{"id":"https://openalex.org/W3207710400","doi":"https://doi.org/10.18653/v1/2022.findings-naacl.130","title":"Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models","display_name":"Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models","alternate_titles":[],"publication_year":2022,"publication_date":"2022-01-01","ids":{"openalex":"https://openalex.org/W3207710400","doi":"https://doi.org/10.18653/v1/2022.findings-naacl.130","mag":"3207710400"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2022.findings-naacl.130","doi":"https://doi.org/10.18653/v1/2022.findings-naacl.130","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.findings-naacl.130","pdf_url":"https://aclanthology.org/2022.findings-naacl.130.pdf","source":{"id":"https://openalex.org/S4363605604","display_name":"Findings of the Association for Computational Linguistics: NAACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: NAACL 2022","raw_type":"proceedings-article"},"parent_works":[],"lineage":["https://openalex.org/W3207710400"],"child_works_count":0,"type":"conference-paper","is_traditional":true,"indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2022.findings-naacl.130.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5026227210","display_name":"Tianlu Wang","orcid":"https://orcid.org/0000-0002-7480-9705","observed_orcids":["https://orcid.org/0000-0002-7480-9705"]},"institutions":[{"id":"https://openalex.org/I51556381","display_name":"University of Virginia","ror":"https://ror.org/0153tk833","country_code":"US","type":"education","lineage":["https://openalex.org/I51556381"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Tianlu Wang","raw_affiliation_strings":["University of Virginia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Virginia","institution_ids":["https://openalex.org/I51556381"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002296599","display_name":"Rohit Sridhar","orcid":"https://orcid.org/0000-0001-5239-5698","observed_orcids":["https://orcid.org/0000-0001-5239-5698"]},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Rohit Sridhar","raw_affiliation_strings":["Georgia Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089413311","display_name":"Diyi Yang","orcid":"https://orcid.org/0000-0003-1220-3983","observed_orcids":["https://orcid.org/0000-0003-1220-3983"]},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Diyi Yang","raw_affiliation_strings":["Georgia Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5040824554","display_name":"Xuezhi Wang","orcid":"https://orcid.org/0000-0001-5222-248X","observed_orcids":["https://orcid.org/0000-0001-5222-248X"]},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Xuezhi Wang","raw_affiliation_strings":["Google Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Research","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":6.7075,"has_fulltext":true,"cited_by_count":116,"citation_normalized_percentile":{"value":0.9788569,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"1719","last_page":"1729"},"is_retracted":false,"is_pulled":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.7807999849319458,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.7807999849319458,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.19979999959468842,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.008700000122189522,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/spurious-correlations","display_name":"spurious correlations","score":0.9950000047683716},{"id":"https://openalex.org/keywords/shortcut-learning","display_name":"shortcut learning","score":0.6940000057220459},{"id":"https://openalex.org/keywords/robustness","display_name":"robustness","score":0.8029999732971191},{"id":"https://openalex.org/keywords/interpretability","display_name":"interpretability","score":0.5929999947547913},{"id":"https://openalex.org/keywords/nlp-models","display_name":"NLP models","score":0.5709999799728394}],"concepts":[{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.9279812574386597},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.9275155067443848},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8254009485244751},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.7088618874549866},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.6989611387252808},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6485350131988525},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6092907786369324},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6050383448600769},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.4469892382621765},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.4129774570465088},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.33771464228630066},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.3334546983242035},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2022.findings-naacl.130","doi":"https://doi.org/10.18653/v1/2022.findings-naacl.130","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.findings-naacl.130","pdf_url":"https://aclanthology.org/2022.findings-naacl.130.pdf","source":{"id":"https://openalex.org/S4363605604","display_name":"Findings of the Association for Computational Linguistics: NAACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: NAACL 2022","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2110.07736","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2110.07736","pdf_url":"https://arxiv.org/pdf/2110.07736","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2110.07736","doi":"https://doi.org/10.48550/arxiv.2110.07736","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2110.07736","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.18653/v1/2022.findings-naacl.130","doi":"https://doi.org/10.18653/v1/2022.findings-naacl.130","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.findings-naacl.130","pdf_url":"https://aclanthology.org/2022.findings-naacl.130.pdf","source":{"id":"https://openalex.org/S4363605604","display_name":"Findings of the Association for Computational Linguistics: NAACL","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/P4310320244","host_organization_name":"Association for Computational Linguistics","host_organization_lineage":["https://openalex.org/P4310320244"],"host_organization_lineage_names":["Association for Computational Linguistics"],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: NAACL 2022","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3207710400.pdf","grobid_xml":"https://content.openalex.org/works/W3207710400.grobid-xml"},"referenced_works_count":47,"referenced_works":["https://openalex.org/W102708294","https://openalex.org/W2081580037","https://openalex.org/W2136148671","https://openalex.org/W2158108973","https://openalex.org/W2163302275","https://openalex.org/W2165698076","https://openalex.org/W2251939518","https://openalex.org/W2282821441","https://openalex.org/W2289952147","https://openalex.org/W2396427021","https://openalex.org/W2516809705","https://openalex.org/W2594633041","https://openalex.org/W2897355816","https://openalex.org/W2909212904","https://openalex.org/W2912477312","https://openalex.org/W2951286828","https://openalex.org/W2953337107","https://openalex.org/W2962718684","https://openalex.org/W2963341956","https://openalex.org/W2963969878","https://openalex.org/W2969624041","https://openalex.org/W2970019270","https://openalex.org/W2970928735","https://openalex.org/W2973291083","https://openalex.org/W2984256198","https://openalex.org/W2996851481","https://openalex.org/W3007373432","https://openalex.org/W3016970897","https://openalex.org/W3021336872","https://openalex.org/W3023533951","https://openalex.org/W3023690688","https://openalex.org/W3034350582","https://openalex.org/W3034408878","https://openalex.org/W3034584102","https://openalex.org/W3035032873","https://openalex.org/W3035219095","https://openalex.org/W3035422656","https://openalex.org/W3095992020","https://openalex.org/W3099536922","https://openalex.org/W3104169042","https://openalex.org/W3134100302","https://openalex.org/W3134374554","https://openalex.org/W3170037207","https://openalex.org/W3173787059","https://openalex.org/W3176265725","https://openalex.org/W3177397812","https://openalex.org/W4297971002"],"related_works":["https://openalex.org/W3205598847","https://openalex.org/W7202191109","https://openalex.org/W7134067367","https://openalex.org/W6964939956","https://openalex.org/W4385565551","https://openalex.org/W4285298468","https://openalex.org/W3095992020","https://openalex.org/W4385573608","https://openalex.org/W2996601440","https://openalex.org/W3213681714"],"abstract_inverted_index":{"Recently,":[0],"NLP":[1,74],"models":[2,28,75,139],"have":[3,14],"achieved":[4],"remarkable":[5],"progress":[6],"across":[7,107],"a":[8,45,127],"variety":[9],"of":[10,48,130],"tasks;":[11],"however,":[12],"they":[13],"also":[15],"been":[16],"criticized":[17],"for":[18],"being":[19],"not":[20],"robust.Many":[21],"robustness":[22],"problems":[23],"can":[24,122],"be":[25],"attributed":[26],"to":[27,67,83,136],"exploiting":[29],"spurious":[30,71],"correlations,":[31],"or":[32,54],"shortcuts":[33,50],"between":[34],"the":[35,39,93],"training":[36],"data":[37],"and":[38,61,100,110,124,132],"task":[40],"labels.Most":[41],"existing":[42,80],"work":[43],"identifies":[44],"limited":[46],"set":[47,129],"task-specific":[49],"via":[51],"human":[52],"priors":[53],"error":[55],"analyses,":[56],"which":[57],"requires":[58],"extensive":[59],"expertise":[60],"efforts.In":[62],"this":[63],"paper,":[64],"we":[65],"aim":[66],"automatically":[68],"identify":[69,126],"such":[70],"correlations":[72],"in":[73,140],"at":[76],"scale.We":[77],"first":[78],"leverage":[79],"interpretability":[81],"methods":[82],"extract":[84],"tokens":[85,99,102],"that":[86,118],"significantly":[87],"affect":[88],"model's":[89],"decision":[90],"process":[91],"from":[92],"input":[94],"text.We":[95],"then":[96],"distinguish":[97],"\"genuine\"":[98],"\"spurious\"":[101],"by":[103],"analyzing":[104],"model":[105],"predictions":[106],"multiple":[108,141],"corpora":[109],"further":[111],"verify":[112],"them":[113],"through":[114],"knowledge-aware":[115],"perturbations.We":[116],"show":[117],"our":[119],"proposed":[120],"method":[121],"effectively":[123],"efficiently":[125],"scalable":[128],"\"shortcuts\",":[131],"mitigating":[133],"these":[134],"leads":[135],"more":[137],"robust":[138],"applications.":[142]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W3207710400","counts_by_year":[{"year":2026,"cited_by_count":17},{"year":2025,"cited_by_count":36},{"year":2024,"cited_by_count":24},{"year":2023,"cited_by_count":33},{"year":2022,"cited_by_count":6}],"updated_date":"2026-11-20T13:11:24.684918","created_date":"2021-10-25T00:00:00"}
