{"id":"https://openalex.org/W2951328433","doi":"https://doi.org/10.18653/v1/p19-1262","title":"Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA","display_name":"Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA","publication_year":2019,"publication_date":"2019-01-01","ids":{"openalex":"https://openalex.org/W2951328433","doi":"https://doi.org/10.18653/v1/p19-1262","mag":"2951328433"},"language":"en","primary_location":{"id":"doi:10.18653/v1/p19-1262","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p19-1262","pdf_url":"https://www.aclweb.org/anthology/P19-1262.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/P19-1262.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101705761","display_name":"Yichen Jiang","orcid":"https://orcid.org/0000-0002-9139-3799"},"institutions":[{"id":"https://openalex.org/I114027177","display_name":"University of North Carolina at Chapel Hill","ror":"https://ror.org/0130frc33","country_code":"US","type":"education","lineage":["https://openalex.org/I114027177"]},{"id":"https://openalex.org/I1333535994","display_name":"University of North Carolina Health Care","ror":"https://ror.org/00qz24g20","country_code":"US","type":"healthcare","lineage":["https://openalex.org/I1333535994"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Yichen Jiang","raw_affiliation_strings":["UNC Chapel Hill"],"affiliations":[{"raw_affiliation_string":"UNC Chapel Hill","institution_ids":["https://openalex.org/I1333535994","https://openalex.org/I114027177"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001987532","display_name":"Mohit Bansal","orcid":"https://orcid.org/0000-0001-5522-1351"},"institutions":[{"id":"https://openalex.org/I114027177","display_name":"University of North Carolina at Chapel Hill","ror":"https://ror.org/0130frc33","country_code":"US","type":"education","lineage":["https://openalex.org/I114027177"]},{"id":"https://openalex.org/I1333535994","display_name":"University of North Carolina Health Care","ror":"https://ror.org/00qz24g20","country_code":"US","type":"healthcare","lineage":["https://openalex.org/I1333535994"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mohit Bansal","raw_affiliation_strings":["UNC Chapel Hill"],"affiliations":[{"raw_affiliation_string":"UNC Chapel Hill","institution_ids":["https://openalex.org/I1333535994","https://openalex.org/I114027177"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5101705761"],"corresponding_institution_ids":["https://openalex.org/I114027177","https://openalex.org/I1333535994"],"apc_list":null,"apc_paid":null,"fwci":9.249,"has_fulltext":true,"cited_by_count":83,"citation_normalized_percentile":{"value":0.98265159,"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":"2726","last_page":"2736"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9998999834060669,"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/T10028","display_name":"Topic Modeling","score":0.9998999834060669,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9986000061035156,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9785000085830688,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/adversarial-system","display_name":"Adversarial system","score":0.9005200862884521},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7975230813026428},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.7124238610267639},{"id":"https://openalex.org/keywords/hop","display_name":"Hop (telecommunications)","score":0.5816375017166138},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5625470876693726},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.479136198759079},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.4575594365596771},{"id":"https://openalex.org/keywords/sentence","display_name":"Sentence","score":0.45221981406211853},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.41616299748420715},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.08016461133956909}],"concepts":[{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.9005200862884521},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7975230813026428},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.7124238610267639},{"id":"https://openalex.org/C25906391","wikidata":"https://www.wikidata.org/wiki/Q1432381","display_name":"Hop (telecommunications)","level":2,"score":0.5816375017166138},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5625470876693726},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.479136198759079},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.4575594365596771},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.45221981406211853},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.41616299748420715},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08016461133956909},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.0},{"id":"https://openalex.org/C111368507","wikidata":"https://www.wikidata.org/wiki/Q43518","display_name":"Oceanography","level":1,"score":0.0},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/p19-1262","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p19-1262","pdf_url":"https://www.aclweb.org/anthology/P19-1262.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/p19-1262","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p19-1262","pdf_url":"https://www.aclweb.org/anthology/P19-1262.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320309480","display_name":"Nvidia","ror":"https://ror.org/03jdj4y14"},{"id":"https://openalex.org/F4320332180","display_name":"Defense Advanced Research Projects Agency","ror":"https://ror.org/02caytj08"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2951328433.pdf","grobid_xml":"https://content.openalex.org/works/W2951328433.grobid-xml"},"referenced_works_count":32,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W1525961042","https://openalex.org/W1544827683","https://openalex.org/W1995945562","https://openalex.org/W2064675550","https://openalex.org/W2250539671","https://openalex.org/W2251818205","https://openalex.org/W2295953541","https://openalex.org/W2471900581","https://openalex.org/W2551396370","https://openalex.org/W2552027021","https://openalex.org/W2740747242","https://openalex.org/W2770970123","https://openalex.org/W2783905278","https://openalex.org/W2889344053","https://openalex.org/W2889787757","https://openalex.org/W2949615363","https://openalex.org/W2951534261","https://openalex.org/W2962718483","https://openalex.org/W2962790689","https://openalex.org/W2962808855","https://openalex.org/W2962985038","https://openalex.org/W2963223524","https://openalex.org/W2963323070","https://openalex.org/W2963339397","https://openalex.org/W2963748441","https://openalex.org/W2963783970","https://openalex.org/W2963866616","https://openalex.org/W2963871484","https://openalex.org/W2963969878","https://openalex.org/W4297749157","https://openalex.org/W4310299640"],"related_works":["https://openalex.org/W2502115930","https://openalex.org/W4246396837","https://openalex.org/W3176240006","https://openalex.org/W3126451824","https://openalex.org/W2482350142","https://openalex.org/W1561927205","https://openalex.org/W3191453585","https://openalex.org/W4297672492","https://openalex.org/W4310988119","https://openalex.org/W4285226279"],"abstract_inverted_index":{"Multi-hop":[0],"question":[1,107],"answering":[2],"requires":[3],"a":[4,14,99],"model":[5,123,144],"to":[6,17,43,105,112,132],"connect":[7],"multiple":[8],"pieces":[9],"of":[10,52,58,175],"evidence":[11],"scattered":[12],"in":[13,27],"long":[15],"context":[16],"answer":[18],"the":[19,28,44,50,53,74,84,93,106,114,126,133,136,159,173],"question.":[20],"In":[21],"this":[22,33,121,142],"paper,":[23],"we":[24,97],"show":[25,119],"that":[26,39,69,102,120,165,178],"multihop":[29,116],"HotpotQA":[30],"We":[31,118,163],"demonstrate":[32],"issue":[34],"by":[35],"constructing":[36],"adversarial":[37,66,82,94,140,184],"documents":[38],"create":[40],"contradicting":[41],"answers":[42],"shortcut":[45],"but":[46,88,156],"do":[47],"not":[48,145],"affect":[49],"validity":[51],"original":[54],"answer.":[55],"The":[56],"performance":[57,86],"strong":[59],"baseline":[60,137],"models":[61,177],"drops":[62],"significantly":[63],"on":[64,92,125,153],"our":[65],"evaluation,":[67],"indicating":[68],"they":[70],"are":[71],"indeed":[72],"exploiting":[73],"shortcuts":[75],"rather":[76],"than":[77,135],"performing":[78],"multi-hop":[79,189],"reasoning.":[80,117],"After":[81,139],"training,":[83,141],"baseline's":[85],"improves":[87],"is":[89,129],"still":[90],"limited":[91],"evaluation.":[95],"Hence,":[96],"use":[98],"control":[100],"unit":[101],"dynamically":[103],"attends":[104],"at":[108],"different":[109],"reasoning":[110,182],"hops":[111],"guide":[113],"model's":[115],"2-hop":[122,143],"trained":[124,152],"regular":[127,154],"data":[128],"more":[130],"robust":[131],"adversaries":[134],"model.":[138],"only":[146],"achieves":[147],"improvements":[148,170],"over":[149],"its":[150],"counterpart":[151],"data,":[155],"also":[157],"outperforms":[158],"adversarially-trained":[160],"1-hop":[161],"baseline.":[162],"hope":[164],"these":[166],"insights":[167],"and":[168],"initial":[169],"will":[171],"motivate":[172],"development":[174],"new":[176],"combine":[179],"explicit":[180],"compositional":[181],"with":[183],"training.":[185],"1":[186],"Recently,":[187],"several":[188],"QA":[190],"datasets,":[191],"such":[192]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":14},{"year":2022,"cited_by_count":11},{"year":2021,"cited_by_count":30},{"year":2020,"cited_by_count":18},{"year":2019,"cited_by_count":5}],"updated_date":"2026-03-17T09:09:15.849793","created_date":"2025-10-10T00:00:00"}