{"id":"https://openalex.org/W7224092647","doi":"https://doi.org/10.48550/arxiv.2610.36578","title":"AudioRSI: Recursively Self-Improving Audio Reasoners","display_name":"AudioRSI: Recursively Self-Improving Audio Reasoners","publication_year":2026,"publication_date":"2026-10-27","ids":{"openalex":"https://openalex.org/W7224092647","doi":"https://doi.org/10.48550/arxiv.2610.36578"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.36578","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36578","pdf_url":"https://arxiv.org/pdf/2610.36578","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.36578","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5159995569","display_name":"Jiajun Fan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Fan, Jiajun","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076628167","display_name":"Roger Ren","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ren, Roger","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160017388","display_name":"Jingyuan Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li, Jingyuan","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020350278","display_name":"Rahul Pandey","orcid":"https://orcid.org/0000-0003-2748-0502","observed_orcids":["https://orcid.org/0000-0003-2748-0502"]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Pandey, Rahul","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088032905","display_name":"Prashanth Gurunath Shivakumar","orcid":"https://orcid.org/0000-0003-1632-3309","observed_orcids":["https://orcid.org/0000-0003-1632-3309"]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shivakumar, Prashanth Gurunath","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014054572","display_name":"Ankur Gandhe","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gandhe, Ankur","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160018795","display_name":"Yile Gu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gu, Yile","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160010337","display_name":"Ge Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Ge","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5109934165","display_name":"Ivan Bulyko","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bulyko, Ivan","raw_affiliation_strings":["UIUC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UIUC","institution_ids":["https://openalex.org/I157725225"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157725225"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.7949000000953674,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.7949000000953674,"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/T10860","display_name":"Speech and Audio Processing","score":0.03819999843835831,"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"}},{"id":"https://openalex.org/T11309","display_name":"Music and Audio Processing","score":0.030700000002980232,"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/large-audio-language-models","display_name":"large audio language models","score":0.9179999828338623},{"id":"https://openalex.org/keywords/recursive-self-improvement","display_name":"recursive self-improvement","score":0.800000011920929},{"id":"https://openalex.org/keywords/hallucination-mitigation","display_name":"hallucination mitigation","score":0.7179999947547913}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7480000257492065},{"id":"https://openalex.org/C168773036","wikidata":"https://www.wikidata.org/wiki/Q264164","display_name":"Recursion (computer science)","level":2,"score":0.6603000164031982},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6087999939918518},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5271999835968018},{"id":"https://openalex.org/C64922751","wikidata":"https://www.wikidata.org/wiki/Q4650799","display_name":"Audio signal","level":3,"score":0.3684999942779541},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36320000886917114},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.35010001063346863},{"id":"https://openalex.org/C45340560","wikidata":"https://www.wikidata.org/wiki/Q215382","display_name":"Disjoint sets","level":2,"score":0.3407000005245209},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3310999870300293},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.3260999917984009},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.28859999775886536},{"id":"https://openalex.org/C128422554","wikidata":"https://www.wikidata.org/wiki/Q20077126","display_name":"Sound recording and reproduction","level":2,"score":0.28619998693466187},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.2768999934196472},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.2766999900341034},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.2596000134944916},{"id":"https://openalex.org/C38956757","wikidata":"https://www.wikidata.org/wiki/Q716215","display_name":"Audio feedback","level":2,"score":0.25130000710487366}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.36578","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36578","pdf_url":"https://arxiv.org/pdf/2610.36578","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.2610.36578","doi":"https://doi.org/10.48550/arxiv.2610.36578","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.36578","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":"pmh:oai:arXiv.org:2610.36578","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.36578","pdf_url":"https://arxiv.org/pdf/2610.36578","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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7224092647.pdf","grobid_xml":"https://content.openalex.org/works/W7224092647.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7223845761","https://openalex.org/W7172533354","https://openalex.org/W7138876343","https://openalex.org/W4417322621","https://openalex.org/W6929674484","https://openalex.org/W7155062258","https://openalex.org/W4417166429","https://openalex.org/W7204964929","https://openalex.org/W7129519707","https://openalex.org/W7134905021"],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1,87],"has":[2],"made":[3],"Audio":[4,47,83,208],"Large":[5],"Language":[6],"Models":[7],"(Audio":[8],"LLMs)":[9],"better":[10],"reasoners,":[11],"but":[12],"the":[13,22,40,63,70,76,126,131,135,138,142,152,186,199,227,235],"reward":[14,52,159,245,249],"that":[15,36,45,65,148,161],"drives":[16],"it":[17,100,108,144,190],"still":[18],"comes":[19],"from":[20,185,198],"outside":[21],"model:":[23],"a":[24,27,85,91,106,117,172,182,223,274],"final-answer":[25],"check,":[26],"hand-crafted":[28],"process":[29],"score,":[30],"or":[31],"an":[32,46,98,269],"external":[33,174,244,270],"LLM":[34,48],"judge":[35],"stays":[37],"fixed":[38],"while":[39,181],"policy":[41,188],"improves.":[42],"We":[43,78],"show":[44],"can":[49,55,145],"supply":[50],"this":[51],"itself,":[53],"and":[54,104,211,233,264,273],"do":[56],"so":[57,69],"recursively:":[58],"each":[59],"improved":[60],"model":[61,93],"becomes":[62],"critic":[64,107,121,136,183],"trains":[66],"its":[67,113],"successor,":[68],"training":[71,201],"signal":[72],"improves":[73],"along":[74],"with":[75],"model.":[77],"introduce":[79],"AudioRSI":[80,203,254],"(Recursively":[81],"Self-Improving":[82],"Reasoners),":[84],"reinforcement":[86],"framework":[88],"in":[89],"which":[90,129],"single":[92],"plays":[94],"two":[95],"roles:":[96],"as":[97,105,141],"actor":[99],"generates":[101],"reasoning":[102,147,160,258],"paths,":[103],"compares":[109],"them":[110],"pairwise,":[111],"turning":[112],"own":[114],"judgments":[115],"into":[116],"process-level":[118],"reward.":[119],"The":[120,165],"is":[122,167],"periodically":[123],"replaced":[124],"by":[125,268],"current":[127],"policy,":[128],"closes":[130],"recursive":[132],"loop.":[133],"Because":[134],"hears":[137],"same":[139],"recording":[140],"actor,":[143],"reject":[146],"cites":[149],"acoustic":[150],"evidence":[151],"audio":[153],"does":[154],"not":[155],"contain,":[156],"rather":[157],"than":[158],"merely":[162],"sounds":[163],"plausible.":[164],"recursion":[166],"what":[168],"sustains":[169],"progress:":[170],"under":[171],"frozen":[173],"judge,":[175],"out-of-distribution":[176,194],"accuracy":[177],"plateaus":[178],"during":[179],"training,":[180],"refreshed":[184],"improving":[187],"keeps":[189],"rising.":[191],"Across":[192],"four":[193],"benchmarks,":[195],"all":[196],"disjoint":[197],"AVQA":[200],"distribution,":[202],"outperforms":[204],"strong":[205],"baselines":[206],"including":[207],"Flamingo":[209],"3":[210],"Gemini":[212],"2.5":[213],"Pro":[214],"on":[215,220,226,239],"MMAU":[216],"test-mini,":[217],"ranks":[218],"first":[219],"MMSU,":[221],"achieves":[222],"clear":[224],"lead":[225],"challenging":[228],"Multi-Audio":[229],"split":[230],"of":[231],"MMAU-Pro,":[232],"leads":[234],"7B":[236],"open-source":[237],"tier":[238],"MMAR":[240],"among":[241],"methods":[242],"without":[243],"models.":[246],"With":[247],"no":[248],"specifying":[250],"how":[251],"to":[252],"reason,":[253],"also":[255],"develops":[256],"structured":[257],"behaviors":[259],"(logical":[260],"reasoning,":[261],"perception-aware":[262],"analysis,":[263],"reduced":[265],"hallucination),":[266],"corroborated":[267],"LLM-as-Judge":[271],"evaluation":[272],"keyword-based":[275],"behavior":[276],"analysis.":[277]},"counts_by_year":[],"updated_date":"2026-10-31T12:09:11.096373","created_date":"2026-10-29T00:00:00"}