{"id":"https://openalex.org/W7135416617","doi":"https://doi.org/10.5715/jnlp.33.207","title":"GPT-MM: Improving Multimodal In-context Learning with Task-specific Retrieval and Reasoning","display_name":"GPT-MM: Improving Multimodal In-context Learning with Task-specific Retrieval and Reasoning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7135416617","doi":"https://doi.org/10.5715/jnlp.33.207"},"language":"en","primary_location":{"id":"doi:10.5715/jnlp.33.207","is_oa":true,"landing_page_url":"https://doi.org/10.5715/jnlp.33.207","pdf_url":"https://www.jstage.jst.go.jp/article/jnlp/33/1/33_207/_pdf","source":{"id":"https://openalex.org/S4210212357","display_name":"Journal of Natural Language Processing","issn_l":"1340-7619","issn":["1340-7619","2185-8314"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Natural Language Processing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://www.jstage.jst.go.jp/article/jnlp/33/1/33_207/_pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100959340","display_name":"Zhen Wan","orcid":null},"institutions":[{"id":"https://openalex.org/I22299242","display_name":"Kyoto University","ror":"https://ror.org/02kpeqv85","country_code":"JP","type":"education","lineage":["https://openalex.org/I22299242"]}],"countries":["JP"],"is_corresponding":true,"raw_author_name":"Zhen Wan","raw_affiliation_strings":["Graduate School of Informatics, Kyoto University"],"affiliations":[{"raw_affiliation_string":"Graduate School of Informatics, Kyoto University","institution_ids":["https://openalex.org/I22299242"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129182399","display_name":"Fei Cheng","orcid":null},"institutions":[{"id":"https://openalex.org/I22299242","display_name":"Kyoto University","ror":"https://ror.org/02kpeqv85","country_code":"JP","type":"education","lineage":["https://openalex.org/I22299242"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Fei Cheng","raw_affiliation_strings":["Graduate School of Informatics, Kyoto University"],"affiliations":[{"raw_affiliation_string":"Graduate School of Informatics, Kyoto University","institution_ids":["https://openalex.org/I22299242"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129167585","display_name":"Sadao Kurohashi","orcid":null},"institutions":[{"id":"https://openalex.org/I184597095","display_name":"National Institute of Informatics","ror":"https://ror.org/04ksd4g47","country_code":"JP","type":"facility","lineage":["https://openalex.org/I1319490839","https://openalex.org/I184597095","https://openalex.org/I4210158934"]},{"id":"https://openalex.org/I22299242","display_name":"Kyoto University","ror":"https://ror.org/02kpeqv85","country_code":"JP","type":"education","lineage":["https://openalex.org/I22299242"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Sadao Kurohashi","raw_affiliation_strings":["Graduate School of Informatics, Kyoto University","National Institute of Informatics"],"affiliations":[{"raw_affiliation_string":"Graduate School of Informatics, Kyoto University","institution_ids":["https://openalex.org/I22299242"]},{"raw_affiliation_string":"National Institute of Informatics","institution_ids":["https://openalex.org/I184597095"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5100959340"],"corresponding_institution_ids":["https://openalex.org/I22299242"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.94020614,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":null,"biblio":{"volume":"33","issue":"1","first_page":"207","last_page":"239"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.17579999566078186,"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/T12031","display_name":"Speech and dialogue systems","score":0.17579999566078186,"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":0.16120000183582306,"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"}},{"id":"https://openalex.org/T11902","display_name":"Intelligent Tutoring Systems and Adaptive Learning","score":0.15520000457763672,"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/feature","display_name":"Feature (linguistics)","score":0.31369999051094055},{"id":"https://openalex.org/keywords/case-based-reasoning","display_name":"Case-based reasoning","score":0.29260000586509705},{"id":"https://openalex.org/keywords/multimodality","display_name":"Multimodality","score":0.27810001373291016},{"id":"https://openalex.org/keywords/field","display_name":"Field (mathematics)","score":0.2662999927997589},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.2646999955177307}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6560999751091003},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5034999847412109},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3149000108242035},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.31369999051094055},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.29260000586509705},{"id":"https://openalex.org/C2780910867","wikidata":"https://www.wikidata.org/wiki/Q1952416","display_name":"Multimodality","level":2,"score":0.27810001373291016},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.2646999955177307},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.2630000114440918},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2614000141620636}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.5715/jnlp.33.207","is_oa":true,"landing_page_url":"https://doi.org/10.5715/jnlp.33.207","pdf_url":"https://www.jstage.jst.go.jp/article/jnlp/33/1/33_207/_pdf","source":{"id":"https://openalex.org/S4210212357","display_name":"Journal of Natural Language Processing","issn_l":"1340-7619","issn":["1340-7619","2185-8314"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Natural Language Processing","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.5715/jnlp.33.207","is_oa":true,"landing_page_url":"https://doi.org/10.5715/jnlp.33.207","pdf_url":"https://www.jstage.jst.go.jp/article/jnlp/33/1/33_207/_pdf","source":{"id":"https://openalex.org/S4210212357","display_name":"Journal of Natural Language Processing","issn_l":"1340-7619","issn":["1340-7619","2185-8314"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of Natural Language Processing","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7135416617.pdf","grobid_xml":"https://content.openalex.org/works/W7135416617.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4],"exhibited":[5],"impressive":[6],"generalization":[7],"through":[8],"in-context":[9,181],"learning":[10,182],"(ICL),":[11],"yet":[12],"most":[13],"studies":[14],"focus":[15],"on":[16],"textual":[17,65,142],"tasks,":[18],"leaving":[19],"the":[20,62,92,109],"mechanisms":[21,106],"that":[22,41,74,85,167],"enable":[23],"ICL":[24,39,113],"to":[25,77,123],"generalize":[26],"across":[27,184],"modalities":[28],"largely":[29],"unexplored.":[30],"To":[31],"bridge":[32],"this":[33,121],"gap,":[34],"we":[35],"propose":[36],"a":[37,69,175,179],"unified":[38,180],"framework":[40,63,122,147],"integrates":[42],"task-aware":[43,168],"demonstration":[44,100],"retrieval":[45,83,169],"and":[46,57,114,133,143,151,154,170],"label-induced":[47,96,171],"reasoning":[48,97,172],"as":[49],"two":[50],"complementary":[51],"components":[52],"for":[53,128,135,178],"improving":[54],"both":[55,141],"accuracy":[56],"interpretability.":[58],"We":[59,118],"first":[60],"validate":[61],"in":[64],"relation":[66],"extraction":[67],"(RE),":[68],"representative":[70],"structured":[71],"prediction":[72],"task":[73],"challenges":[75],"LLMs":[76],"infer":[78],"fine-grained":[79],"entity\u2013relation":[80],"semantics.":[81],"Task-aware":[82],"ensures":[84],"retrieved":[86],"examples":[87],"are":[88],"semantically":[89],"aligned":[90],"with":[91,101,161],"target":[93],"instance,":[94],"while":[95],"enriches":[98],"each":[99],"label-grounded":[102],"explanatory":[103],"logic.":[104],"These":[105,164],"substantially":[107],"narrow":[108],"performance":[110],"gap":[111],"between":[112],"fully":[115],"supervised":[116],"models.":[117,163],"then":[119],"extend":[120],"multimodal":[124,144],"ICL,":[125],"leveraging":[126],"GPT-4o":[127],"visual":[129],"question":[130,137],"answering":[131,138],"(VQA)":[132],"Whisper-large-v3":[134],"audio":[136],"(AudioQA).":[139],"Across":[140],"benchmarks,":[145],"our":[146],"consistently":[148],"outperforms":[149],"GPT-3":[150],"GPT-4":[152],"baselines":[153],"achieves":[155],"competitive":[156],"or":[157],"superior":[158],"results":[159],"compared":[160],"fine-tuned":[162],"findings":[165],"demonstrate":[166],"together":[173],"form":[174],"generalizable":[176],"foundation":[177],"paradigm":[183],"modalities.":[185]},"counts_by_year":[],"updated_date":"2026-03-16T07:19:36.451410","created_date":"2026-03-15T00:00:00"}