{"id":"https://openalex.org/W4385572958","doi":"https://doi.org/10.18653/v1/2022.emnlp-main.285","title":"FiE: Building a Global Probability Space by Leveraging Early Fusion in Encoder for Open-Domain Question Answering","display_name":"FiE: Building a Global Probability Space by Leveraging Early Fusion in Encoder for Open-Domain Question Answering","publication_year":2022,"publication_date":"2022-01-01","ids":{"openalex":"https://openalex.org/W4385572958","doi":"https://doi.org/10.18653/v1/2022.emnlp-main.285"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2022.emnlp-main.285","doi":"https://doi.org/10.18653/v1/2022.emnlp-main.285","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.emnlp-main.285","pdf_url":"https://aclanthology.org/2022.emnlp-main.285.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2022 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2022.emnlp-main.285.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5030211209","display_name":"Akhil Kedia","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Akhil Kedia","raw_affiliation_strings":["Samsung Research, Seoul"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Samsung Research, Seoul","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060189444","display_name":"Mohd Abbas Zaidi","orcid":"https://orcid.org/0000-0003-2961-4404","observed_orcids":["https://orcid.org/0000-0003-2961-4404"]},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Mohd Abbas Zaidi","raw_affiliation_strings":["Samsung Research, Seoul"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Samsung Research, Seoul","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5052875434","display_name":"Haejun Lee","orcid":"https://orcid.org/0000-0002-6284-2903","observed_orcids":["https://orcid.org/0000-0002-6284-2903"]},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Haejun Lee","raw_affiliation_strings":["Samsung Research, Seoul"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Samsung Research, Seoul","institution_ids":["https://openalex.org/I2250650973"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I2250650973"],"apc_list":null,"apc_paid":null,"fwci":1.3813,"has_fulltext":true,"cited_by_count":14,"citation_normalized_percentile":{"value":0.84964302,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"4246","last_page":"4260"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9771000146865845,"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":0.9771000146865845,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.008799999952316284,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.007699999958276749,"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/open-domain-question-answering","display_name":"open-domain question answering","score":0.9980000257492065},{"id":"https://openalex.org/keywords/early-fusion","display_name":"feature-level fusion","score":0.781000018119812},{"id":"https://openalex.org/keywords/parameter-efficiency","display_name":"parameter efficiency","score":0.4860000014305115},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.7940000295639038},{"id":"https://openalex.org/keywords/hallucination-mitigation","display_name":"hallucination mitigation","score":0.722000002861023},{"id":"https://openalex.org/keywords/synthetic-data-augmentation","display_name":"synthetic data augmentation","score":0.6669999957084656},{"id":"https://openalex.org/keywords/natural-questions","display_name":"Natural Questions","score":0.5270000100135803},{"id":"https://openalex.org/keywords/triviaqa","display_name":"TriviaQA","score":0.996999979019165}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8029143810272217},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7026713490486145},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.6326253414154053},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.5431129336357117},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5207859873771667},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.503473699092865},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4760843813419342},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.46853581070899963},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4670274257659912},{"id":"https://openalex.org/C2988773926","wikidata":"https://www.wikidata.org/wiki/Q25104379","display_name":"Generative adversarial network","level":3,"score":0.44305768609046936},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.41538557410240173},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.1314813196659088},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2022.emnlp-main.285","doi":"https://doi.org/10.18653/v1/2022.emnlp-main.285","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.emnlp-main.285","pdf_url":"https://aclanthology.org/2022.emnlp-main.285.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2022 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2211.10147","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2211.10147","pdf_url":"https://arxiv.org/pdf/2211.10147","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2211.10147","doi":"https://doi.org/10.48550/arxiv.2211.10147","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2211.10147","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.18653/v1/2022.emnlp-main.285","doi":"https://doi.org/10.18653/v1/2022.emnlp-main.285","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2022.emnlp-main.285","pdf_url":"https://aclanthology.org/2022.emnlp-main.285.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2022 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4385572958.pdf","grobid_xml":"https://content.openalex.org/works/W4385572958.grobid-xml"},"referenced_works_count":40,"referenced_works":["https://openalex.org/W2252136820","https://openalex.org/W2612431505","https://openalex.org/W2626778328","https://openalex.org/W2765390718","https://openalex.org/W2769395616","https://openalex.org/W2912924812","https://openalex.org/W2947497897","https://openalex.org/W2947813521","https://openalex.org/W2963341956","https://openalex.org/W2971105107","https://openalex.org/W2979826702","https://openalex.org/W2996035354","https://openalex.org/W3015468748","https://openalex.org/W3015883388","https://openalex.org/W3022265721","https://openalex.org/W3027879771","https://openalex.org/W3034671305","https://openalex.org/W3035187010","https://openalex.org/W3035231859","https://openalex.org/W3039017601","https://openalex.org/W3045733172","https://openalex.org/W3104777900","https://openalex.org/W3115513107","https://openalex.org/W3118668786","https://openalex.org/W3119437245","https://openalex.org/W3132730484","https://openalex.org/W3141362964","https://openalex.org/W3151130473","https://openalex.org/W3154800474","https://openalex.org/W3166576616","https://openalex.org/W3176390686","https://openalex.org/W3196697103","https://openalex.org/W3199011812","https://openalex.org/W3206427007","https://openalex.org/W4225727438","https://openalex.org/W4226209916","https://openalex.org/W4285126267","https://openalex.org/W4285151662","https://openalex.org/W4292956554","https://openalex.org/W4309417034"],"related_works":["https://openalex.org/W4285284346","https://openalex.org/W3039017601","https://openalex.org/W4226209916","https://openalex.org/W3118244628","https://openalex.org/W3176182290","https://openalex.org/W2922234509","https://openalex.org/W3206427007","https://openalex.org/W4386566577","https://openalex.org/W4385212420","https://openalex.org/W6885090953"],"abstract_inverted_index":{"Generative":[0],"models":[1,8,30,37,156,180],"have":[2],"recently":[3],"started":[4],"to":[5,19,32,39,50,64,71,80,97],"outperform":[6,113,154],"extractive":[7,36],"in":[9,102],"Open":[10],"Domain":[11],"Question":[12,126,188,195],"Answering,":[13],"largely":[14],"by":[15,118],"leveraging":[16],"their":[17,27,56],"decoder":[18,52],"attend":[20],"over":[21,84],"multiple":[22,75],"encoded":[23],"passages":[24],"and":[25,55,134,141,164],"combining":[26],"information.However,":[28],"generative":[29],"tend":[31],"be":[33],"larger":[34,155],"than":[35],"due":[38,49],"the":[40,69,103,114,124,137,158],"need":[41],"for":[42,174],"a":[43],"decoder,":[44],"run":[45],"slower":[46],"during":[47,139],"inference":[48],"auto-regressive":[51],"beam":[53],"search,":[54],"generated":[57],"output":[58],"often":[59,182],"suffers":[60],"from":[61,74],"hallucinations.We":[62],"propose":[63,90],"extend":[65],"transformer":[66],"encoders":[67],"with":[68,149],"ability":[70],"fuse":[72],"information":[73],"passages,":[76],"using":[77,129],"global":[78,104],"representation":[79],"provide":[81],"cross-sample":[82],"attention":[83],"all":[85,107],"tokens":[86],"across":[87],"samples.Furthermore,":[88],"we":[89,112,153],"an":[91],"alternative":[92],"answer":[93,100],"span":[94],"probability":[95],"calculation":[96],"better":[98],"aggregate":[99],"scores":[101],"space":[105],"of":[106,132,136,167],"samples.Using":[108],"our":[109,168],"proposed":[110],"method,":[111],"current":[115],"stateof-the-art":[116],"method":[117,169],"2.5":[119],"Exact":[120,143],"Match":[121,144],"score":[122],"on":[123,145,157],"Natural":[125],"dataset":[127,160],"while":[128],"only":[130],"25%":[131],"parameters":[133],"35%":[135],"latency":[138,163],"inference,":[140],"4.4":[142],"We-bQuestions":[146],"dataset.When":[147],"coupled":[148],"synthetic":[150],"data":[151],"augmentation,":[152],"TriviaQA":[159],"as":[161,178],"well.The":[162],"parameter":[165],"savings":[166],"make":[170],"it":[171],"particularly":[172],"attractive":[173],"open-domain":[175],"question":[176],"answering,":[177],"these":[179],"are":[181],"compute-intensive.Global":[183],"Tokens":[184],"Global":[185],"Encoder":[186,193,200,224,227],"Layer":[187,194,201],"+":[189,196],"Passage":[190,192,197,199],"1":[191,207],"2":[198,214],"Span":[202,209],"Classifier":[203,210],"\ud835\udc43\ud835\udc43":[204,211,216],"\ud835\udc60\ud835\udc60":[205,212],"(\ud835\udc60\ud835\udc60":[206,213],")":[208,215],"\ud835\udc34\ud835\udc34":[217],"(\"\ud835\udc41\ud835\udc41\ud835\udc41\ud835\udc41\ud835\udc41\ud835\udc41":[218],"\ud835\udc4c\ud835\udc4c\ud835\udc4c\ud835\udc4c\ud835\udc4c\ud835\udc4c\ud835\udc4c\ud835\udc4c\")":[219],"\u03a3":[220],"Multi-Head":[221],"Attention":[222],"(Global":[223],"Layer)x":[225],"N":[226],"Layers":[228],"Q":[229],"KV":[230]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":4}],"updated_date":"2026-10-30T12:39:25.321706","created_date":"2023-08-05T00:00:00"}