{"id":"https://openalex.org/W7220920701","doi":"https://doi.org/10.48550/arxiv.2610.10437","title":"Q-Learning with Scalar Adjoint Matching","display_name":"Q-Learning with Scalar Adjoint Matching","publication_year":2026,"publication_date":"2026-10-07","ids":{"openalex":"https://openalex.org/W7220920701","doi":"https://doi.org/10.48550/arxiv.2610.10437"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.10437","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.10437","pdf_url":"https://arxiv.org/pdf/2610.10437","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.10437","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5157888784","display_name":"Yonghoon Dong","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Dong, Yonghoon","raw_affiliation_strings":["KAIST ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071352137","display_name":"Minsung Yoon","orcid":"https://orcid.org/0000-0001-9860-9647","observed_orcids":["https://orcid.org/0000-0001-9860-9647","https://orcid.org/0000-0002-1440-6656"]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yoon, Minsung","raw_affiliation_strings":["KAIST ,","RLWRLD"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]},{"raw_affiliation_string":"RLWRLD","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137092191","display_name":"Jaehyuk Kim","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Kim, Jaehyuk","raw_affiliation_strings":["KAIST ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134796330","display_name":"Jungwoo Park","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Park, Jungwoo","raw_affiliation_strings":["KAIST ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5157956238","display_name":"Changyeon Kim","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Kim, Changyeon","raw_affiliation_strings":["KAIST ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102928677","display_name":"Jinwoo Shin","orcid":"https://orcid.org/0000-0003-4313-4669","observed_orcids":["https://orcid.org/0000-0003-4313-4669"]},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Shin, Jinwoo","raw_affiliation_strings":["KAIST ,","RLWRLD"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KAIST ,","institution_ids":["https://openalex.org/I157485424"]},{"raw_affiliation_string":"RLWRLD","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157485424"],"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9984999895095825,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9984999895095825,"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/T12794","display_name":"Adaptive Dynamic Programming Control","score":0.00039999998989515007,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":9.999999747378752e-05,"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/q-learning","display_name":"Q-learning","score":0.6480000019073486},{"id":"https://openalex.org/keywords/vision-language-action-policies","display_name":"vision-language-action policies","score":0.8539999723434448}],"concepts":[{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.7120000123977661},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5687000155448914},{"id":"https://openalex.org/C200331156","wikidata":"https://www.wikidata.org/wiki/Q506041","display_name":"Jacobian matrix and determinant","level":2,"score":0.5508999824523926},{"id":"https://openalex.org/C38349280","wikidata":"https://www.wikidata.org/wiki/Q1434290","display_name":"Flow (mathematics)","level":2,"score":0.5407999753952026},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.49000000953674316},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.46050000190734863},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.4325999915599823},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4318999946117401},{"id":"https://openalex.org/C14646407","wikidata":"https://www.wikidata.org/wiki/Q1430750","display_name":"Bellman equation","level":2,"score":0.3831999897956848},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.3682999908924103},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.3465000092983246},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3160000145435333},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.3158000111579895},{"id":"https://openalex.org/C110521144","wikidata":"https://www.wikidata.org/wiki/Q193460","display_name":"Scalar field","level":2,"score":0.3070000112056732},{"id":"https://openalex.org/C167879884","wikidata":"https://www.wikidata.org/wiki/Q727568","display_name":"Balanced flow","level":2,"score":0.2865000069141388},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.2759999930858612}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.10437","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.10437","pdf_url":"https://arxiv.org/pdf/2610.10437","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.2610.10437","doi":"https://doi.org/10.48550/arxiv.2610.10437","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.10437","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:2610.10437","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.10437","pdf_url":"https://arxiv.org/pdf/2610.10437","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/W7220920701.pdf","grobid_xml":"https://content.openalex.org/works/W7220920701.grobid-xml"},"referenced_works_count":10,"referenced_works":["https://openalex.org/W2741122588","https://openalex.org/W4327673663","https://openalex.org/W4366850766","https://openalex.org/W4414901575","https://openalex.org/W4416967268","https://openalex.org/W7157012058","https://openalex.org/W7160425969","https://openalex.org/W7162393363","https://openalex.org/W7165021285","https://openalex.org/W7204808299"],"related_works":["https://openalex.org/W7214728399","https://openalex.org/W7162490384","https://openalex.org/W7125202567","https://openalex.org/W7164078567","https://openalex.org/W4407186878","https://openalex.org/W7167712518","https://openalex.org/W7161044806","https://openalex.org/W7161158341","https://openalex.org/W4415057179","https://openalex.org/W7160596644"],"abstract_inverted_index":{"Flow":[0],"policies":[1,107],"capture":[2],"rich":[3],"and":[4,8,93],"diverse":[5],"action":[6,41,65,130],"distributions,":[7],"fine-tuning":[9,24,234],"them":[10],"with":[11,87,165,175],"off-policy":[12],"RL":[13],"to":[14,52,67,206,215],"improve":[15],"beyond":[16],"the":[17,37,54,63,78,88,94,100,124,128,132,136,145,155,172,186,198],"demonstrations":[18],"has":[19],"drawn":[20],"growing":[21],"interest.":[22],"However,":[23],"a":[25,29,49,74,83,118,176,222,226],"flow":[26,44,55,69,91,106,133],"policy":[27,38,79,95,224],"against":[28],"learned":[30],"value":[31,60,125,147,177],"function":[32],"is":[33,151],"not":[34],"trivial,":[35],"because":[36],"generates":[39],"its":[40,110,192],"over":[42,232],"many":[43],"steps.":[45],"Adjoint":[46,167],"matching":[47],"offers":[48],"principled":[50],"way":[51],"update":[53],"model":[56],"itself":[57],"by":[58,113,131,204],"propagating":[59],"information":[61],"from":[62],"final":[64,129],"back":[66],"each":[68,202],"step,":[70,82],"but":[71],"it":[72],"requires":[73],"vector--Jacobian":[75,138],"product":[76],"through":[77],"at":[80,127,148,179],"every":[81],"cost":[84],"that":[85,99,122,143,196],"grows":[86],"number":[89],"of":[90,104,197],"steps":[92],"size.":[96],"We":[97,140],"observe":[98],"batch-averaged":[101],"velocity":[102],"Jacobian":[103],"pretrained":[105,217],"concentrates":[108],"on":[109,159,185,225,235],"diagonal.":[111],"Motivated":[112],"this":[114],"finding,":[115],"we":[116,162,219],"derive":[117],"closed-form":[119],"scalar":[120,156,173],"adjoint":[121,174],"scales":[123],"gradient":[126],"time,":[134],"eliminating":[135],"per-step":[137],"products.":[139],"further":[141],"find":[142],"controlling":[144],"critic's":[146],"policy-generated":[149],"actions":[150],"particularly":[152],"important":[153],"under":[154],"adjoint.":[157],"Based":[158],"these":[160],"findings,":[161],"propose":[163],"Q-learning":[164],"Scalar":[166],"Matching":[168],"(SQAM),":[169],"which":[170],"combines":[171],"penalty":[178],"those":[180],"actions.":[181],"SQAM's":[182],"gains":[183],"concentrate":[184],"four":[187],"hardest":[188],"OGBench":[189],"domains,":[190],"where":[191],"success":[193],"rate":[194],"exceeds":[195],"strongest":[199],"baseline":[200],"in":[201],"domain":[203],"18":[205],"35":[207],"percentage":[208],"points.":[209],"To":[210],"test":[211],"whether":[212],"SQAM":[213,230],"extends":[214],"large":[216],"policies,":[218],"also":[220],"fine-tune":[221],"vision-language-action":[223],"real":[227],"bimanual":[228],"robot.":[229],"improves":[231],"supervised":[233],"all":[236],"three":[237],"tasks.":[238]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-09T00:00:00"}