets.  ...  However, the crowdsourcing protocol that was used to collect this data has known issues and was not explicitly optimized for either of these purposes, so it is likely far from ideal.  ...  Acknowledgments We thank the annotators who spent time and effort on this project and the many members of the natural language processing community at Google who provided feedback.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.11997v2">arXiv:2004.11997v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/xfhgsljflfd2jc3ioabbshdcri">fatcat:xfhgsljflfd2jc3ioabbshdcri</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201005072834/https://arxiv.org/pdf/2004.11997v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.11997v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

New Protocols and Negative Results for Textual Entailment Data Collection

Samuel R. Bowman, Jennimaria Palomaki, Livio Baldini Soares, Emily Pitler
<span title="">2020</span> <i title="Association for Computational Linguistics"> Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) </i> &nbsp; <span class="release-stage">unpublished</span>
Using these alternatives and a fifth baseline protocol, we collect and compare five new 8.5k-example training sets.  ...  However, the crowdsourcing protocol that was used to collect this data has known issues and was not explicitly optimized for either of these purposes, so it is likely far from ideal.  ...  Acknowledgments We thank the annotators who spent time and effort on this project and the many members of the natural language processing community at Google who provided feedback.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/2020.emnlp-main.658">doi:10.18653/v1/2020.emnlp-main.658</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4rgfrpaiovgstjf5hffj3ngk7a">fatcat:4rgfrpaiovgstjf5hffj3ngk7a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201211160408/https://www.aclweb.org/anthology/2020.emnlp-main.658.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ca/c1/cac1849c8f136903b19ec19fa1f83b631b0ddfcc.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/2020.emnlp-main.658"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Asking Crowdworkers to Write Entailment Examples: The Best of Bad Options [article]

Clara Vania, Ruijie Chen, Samuel R. Bowman
<span title="2020-10-13">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We conclude that crowdworker writing still the best known option for entailment data, highlighting the need for further data collection work to focus on improving writing-based annotation processes.  ...  Our experiments on NLI and transfer learning show negative results: None of the alternative protocols outperforms the baseline in evaluations of generalization within NLI or on transfer to outside target  ...  Deep Learning using Latent Structure), by Intuit, Inc., and in-kind support by the NYU High-Performance Computing Center and by NVIDIA Corporation (with the donation of a Titan V GPU).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.06122v1">arXiv:2010.06122v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/kr6ccdzwivdi5n3mfwlmrb3f2u">fatcat:kr6ccdzwivdi5n3mfwlmrb3f2u</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201018023611/https://arxiv.org/pdf/2010.06122v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/82/94/82944c454086b0c18a6df076dd60c4789e70ae8c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.06122v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Don't 'have a clue'? Unsupervised co-learning of downward-entailing operators [article]

Cristian Danescu-Niculescu-Mizil, Lillian Lee
<span title="2010-11-27">2010</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Recent work proposed a method for learning English downward-entailing operators that requires access to a high-quality collection of 'negative polarity items' (NPIs).  ...  Researchers in textual entailment have begun to consider inferences involving 'downward-entailing operators', an interesting and important class of lexical items that change the way inferences are made  ...  Paula Muchado, Stephen Purpura, Mark Yatskar, Ainur Yessenalina, and the anonymous reviewers for their helpful comments.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1008.3169v2">arXiv:1008.3169v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/hratrudug5edfenksfwhg7o67q">fatcat:hratrudug5edfenksfwhg7o67q</a> </span>
<a target="_blank" rel="noopener" href="https://archive.org/download/arxiv-1008.3169/1008.3169.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> File Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/75/47/7547b903ef3a103022021b7802f55c93899420b4.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1008.3169v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Ordinal Common-sense Inference [article]

Sheng Zhang, Rachel Rudinger, Kevin Duh, Benjamin Van Durme
<span title="2017-06-02">2017</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We describe a framework for extracting common-sense knowledge from corpora, which is then used to construct a dataset for this ordinal entailment task.  ...  We propose an evaluation of automated common-sense inference based on an extension of recognizing textual entailment: predicting ordinal human responses on the subjective likelihood of an inference holding  ...  Acknowledgments Thank you to action editor Mark Steedman and the anonymous reviewers for their feedback, as well as colleagues including Lenhart Schubert, Kyle Rawlins, Aaron White, and Keisuke Sakaguchi  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1611.00601v3">arXiv:1611.00601v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/kezt3fidsvgkxk5riypb3o7bfi">fatcat:kezt3fidsvgkxk5riypb3o7bfi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200824200700/https://arxiv.org/pdf/1611.00601v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/c2/bd/c2bd37f67b88a89b4183a7fafb6d0ab830f2a078.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1611.00601v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Analyzing Dynamic Adversarial Training Data in the Limit [article]

Eric Wallace, Adina Williams, Robin Jia, Douwe Kiela
<span title="2021-10-16">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We present the first study of longer-term DADC, where we collect 20 rounds of NLI examples for a small set of premise paragraphs, with both adversarial and non-adversarial approaches.  ...  Dynamic adversarial data collection (DADC), where annotators craft examples that challenge continually improving models, holds promise as an approach for generating such diverse training sets.  ...  We also thank Alicia Parrish, Nicholas Tomlin, Max Bartolo, Jessy Lin, and Nelson Liu for help annotating our test set.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.08514v1">arXiv:2110.08514v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/rclybepweneyjdbem2cvpmkvxi">fatcat:rclybepweneyjdbem2cvpmkvxi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211021043513/https://arxiv.org/pdf/2110.08514v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/80/5a/805af32422768436b5e3d9a4dea1f8b0811280e7.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.08514v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Evaluating the Factual Consistency of Abstractive Text Summarization [article]

Wojciech Kryściński, Bryan McCann, Caiming Xiong, Richard Socher
<span title="2019-10-28">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We propose a weakly-supervised, model-based approach for verifying factual consistency and identifying conflicts between source documents and a generated summary.  ...  Training data is generated by applying a series of rule-based transformations to the sentences of source documents.  ...  Acknowledgements We thank Nitish Shirish Keskar, Dragomir Radev, Ben Krause, and Wenpeng Yin for reviewing this manuscript and providing valuable feedback.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1910.12840v1">arXiv:1910.12840v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/q5l6ih2pz5gtngy2ujfficc7ly">fatcat:q5l6ih2pz5gtngy2ujfficc7ly</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200824170101/https://arxiv.org/pdf/1910.12840v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/0c/55/0c5598424cc96d8fb500eb553cb7969f86a0ede0.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1910.12840v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Pairwise Supervised Contrastive Learning of Sentence Representations [article]

Dejiao Zhang, Shang-Wen Li, Wei Xiao, Henghui Zhu, Ramesh Nallapati, Andrew O. Arnold, Bing Xiang
<span title="2022-01-29">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Therefore, optimizing the semantic entailment and contradiction reasoning objective alone is inadequate to capture the high-level semantic structure.  ...  We outperform the previous state-of-the-art method with 10%–13% averaged improvement on eight clustering tasks, and 5%–6% averaged improvement on seven semantic textual similarity (STS) tasks.  ...  More recently, SBERT (Reimers and Gurevych, 2019b) finetunes a siamese BERT (Devlin et al., 2018) model on NLI and sets new state-of-the-art results.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2109.05424v2">arXiv:2109.05424v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/niqqqqwmjnaz3noadxn7mdwvbq">fatcat:niqqqqwmjnaz3noadxn7mdwvbq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220202223143/https://arxiv.org/pdf/2109.05424v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/29/6a/296aaba05c9a790f48b501d14f82c562d202c2e5.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2109.05424v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Climate change opinions in online debate sites

Adrian Groza, Pinar Ozturk, Radu Razvan-Slavescu, Anca Marginean
<span title="">2019</span> <i title="National Library of Serbia"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/rhtuh2ifczhapmhplqzald63za" style="color: black;">Computer Science and Information Systems</a> </i> &nbsp;
For the second objective, we assess the semantic similarity between two debate topics based on textual entailment [28] .  ...  We focus here on developing a technical instrumentation for making sense of a set of online arguments and aggregating them into usable results for policy making and climate science communication.  ...  modeling" between Norwegian University of Science and Technology, Trondheim, Norway and Technical University of Cluj-Napoca, Romania.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.2298/csis180601015g">doi:10.2298/csis180601015g</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ynl6dgltirfm7eqtl57t5npybu">fatcat:ynl6dgltirfm7eqtl57t5npybu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200505044404/http://www.doiserbia.nb.rs/img/doi/1820-0214/2020/1820-02141900015G.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/dd/62/dd62827fe1f2976b210cdcde848ee29c93374652.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.2298/csis180601015g"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> Publisher / doi.org </button> </a>

A New Dataset for Natural Language Inference from Code-mixed Conversations [article]

Simran Khanuja, Sandipan Dandapat, Sunayana Sitaram, Monojit Choudhury
<span title="2020-04-13">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We evaluate the dataset using a standard mBERT-based pipeline for NLI and report results.  ...  Currently, the data collected consists of 400 premises in the form of code-mixed conversation snippets and 2240 code-mixed hypotheses.  ...  collection of 433k sentence pairs annotated with entailment information.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.05051v2">arXiv:2004.05051v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/gwic2jrs3zhqnbpawjwhd6xhrq">fatcat:gwic2jrs3zhqnbpawjwhd6xhrq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200415022055/https://arxiv.org/pdf/2004.05051v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.05051v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Annotation Artifacts in Natural Language Inference Data [article]

Suchin Gururangan, Swabha Swayamdipta, Omer Levy, Roy Schwartz, Samuel R. Bowman, Noah A. Smith
<span title="2018-04-16">2018</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Large-scale datasets for natural language inference are created by presenting crowd workers with a sentence (premise), and asking them to generate three new sentences (hypotheses) that it entails, contradicts  ...  We show that, in a significant portion of such data, this protocol leaves clues that make it possible to identify the label by looking only at the hypothesis, without observing the premise.  ...  SB acknowledges gift support from Google and Tencent Holdings and support from Samsung Research.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1803.02324v2">arXiv:1803.02324v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/3o57ixhbffgljnlbfhybydco2i">fatcat:3o57ixhbffgljnlbfhybydco2i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20191024084137/https://arxiv.org/pdf/1803.02324v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/a3/31/a33134b82ccfd22043451d3db2735b0218d43ff7.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1803.02324v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Annotation Artifacts in Natural Language Inference Data

Suchin Gururangan, Swabha Swayamdipta, Omer Levy, Roy Schwartz, Samuel Bowman, Noah A. Smith
<span title="">2018</span> <i title="Association for Computational Linguistics"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/d5ex6ucxtrfz3clshlkh3f6w2q" style="color: black;">Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers)</a> </i> &nbsp;
Large-scale datasets for natural language inference are created by presenting crowd workers with a sentence (premise), and asking them to generate three new sentences (hypotheses) that it entails, contradicts  ...  We show that, in a significant portion of such data, this protocol leaves clues that make it possible to identify the label by looking only at the hypothesis, without observing the premise.  ...  SB acknowledges gift support from Google and Tencent Holdings and support from Samsung Research.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/n18-2017">doi:10.18653/v1/n18-2017</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/naacl/GururanganSLSBS18.html">dblp:conf/naacl/GururanganSLSBS18</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/up67r2jwmnawrdsnigw3nuc6h4">fatcat:up67r2jwmnawrdsnigw3nuc6h4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200309092649/https://www.aclweb.org/anthology/N18-2017.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/48/3e/483e19f50ff47b0bf5e57b0cea65a7f084779b92.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.18653/v1/n18-2017"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Classifying Argumentative Relations Using Logical Mechanisms and Argumentation Schemes

Yohan Jo, Seojin Bang, Chris Reed, Eduard Hovy
<span title="">2021</span> <i title="MIT Press - Journals"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/jtrizn7izbhdlfmye6eriw6ub4" style="color: black;">Transactions of the Association for Computational Linguistics</a> </i> &nbsp;
, (iii) causal relation, and (iv) normative relation.  ...  We demonstrate that our operationalization of these logical mechanisms classifies argumentative relations without directly training on data labeled with the relations, significantly better than several  ...  Acknowledgments We thank the reviewers and action editor for valuable comments.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1162/tacl_a_00394">doi:10.1162/tacl_a_00394</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/csyhtbvq3bhslbhk5n5zkgg4mm">fatcat:csyhtbvq3bhslbhk5n5zkgg4mm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210922011723/https://watermark.silverchair.com/tacl_a_00394.pdf?token=AQECAHi208BE49Ooan9kkhW_Ercy7Dm3ZL_9Cf3qfKAc485ysgAAAs8wggLLBgkqhkiG9w0BBwagggK8MIICuAIBADCCArEGCSqGSIb3DQEHATAeBglghkgBZQMEAS4wEQQMAS28Zr4hETREKHPOAgEQgIICggpr3PcBGHDyQWRsgs__bLybMgs15T-XK7Niac7Pz4BPbo_Wfs-jV4-nw2wR1Sda5UNTc9IbQFtddITJnUwiY5iyYhqp-vhBch9I98RcEf88KYxjSCYQEmhkCnE3x-wkkfuyslArKdgqykMzXoSayGIULMg5eB9aMH0KbI4zfZ03vxzNEbumfX0yqtPiyFphQGBPV74QQh3hdupRrTcSxtAJ3VvVH1X6Nid0ji-493Chv4uGboMpJHQ6S-hV98vY_BvmzRBZIjscNIyhIRUYfvQDhPMhIbtNN7PDlZqQaxftWPU-gopKyyhWYCd-Tvaye8mWMQzeHhf-WPkLpRSorcYd-J-8GqCmULFuPDf_LJ6oRa5Zb8wqR87Y1m6HQkZMPRZXTUSvretWN2YE9E9GgtmVF0fnlkVeUvdzuXvK7H2-rRgGjih5btSvBBtu0Uhna4279js6R1S9uLXHlE73YtjIwkv-r6f86fabDKpDqDCX6WlrBnVjBxI__lhxI0Xd2EGwqALchjCMf4ak82asPrYFpxbnKUHUXSpP0PAjuF0X3DPZjzuhZevpGBAhgJSJVqDmmqopcSgfia0M6RzsjdIkEujx4qi2evsKk6R__k_9w19AiQxX-0KVEePWG93PBoCWzpIr7BPJXxpc-RQVQVpqcPs4RyZ4Vfw2LmYVficzi-H38GtdyCtR4-n1fMQQIUtwVLeFY7QmXzhRxtZ1SaKW-TyibOTk174EPOqnInlol2GjrPwS9Vc2OGfX2xdmCyNhx5DaFgQCQ30QJnYgKALBtTr_PsK3g-7GhTsoSzYbwd-13XWqxNqKpDL1Q_W0ggErvnBv7nvEl2Y3JqGkw3GdjA" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/f1/35/f135cc0bb4fe8bb39d50146460e28261edca4777.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1162/tacl_a_00394"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> mitpressjournals.org </button> </a>

VIOLIN: A Large-Scale Dataset for Video-and-Language Inference [article]

Jingzhou Liu, Wenhu Chen, Yu Cheng, Zhe Gan, Licheng Yu, Yiming Yang, Jingjing Liu
<span title="2020-03-25">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We introduce a new task, Video-and-Language Inference, for joint multimodal understanding of video and text.  ...  A new large-scale dataset, named Violin (VIdeO-and-Language INference), is introduced for this task, which consists of 95,322 video-hypothesis pairs from 15,887 video clips, spanning over 582 hours of  ...  Acknowledgement We would like to thank Yandong Li, Liqun Chen, Shuyang Dai, Linjie Li, Chen Zhu, Jiacheng Xu and Boyi Li for providing useful feedback on the project and their help in collecting and annotating  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2003.11618v1">arXiv:2003.11618v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/zs5m6fonynh27ogqjarxtwrqie">fatcat:zs5m6fonynh27ogqjarxtwrqie</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200328020709/https://arxiv.org/pdf/2003.11618v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/70/4e/704ec27b8399df574a96da338c428a923509385e.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2003.11618v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models [article]

Boxin Wang, Chejian Xu, Shuohang Wang, Zhe Gan, Yu Cheng, Jianfeng Gao, Ahmed Hassan Awadallah, Bo Li
<span title="2022-01-10">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In particular, we systematically apply 14 textual adversarial attack methods to GLUE tasks to construct AdvGLUE, which is further validated by humans for reliable annotations.  ...  We hope our work will motivate the development of new adversarial attacks that are more stealthy and semantic-preserving, as well as new robust language models against sophisticated adversarial attacks  ...  Adina Williams, Nikita Nangia, Jinfeng Li, and many others for the helpful discussion. We thank Prof. Robin Jia and Yixin Nie for allowing us to incorporate their datasets as part of the evaluation.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.02840v2">arXiv:2111.02840v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/3wzxt4cdrjettcmsajvp52skoe">fatcat:3wzxt4cdrjettcmsajvp52skoe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220112001611/https://arxiv.org/pdf/2111.02840v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/b1/32/b1324a2a1ebc66b8d8a7ae3ff35a8e84319f06e0.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.02840v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 9,384 results