{"id":"https://openalex.org/W4379924477","doi":"https://doi.org/10.48550/arxiv.2306.03521","title":"Machine learning in and out of equilibrium","display_name":"Machine learning in and out of equilibrium","publication_year":2023,"publication_date":"2023-06-06","ids":{"openalex":"https://openalex.org/W4379924477","doi":"https://doi.org/10.48550/arxiv.2306.03521"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2306.03521","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2306.03521","pdf_url":"https://arxiv.org/pdf/2306.03521","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","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/2306.03521","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5112360435","display_name":"Shishir Adhikari","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Adhikari, Shishir","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041370925","display_name":"Alkan Kabak\u00e7\u0131o\u011flu","orcid":"https://orcid.org/0000-0002-9831-3632","observed_orcids":["https://orcid.org/0000-0002-9831-3632"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kabak\u00e7\u0131o\u011flu, Alkan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036898670","display_name":"Alexander Strang","orcid":"https://orcid.org/0000-0001-6618-631X","observed_orcids":["https://orcid.org/0000-0001-6618-631X"]},"institutions":[{"id":"https://openalex.org/I40347166","display_name":"University of Chicago","ror":"https://ror.org/024mw5h28","country_code":"US","type":"education","lineage":["https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Strang, Alexander","raw_affiliation_strings":["Department of Statistics , University of Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics , University of Chicago","institution_ids":["https://openalex.org/I40347166"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013522667","display_name":"Deniz Y\u00fcret","orcid":"https://orcid.org/0000-0002-7039-0046","observed_orcids":["https://orcid.org/0000-0002-7039-0046"]},"institutions":[{"id":"https://openalex.org/I1351752","display_name":"Ko\u00e7 University","ror":"https://ror.org/00jzwgz36","country_code":"TR","type":"education","lineage":["https://openalex.org/I1351752"]}],"countries":["TR"],"is_corresponding":false,"raw_author_name":"Yuret, Deniz","raw_affiliation_strings":["Department of Computer Engineering , Koc \u00b8University , Istanbul , Turkey","Koc \u00b8University , KUIS AI Center , Istanbul , Turkey"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Engineering , Koc \u00b8University , Istanbul , Turkey","institution_ids":["https://openalex.org/I1351752"]},{"raw_affiliation_string":"Koc \u00b8University , KUIS AI Center , Istanbul , Turkey","institution_ids":["https://openalex.org/I1351752"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5051732988","display_name":"Michael Hinczewski","orcid":"https://orcid.org/0000-0003-2837-7697","observed_orcids":["https://orcid.org/0000-0003-2837-7697"]},"institutions":[{"id":"https://openalex.org/I58956616","display_name":"Case Western Reserve University","ror":"https://ror.org/051fd9666","country_code":"US","type":"education","lineage":["https://openalex.org/I58956616"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hinczewski, Michael","raw_affiliation_strings":["Department of Physics , Case Western Reserve University , Cleveland , OH , U.S.A"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Physics , Case Western Reserve University , Cleveland , OH , U.S.A","institution_ids":["https://openalex.org/I58956616"]}]}],"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":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/T11520","display_name":"Advanced Thermodynamics and Statistical Mechanics","score":0.4514000117778778,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11520","display_name":"Advanced Thermodynamics and Statistical Mechanics","score":0.4514000117778778,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.34769999980926514,"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/T12056","display_name":"Markov Chains and Monte Carlo Methods","score":0.1177000030875206,"subfield":{"id":"https://openalex.org/subfields/2613","display_name":"Statistics and Probability"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"stochastic gradient descent","score":0.8659999966621399},{"id":"https://openalex.org/keywords/non-equilibrium-steady-state","display_name":"nonequilibrium steady states","score":0.7300000190734863},{"id":"https://openalex.org/keywords/entropy-generation","display_name":"entropy production","score":0.6600000262260437},{"id":"https://openalex.org/keywords/fluctuation-theorems","display_name":"fluctuation theorems","score":0.9739999771118164},{"id":"https://openalex.org/keywords/fokker-planck-equation","display_name":"Fokker-Planck equation","score":0.8539999723434448},{"id":"https://openalex.org/keywords/persistent-currents","display_name":"persistent currents","score":0.3959999978542328},{"id":"https://openalex.org/keywords/loss-landscape","display_name":"loss landscape","score":0.6759999990463257},{"id":"https://openalex.org/keywords/diffusion-matrix","display_name":"diffusion matrix","score":0.6949999928474426},{"id":"https://openalex.org/keywords/stochastic-gradient-langevin-dynamics","display_name":"stochastic gradient Langevin dynamics","score":0.9240000247955322},{"id":"https://openalex.org/keywords/bayesian-posterior-sampling","display_name":"Bayesian posterior sampling","score":0.6790000200271606},{"id":"https://openalex.org/keywords/neural-network-training-dynamics","display_name":"neural network training dynamics","score":0.5630000233650208}],"concepts":[{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.5915040969848633},{"id":"https://openalex.org/C98951983","wikidata":"https://www.wikidata.org/wiki/Q7604341","display_name":"Stationary distribution","level":3,"score":0.5786156058311462},{"id":"https://openalex.org/C190502265","wikidata":"https://www.wikidata.org/wiki/Q17069496","display_name":"MNIST database","level":3,"score":0.5508463978767395},{"id":"https://openalex.org/C2780004032","wikidata":"https://www.wikidata.org/wiki/Q6485978","display_name":"Langevin dynamics","level":2,"score":0.5095399022102356},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4510497450828552},{"id":"https://openalex.org/C47458137","wikidata":"https://www.wikidata.org/wiki/Q911951","display_name":"Stationary state","level":2,"score":0.4456402659416199},{"id":"https://openalex.org/C74859849","wikidata":"https://www.wikidata.org/wiki/Q2466686","display_name":"Non-equilibrium thermodynamics","level":2,"score":0.41869181394577026},{"id":"https://openalex.org/C72434380","wikidata":"https://www.wikidata.org/wiki/Q230930","display_name":"State space","level":2,"score":0.41642439365386963},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.37929755449295044},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3499819338321686},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.32780805230140686},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3230644762516022},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.31428730487823486},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26009294390678406},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.1418229639530182},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.10662046074867249},{"id":"https://openalex.org/C98763669","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov chain","level":2,"score":0.0},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2306.03521","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2306.03521","pdf_url":"https://arxiv.org/pdf/2306.03521","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2306.03521","doi":"https://doi.org/10.48550/arxiv.2306.03521","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2306.03521","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:2306.03521","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2306.03521","pdf_url":"https://arxiv.org/pdf/2306.03521","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320322626","display_name":"T\u00fcrkiye Bilimsel ve Teknolojik Ara\u015ft\u0131rma Kurumu","ror":"https://ror.org/04w9kkr77"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4379924477.pdf","grobid_xml":"https://content.openalex.org/works/W4379924477.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2893110193","https://openalex.org/W4411087767","https://openalex.org/W7162141282","https://openalex.org/W2790013445","https://openalex.org/W4404571285","https://openalex.org/W4416159192","https://openalex.org/W6966785410","https://openalex.org/W2906991194","https://openalex.org/W3102766317","https://openalex.org/W3207512848"],"abstract_inverted_index":{"The":[0,101],"algorithms":[1],"used":[2],"to":[3,15,41,157,198,252],"train":[4],"neural":[5],"networks,":[6],"like":[7],"stochastic":[8,226],"gradient":[9,227],"descent":[10],"(SGD),":[11],"have":[12],"close":[13],"parallels":[14,44],"natural":[16],"processes":[17],"that":[18,153,169,231],"navigate":[19],"a":[20,34,46,131,205,210,222,260],"high-dimensional":[21],"parameter":[22],"space":[23,78],"--":[24,114],"for":[25,96,204],"example":[26,238],"protein":[27],"folding":[28],"or":[29,187],"evolution.":[30],"Our":[31],"study":[32],"uses":[33],"Fokker-Planck":[35],"approach,":[36],"adapted":[37],"from":[38,209],"statistical":[39],"physics,":[40],"explore":[42],"these":[43,105,125],"in":[45,52,61,66,76,83,127,216],"single,":[47],"unified":[48],"framework.":[49],"We":[50,123,190,220],"focus":[51],"particular":[53,206],"on":[54,179],"the":[55,59,62,77,87,108,118,140,150,158,162,171,174,180,241,253,263],"stationary":[56,102,151,175,202],"state":[57,152,203],"of":[58,71,79,104,117,121,149,173,183,194,213,225,256,262],"system":[60,239],"long-time":[63],"limit,":[64],"which":[65],"conventional":[67],"SGD":[68],"is":[69,89,243],"out":[70],"equilibrium,":[72],"exhibiting":[73],"persistent":[74],"currents":[75],"network":[80,134,214],"parameters.":[81],"As":[82],"its":[84],"physical":[85],"analogues,":[86],"current":[88],"associated":[90],"with":[91,186],"an":[92,200,237],"entropy":[93],"production":[94],"rate":[95],"any":[97],"given":[98],"training":[99,159],"trajectory.":[100],"distribution":[103,176,212],"rates":[106],"obeys":[107],"integral":[109],"and":[110,135,166],"detailed":[111],"fluctuation":[112,141],"theorems":[113,142],"nonequilibrium":[115,196],"generalizations":[116],"second":[119],"law":[120],"thermodynamics.":[122],"validate":[124],"relations":[126],"two":[128],"numerical":[129],"examples,":[130],"nonlinear":[132],"regression":[133],"MNIST":[136],"digit":[137],"classification.":[138],"While":[139],"are":[143,146,154],"universal,":[144],"there":[145],"other":[147],"aspects":[148],"highly":[155],"sensitive":[156],"details.":[160],"Surprisingly,":[161],"effective":[163],"loss":[164],"landscape":[165],"diffusion":[167],"matrix":[168],"determine":[170],"shape":[172],"vary":[177],"depending":[178],"simple":[181],"choice":[182],"minibatching":[184],"done":[185],"without":[188,233],"replacement.":[189],"can":[191],"take":[192],"advantage":[193],"this":[195,246],"sensitivity":[197],"engineer":[199],"equilibrium":[201],"application:":[207],"sampling":[208],"posterior":[211,242,254],"weights":[215],"Bayesian":[217],"machine":[218],"learning.":[219],"propose":[221],"new":[223],"variation":[224],"Langevin":[228],"dynamics":[229],"(SGLD)":[230],"harnesses":[232],"replacement":[234],"minibatching.":[235],"In":[236],"where":[240],"exactly":[244],"known,":[245],"SGWORLD":[247],"algorithm":[248],"outperforms":[249],"SGLD,":[250],"converging":[251],"orders":[255],"magnitude":[257],"faster":[258],"as":[259],"function":[261],"learning":[264],"rate.":[265]},"counts_by_year":[],"updated_date":"2026-09-30T09:27:27.532063","created_date":"2023-06-09T00:00:00"}