{"id":"https://openalex.org/W7165654692","doi":"https://doi.org/10.48550/arxiv.2606.22276","title":"Learning from Audio-Dependency Errors: Data Curation Strategies Based on Model Confusion Patterns in Audio Question Answering","display_name":"Learning from Audio-Dependency Errors: Data Curation Strategies Based on Model Confusion Patterns in Audio Question Answering","publication_year":2026,"publication_date":"2026-06-20","ids":{"openalex":"https://openalex.org/W7165654692","doi":"https://doi.org/10.48550/arxiv.2606.22276"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.22276","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22276","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.22276","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5068022654","display_name":"Hyeonuk Nam","orcid":"https://orcid.org/0000-0002-1169-5640"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Nam, Hyeonuk","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5068022654"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.40689998865127563,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.40689998865127563,"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":0.1932000070810318,"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/T11309","display_name":"Music and Audio Processing","score":0.08869999647140503,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/counterfactual-thinking","display_name":"Counterfactual thinking","score":0.7175999879837036},{"id":"https://openalex.org/keywords/confusion","display_name":"Confusion","score":0.6672999858856201},{"id":"https://openalex.org/keywords/frame","display_name":"Frame (networking)","score":0.59170001745224},{"id":"https://openalex.org/keywords/question-answering","display_name":"Question answering","score":0.428600013256073},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.39320001006126404},{"id":"https://openalex.org/keywords/data-curation","display_name":"Data curation","score":0.3012000024318695}],"concepts":[{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.7175999879837036},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6953999996185303},{"id":"https://openalex.org/C2781140086","wikidata":"https://www.wikidata.org/wiki/Q557945","display_name":"Confusion","level":2,"score":0.6672999858856201},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.59170001745224},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5541999936103821},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.49790000915527344},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.428600013256073},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.39320001006126404},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.35339999198913574},{"id":"https://openalex.org/C91632574","wikidata":"https://www.wikidata.org/wiki/Q15088675","display_name":"Data curation","level":2,"score":0.3012000024318695},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.29899999499320984},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.29109999537467957},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2786000072956085},{"id":"https://openalex.org/C511192102","wikidata":"https://www.wikidata.org/wiki/Q5156948","display_name":"Comprehension","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C180198813","wikidata":"https://www.wikidata.org/wiki/Q121182","display_name":"Information system","level":2,"score":0.2556000053882599},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.25099998712539673}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.22276","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22276","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.22276","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.22276","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,"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"},"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16","score":0.49825239181518555}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"We":[0],"frame":[1],"the":[2,27,68,95,99],"system":[3,61,102],"as":[4],"diagnostic":[5],"data":[6],"curation":[7],"for":[8,91,112],"a":[9,80,87,127],"large":[10],"audio-language":[11],"model:":[12],"before":[13],"fine-tuning,":[14],"we":[15],"probe":[16],"Qwen3-Omni-30B-A3B-Instruct":[17],"under":[18],"normal,":[19],"empty-audio,":[20],"and":[21,53,86,126],"shuffled-audio":[22],"conditions":[23],"to":[24,44],"identify":[25],"how":[26],"model's":[28],"answers":[29],"change":[30],"when":[31],"audio":[32],"evidence":[33],"is":[34,72],"removed":[35],"or":[36,55],"mismatched.":[37],"These":[38],"model":[39],"confusion":[40],"patterns":[41],"are":[42],"used":[43],"bucket":[45],"training":[46],"samples":[47],"into":[48],"text-prior,":[49],"shuffle-leak,":[50],"strong":[51],"audio-dependent,":[52],"hard":[54],"misleading":[56],"cases.":[57],"Our":[58],"strongest":[59],"train-only":[60,101],"fine-tunes":[62],"only":[63],"on":[64],"strong-audio":[65],"items,":[66],"where":[67],"normal":[69],"audio-question":[70],"pair":[71],"correct":[73],"but":[74],"both":[75],"counterfactual":[76],"variants":[77],"fail,":[78],"plus":[79],"small":[81],"number":[82],"of":[83],"empty-audio":[84],"negatives":[85],"text-only":[88],"response":[89,107],"normalizer":[90],"parse-failed":[92],"generations.":[93],"On":[94],"official":[96],"development":[97],"set,":[98],"best":[100],"reaches":[103],"67.27%":[104],"accuracy":[105],"after":[106],"normalization,":[108],"compared":[109],"with":[110],"65.90%":[111],"our":[113],"local":[114],"Qwen3-Omni":[115],"baseline.":[116],"Final":[117],"submissions":[118],"additionally":[119],"include":[120],"models":[121],"trained":[122],"using":[123],"train+development":[124],"splits":[125],"three-model":[128],"ensemble.":[129]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-24T00:00:00"}