{"id":"https://openalex.org/W7214624573","doi":"https://doi.org/10.48550/arxiv.2609.31066","title":"Modeling quantum neural network gradient with reinforcement learning","display_name":"Modeling quantum neural network gradient with reinforcement learning","publication_year":2026,"publication_date":"2026-09-25","ids":{"openalex":"https://openalex.org/W7214624573","doi":"https://doi.org/10.48550/arxiv.2609.31066"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.31066","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31066","pdf_url":"https://arxiv.org/pdf/2609.31066","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/2609.31066","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5090044114","display_name":"Nhan Trong Luu","orcid":"https://orcid.org/0009-0000-6783-6257","observed_orcids":["https://orcid.org/0009-0000-6783-6257"]},"institutions":[{"id":"https://openalex.org/I177733328","display_name":"Can Tho University","ror":"https://ror.org/0071qz696","country_code":"VN","type":"education","lineage":["https://openalex.org/I177733328"]},{"id":"https://openalex.org/I4210142044","display_name":"VinUniversity","ror":"https://ror.org/052dmdr17","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210142044"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Luu, Nhan Trong","raw_affiliation_strings":["College of Engineering and Computer Science , VinUniversity , Hanoi , Vietnam","College of Information and Communication Technology , Can Tho University , Can Tho , Vietnam"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"College of Engineering and Computer Science , VinUniversity , Hanoi , Vietnam","institution_ids":["https://openalex.org/I4210142044"]},{"raw_affiliation_string":"College of Information and Communication Technology , Can Tho University , Can Tho , Vietnam","institution_ids":["https://openalex.org/I177733328"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153228458","display_name":"Duong Trung Luu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I177733328","display_name":"Can Tho University","ror":"https://ror.org/0071qz696","country_code":"VN","type":"education","lineage":["https://openalex.org/I177733328"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Luu, Duong Trung","raw_affiliation_strings":["Center for Digital Transformation and Communications , Can Tho University , Can Tho , Vietnam"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Center for Digital Transformation and Communications , Can Tho University , Can Tho , Vietnam","institution_ids":["https://openalex.org/I177733328"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126667328","display_name":"Nam Ngoc Pham","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pham, Nam Ngoc","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5110070160","display_name":"Thang Cong Truong","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Truong, Thang Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"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/T10682","display_name":"Quantum Computing Algorithms and Architecture","score":1.0,"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/T10682","display_name":"Quantum Computing Algorithms and Architecture","score":1.0,"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/T11804","display_name":"Quantum many-body systems","score":0.0,"subfield":{"id":"https://openalex.org/subfields/3107","display_name":"Atomic and Molecular Physics, and Optics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.0,"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/quantum-neural-networks","display_name":"quantum neural networks","score":0.9919999837875366},{"id":"https://openalex.org/keywords/barren-plateaus","display_name":"barren plateaus","score":0.8650000095367432},{"id":"https://openalex.org/keywords/surrogate-gradient","display_name":"surrogate gradient","score":0.5709999799728394},{"id":"https://openalex.org/keywords/gradient-variance","display_name":"gradient variance","score":0.6449999809265137},{"id":"https://openalex.org/keywords/parameter-shift-rule","display_name":"parameter-shift rule","score":0.6650000214576721},{"id":"https://openalex.org/keywords/adjoint-differentiation","display_name":"adjoint differentiation","score":0.9150000214576721},{"id":"https://openalex.org/keywords/policy-gradient","display_name":"policy gradient","score":0.531000018119812},{"id":"https://openalex.org/keywords/ppo","display_name":"PPO","score":0.5479999780654907},{"id":"https://openalex.org/keywords/hardware-efficient-ansatz","display_name":"hardware-efficient ansatz","score":0.7620000243186951},{"id":"https://openalex.org/keywords/quantum-machine-learning","display_name":"quantum machine learning","score":0.46799999475479126}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6327999830245972},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6172999739646912},{"id":"https://openalex.org/C130979935","wikidata":"https://www.wikidata.org/wiki/Q568954","display_name":"Ansatz","level":2,"score":0.5861999988555908},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.5708000063896179},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.5295000076293945},{"id":"https://openalex.org/C2780030769","wikidata":"https://www.wikidata.org/wiki/Q4968575","display_name":"Plateau (mathematics)","level":2,"score":0.5095999836921692},{"id":"https://openalex.org/C84114770","wikidata":"https://www.wikidata.org/wiki/Q46344","display_name":"Quantum","level":2,"score":0.49869999289512634},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.49810001254081726},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4706000089645386},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.46939998865127563},{"id":"https://openalex.org/C67820243","wikidata":"https://www.wikidata.org/wiki/Q179164","display_name":"Unitary state","level":2,"score":0.4336000084877014},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.423799991607666},{"id":"https://openalex.org/C151376022","wikidata":"https://www.wikidata.org/wiki/Q168698","display_name":"Exponential function","level":2,"score":0.42010000348091125},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.37700000405311584},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.3767000138759613},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.3684000074863434},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.36629998683929443},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3580999970436096},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3474000096321106},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.34689998626708984},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.31040000915527344},{"id":"https://openalex.org/C58053490","wikidata":"https://www.wikidata.org/wiki/Q176555","display_name":"Quantum computer","level":3,"score":0.3034999966621399},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.299699991941452},{"id":"https://openalex.org/C75235859","wikidata":"https://www.wikidata.org/wiki/Q582659","display_name":"Exponential growth","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C115680565","wikidata":"https://www.wikidata.org/wiki/Q5977448","display_name":"Gradient method","level":2,"score":0.2815999984741211},{"id":"https://openalex.org/C83581934","wikidata":"https://www.wikidata.org/wiki/Q527381","display_name":"Arbitrary-precision arithmetic","level":2,"score":0.2766000032424927},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.2743000090122223},{"id":"https://openalex.org/C75438885","wikidata":"https://www.wikidata.org/wiki/Q3403615","display_name":"Large deviations theory","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C148043351","wikidata":"https://www.wikidata.org/wiki/Q4456944","display_name":"Current (fluid)","level":2,"score":0.27149999141693115},{"id":"https://openalex.org/C62644790","wikidata":"https://www.wikidata.org/wiki/Q3454689","display_name":"Variance reduction","level":3,"score":0.27079999446868896}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.31066","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31066","pdf_url":"https://arxiv.org/pdf/2609.31066","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.2609.31066","doi":"https://doi.org/10.48550/arxiv.2609.31066","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.31066","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.31066","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31066","pdf_url":"https://arxiv.org/pdf/2609.31066","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214624573.pdf","grobid_xml":"https://content.openalex.org/works/W7214624573.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W2518108298","https://openalex.org/W2750384547","https://openalex.org/W2901859177","https://openalex.org/W2958921151","https://openalex.org/W2964121744","https://openalex.org/W2984991932","https://openalex.org/W3083594649","https://openalex.org/W3088321919","https://openalex.org/W3139389937","https://openalex.org/W3153434261","https://openalex.org/W3197971197","https://openalex.org/W4324106947","https://openalex.org/W4392678006","https://openalex.org/W4403346009","https://openalex.org/W4403618225","https://openalex.org/W6926238688","https://openalex.org/W7196928691"],"related_works":["https://openalex.org/W4287250207","https://openalex.org/W7171994585","https://openalex.org/W3098306754","https://openalex.org/W4366341679","https://openalex.org/W7197045232","https://openalex.org/W7163919957","https://openalex.org/W4225864255","https://openalex.org/W4384644635","https://openalex.org/W4364377871","https://openalex.org/W7222938421"],"abstract_inverted_index":{"Training":[0],"quantum":[1],"neural":[2],"networks":[3],"(QNNs)":[4],"on":[5,64,127,206,220],"near-term":[6],"hardware":[7],"remains":[8],"hampered":[9],"by":[10,80,87,98,153,199],"two":[11],"compounding":[12],"difficulties:":[13],"the":[14,22,26,65,75,90,99,109,116,162],"exponential":[15],"vanishing":[16],"of":[17,34,111,158,178,208],"gradient":[18,77],"variance":[19,94],"known":[20],"as":[21],"barren":[23,100,216],"plateau,":[24],"and":[25,31,71,104,124,149,170,180,184,213,229],"$\\mathcal{O}(L":[27],"\\cdot":[28],"2^n)$":[29],"time":[30],"memory":[32,179],"cost":[33,106],"differentiating":[35,88],"through":[36,89],"an":[37],"$n$-qubit,":[38],"$L$-layer":[39],"circuit.":[40],"We":[41,119],"propose":[42,59],"RLQ-Grad,":[43],"a":[44,49,81,128,142],"reinforcement-learning-based":[45],"optimizer":[46,171],"in":[47],"which":[48],"classical":[50,82],"policy":[51],"$\u03c0_\u03d5$":[52],"(a":[53],"spectrally-normalized":[54],"PPO":[55],"agent)":[56],"learns":[57],"to":[58,137,155,201,210,224,233],"parameter":[60],"updates":[61],"directly,":[62],"conditioned":[63],"QNN's":[66],"current":[67],"parameters,":[68],"loss,":[69],"accuracy,":[70],"previous":[72],"update.":[73],"Because":[74],"surrogate":[76],"is":[78,95],"emitted":[79],"network":[83],"rather":[84,114],"than":[85,115,189],"obtained":[86],"unitary":[91],"$U(\u03b8)$,":[92],"its":[93,105],"not":[96],"constrained":[97],"plateau":[101,217],"concentration":[102],"bound,":[103],"scales":[107],"with":[108,135],"number":[110],"trainable":[112],"parameters":[113],"Hilbert-space":[117],"dimension.":[118],"prove":[120],"these":[121,190],"properties":[122],"formally":[123],"verify":[125],"them":[126],"hardware-efficient":[129],"ansatz":[130],"across":[131],"four":[132],"supervised":[133],"benchmarks":[134],"up":[136,200,209],"$n=20$":[138],"qubits.":[139],"RLQ-Grad":[140,173],"preserves":[141],"near-flat":[143],"gradient-variance":[144],"curve":[145],"where":[146,227],"backpropagation,":[147],"parameter-shift,":[148],"adjoint":[150],"differentiation":[151],"decay":[152],"1":[154],"2":[156,176],"orders":[157],"magnitude.":[159],"Accounting":[160],"for":[161],"full":[163],"training":[164],"pipeline":[165],"(PPO":[166],"rollouts,":[167],"actor-critic":[168],"updates,":[169],"states),":[172],"needs":[174],"under":[175],"MB":[177],"runs":[181],"$2490\\times$,":[182],"$7876\\times$,":[183],"$673\\times$":[185],"faster":[186],"per":[187],"iteration":[188],"three":[191],"methods":[192,219],"at":[193,222],"$n=20$.":[194],"It":[195],"improves":[196],"top-1":[197],"accuracy":[198],"$+10\\%$":[202],"over":[203],"gradient-based":[204],"baselines":[205],"circuits":[207],"12":[211],"qubits,":[212,226],"matches":[214],"dedicated":[215],"mitigation":[218],"CIFAR-10":[221],"14":[223],"20":[225],"evolutionary":[228],"gradient-free":[230],"optimizers":[231],"collapse":[232],"chance.":[234]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-29T00:00:00"}