{"id":"https://openalex.org/W7138092950","doi":"https://doi.org/10.1609/aaai.v40i24.39112","title":"OTARo: Once Tuning for All Precisions Toward Robust On-Device LLMs","display_name":"OTARo: Once Tuning for All Precisions Toward Robust On-Device LLMs","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138092950","doi":"https://doi.org/10.1609/aaai.v40i24.39112"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i24.39112","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i24.39112","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1609/aaai.v40i24.39112","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129668745","display_name":"Shaoyuan Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shaoyuan Chen","raw_affiliation_strings":["Houmo AI\nSun Yat-sen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI\nSun Yat-sen University","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129731593","display_name":"Zhixuan Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Zhixuan Chen","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129699859","display_name":"Dawei Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Dawei Yang","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129711707","display_name":"Zhihang Yuan","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhihang Yuan","raw_affiliation_strings":["Peking University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129689832","display_name":"Qiang Wu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Qiang Wu","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"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":"40","issue":"24","first_page":"20244","last_page":"20252"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.4262999892234802,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.4262999892234802,"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/T10028","display_name":"Topic Modeling","score":0.06840000301599503,"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/T11948","display_name":"Machine Learning in Materials Science","score":0.05609999969601631,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.7190999984741211},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.6924999952316284},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.6553999781608582},{"id":"https://openalex.org/keywords/adaptability","display_name":"Adaptability","score":0.48249998688697815},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.42480000853538513},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.383899986743927}],"concepts":[{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.7190999984741211},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6924999952316284},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6586999893188477},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.6553999781608582},{"id":"https://openalex.org/C177606310","wikidata":"https://www.wikidata.org/wiki/Q5674297","display_name":"Adaptability","level":2,"score":0.48249998688697815},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.42480000853538513},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.383899986743927},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.36880001425743103},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.35589998960494995},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3479999899864197},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.30390000343322754},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.3003999888896942},{"id":"https://openalex.org/C2778738651","wikidata":"https://www.wikidata.org/wiki/Q16546687","display_name":"Novelty","level":2,"score":0.2669000029563904}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i24.39112","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i24.39112","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i24.39112","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i24.39112","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"fine-tuning":[4,37],"techniques":[5],"not":[6],"only":[7],"improve":[8],"the":[9,36,81,94,175],"adaptability":[10],"to":[11,55,59,63,79,106,129,142],"diverse":[12],"downstream":[13,147],"tasks,":[14],"but":[15],"also":[16],"mitigate":[17],"adverse":[18],"effects":[19],"of":[20,84,137],"model":[21],"quantization.":[22],"Despite":[23],"this,":[24],"conventional":[25],"quantization":[26,46,109,127],"suffers":[27],"from":[28],"its":[29],"structural":[30],"limitation":[31],"that":[32,73,102,207],"hinders":[33],"flexibility":[34],"during":[35],"and":[38,193,212],"deployment":[39],"stages.":[40],"Practical":[41],"on-device":[42,82,104],"tasks":[43,53],"demand":[44],"different":[45,49,131,158],"precisions":[47,85,110],"(i.e.":[48],"bit-widths),":[50],"e.g.,":[51,203],"understanding":[52],"tend":[54],"exhibit":[56],"higher":[57],"tolerance":[58],"reduced":[60],"precision":[61],"compared":[62],"generation":[64],"tasks.":[65],"Conventional":[66],"quantization,":[67],"typically":[68],"relying":[69],"on":[70,200],"scaling":[71],"factors":[72],"are":[74],"incompatible":[75],"across":[76],"bit-widths,":[77],"fails":[78],"support":[80],"switching":[83],"when":[86],"confronted":[87],"with":[88],"complex":[89],"real\u2011world":[90],"scenarios.":[91],"To":[92],"overcome":[93],"dilemma,":[95],"we":[96],"propose":[97],"OTARo,":[98],"a":[99,125,138,151,179],"novel":[100],"method":[101,161],"enables":[103],"LLMs":[105],"flexibly":[107],"switch":[108],"while":[111],"maintaining":[112],"performance":[113,214],"robustness":[114,145],"through":[115,133],"once":[116],"fine-tuning.":[117],"OTARo":[118,149,208],"introduces":[119],"Shared":[120],"Exponent":[121],"Floating":[122],"Point":[123],"(SEFP),":[124],"distinct":[126],"mechanism,":[128],"produce":[130],"bit-widths":[132],"simple":[134],"mantissa":[135],"truncations":[136],"single":[139],"model.":[140],"Moreover,":[141],"achieve":[143],"bit-width":[144],"in":[146],"applications,":[148],"performs":[150,189],"learning":[152],"process":[153],"toward":[154],"losses":[155],"induced":[156],"by":[157],"bit-widths.":[159,198],"The":[160],"involves":[162],"two":[163],"critical":[164],"strategies:":[165],"(1)":[166],"Exploitation-Exploration":[167],"Bit-Width":[168],"Path":[169],"Search":[170],"(BPS),":[171],"which":[172,188],"iteratively":[173],"updates":[174,195],"search":[176],"path":[177],"via":[178],"designed":[180],"scoring":[181],"mechanism;":[182],"(2)":[183],"Low-Precision":[184],"Asynchronous":[185],"Accumulation":[186],"(LAA),":[187],"asynchronous":[190],"gradient":[191],"accumulations":[192],"delayed":[194],"under":[196],"low":[197],"Experiments":[199],"popular":[201],"LLMs,":[202],"LLaMA3.2-1B,":[204],"LLaMA3-8B,":[205],"demonstrate":[206],"achieves":[209],"consistently":[210],"strong":[211],"robust":[213],"for":[215],"all":[216],"precisions.":[217]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-03-18T00:00:00"}