{"id":"https://openalex.org/W7225005901","doi":"https://doi.org/10.5281/zenodo.23826654","title":"Learning Finite-State Transducers Through Their Approximations, with Proof","display_name":"Learning Finite-State Transducers Through Their Approximations, with Proof","publication_year":2026,"publication_date":"2026-10-31","ids":{"openalex":"https://openalex.org/W7225005901","doi":"https://doi.org/10.5281/zenodo.23826654"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23826654","doi":"https://doi.org/10.5281/zenodo.23826654","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23826654","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23826654","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5065197233","display_name":"David Barkhuizen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Barkhuizen, David","raw_affiliation_strings":["Independent researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent researcher","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5065197233"],"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":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11567","display_name":"semigroups and automata theory","score":0.9945999979972839,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11567","display_name":"semigroups and automata theory","score":0.9945999979972839,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T10320","display_name":"Neural Networks and Applications","score":0.0026000000070780516,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.0006000000284984708,"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/finite-state-transducers","display_name":"finite-state transducers","score":0.8790000081062317},{"id":"https://openalex.org/keywords/learning-from-teachers","display_name":"learning from teachers","score":0.375},{"id":"https://openalex.org/keywords/homomorphic-images","display_name":"homomorphic images","score":0.7960000038146973},{"id":"https://openalex.org/keywords/recurrent-neural-networks","display_name":"recurrent neural networks","score":0.46000000834465027},{"id":"https://openalex.org/keywords/modular-division","display_name":"modular division","score":0.49799999594688416},{"id":"https://openalex.org/keywords/move-to-front","display_name":"move-to-front","score":0.46000000834465027},{"id":"https://openalex.org/keywords/s4","display_name":"S4","score":0.3790000081062317},{"id":"https://openalex.org/keywords/s3","display_name":"S3","score":0.7960000038146973},{"id":"https://openalex.org/keywords/claude-code","display_name":"Claude Code","score":0.39500001072883606}],"concepts":[{"id":"https://openalex.org/C199422724","wikidata":"https://www.wikidata.org/wiki/Q41118","display_name":"Quotient","level":2,"score":0.670199990272522},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5501000285148621},{"id":"https://openalex.org/C96865113","wikidata":"https://www.wikidata.org/wiki/Q2946816","display_name":"Certificate","level":2,"score":0.5356000065803528},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.4593000113964081},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.4553000032901764},{"id":"https://openalex.org/C60798267","wikidata":"https://www.wikidata.org/wiki/Q1226939","display_name":"Division (mathematics)","level":2,"score":0.4542999863624573},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.4399000108242035},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.4318999946117401},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3634999990463257},{"id":"https://openalex.org/C2780990831","wikidata":"https://www.wikidata.org/wiki/Q319141","display_name":"Conjecture","level":2,"score":0.35269999504089355},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.34700000286102295},{"id":"https://openalex.org/C2777686260","wikidata":"https://www.wikidata.org/wiki/Q144037","display_name":"Calculus (dental)","level":2,"score":0.3068999946117401},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3027999997138977},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C46304622","wikidata":"https://www.wikidata.org/wiki/Q374814","display_name":"Certification","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.26030001044273376}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23826654","doi":"https://doi.org/10.5281/zenodo.23826654","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23826654","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.23826655","doi":"https://doi.org/10.5281/zenodo.23826655","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23826655","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"doi:10.5281/zenodo.23826654","doi":"https://doi.org/10.5281/zenodo.23826654","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23826654","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7098893823","https://openalex.org/W7099085981","https://openalex.org/W2515411877","https://openalex.org/W7206169185","https://openalex.org/W2613857310","https://openalex.org/W3145490059","https://openalex.org/W7099725122","https://openalex.org/W7087002368","https://openalex.org/W7096194382","https://openalex.org/W2228014137"],"abstract_inverted_index":{"What":[0],"must":[1],"a":[2,9,17,60,63,66,82,166,219,339,354,418,440,493,509],"recurrent":[3],"network":[4,44,232,495],"be":[5,423,506],"told":[6],"to":[7,46,304,477],"learn":[8,89,195],"finite-state":[10,260],"transducer":[11],"exactly?":[12],"We":[13],"train":[14],"it":[15,277,307,333,353,421],"on":[16,179,210,298,492],"target's":[18],"answers":[19,212],"and":[20,62,86,104,113,191,193,198,208,217,265,270,286,325,331,358,373,382,385,397,402,410,464,484,496,527],"some":[21],"of":[22,31,78,106,125,154,169,182,186,200,221,289,391,405],"its":[23,25,32,47,284,287,386,395,408,411,513],"approximations,":[24,205],"homomorphic":[26],"images":[27],"(the":[28],"congruence":[29],"quotients":[30],"states),":[33],"as":[34,146,148,278,417,517],"extra":[35],"outputs.":[36],"A":[37,502],"complete":[38],"certificate":[39,247,491],"then":[40],"proves":[41],"every":[42,50,301,368,392,482],"learned":[43,223,494],"equal":[45],"target":[48],"at":[49,74,110,144,159,314,367],"length,":[51],"so,":[52],"given":[53,81,94,99,203],"the":[54,84,87,90,95,100,117,120,130,137,140,149,174,222,231,246,257,272,328,356,359,361,365,370,375,378,380,389,398,400,403,406,426,453,461,465,470,479,487,490],"target,":[55],"one":[56,230,446],"certified":[57],"run":[58],"settles":[59],"machine,":[61,91],"rate":[64],"is":[65,119,334,414,439],"cost.":[67],"Approximations":[68],"cut":[69],"it.":[70,535],"On":[71,214],"modular":[72,96],"division":[73,215],"121":[75,112],"states,":[76],"16":[77,103,183,197],"20":[79,107,160,201],"runs":[80,133,202,362],"cover,":[83,101],"sum":[85],"difference,":[88],"against":[92,152,189,206],"none":[93],"quotient":[97],"alone;":[98],"17,":[102],"15":[105],"do":[108,134],"so":[109],"49,":[111],"169":[114],"states.":[115],"Whether":[116],"help":[118],"approximations'":[121],"own":[122],"or":[123,512,533],"that":[124,225,351,457],"what":[126,327],"they":[127,243],"tell":[128],"about":[129],"state,":[131],"our":[132],"not":[135,415,442,521,529],"separate:":[136],"state":[138,175,224],"itself,":[139],"finest":[141],"approximation,":[142],"did":[143,364,432],"least":[145],"well":[147],"cover":[150],"(17":[151],"11":[153],"20,":[155],"no":[156,296],"difference":[157],"detected":[158],"seeds":[161],"an":[162,227,348,518],"arm).":[163],"Beyond":[164],"arithmetic,":[165],"planted":[167],"approximation":[168,228],"4":[170,207,209],"blocks,":[171],"which":[172],"leaves":[173],"undetermined,":[176],"given,":[177],"helps":[178],"random":[180],"machines":[181,188,261,360],"states":[184],"(10":[185],"25":[187],"2),":[190],"S4":[192],"move-to-front-4":[194],"in":[196,469,486,508,524],"18":[199],"two":[204],"their":[211,263,294,323],"alone.":[213],"(exploratory)":[216],"S3,":[218],"factor":[220],"holds":[226,229],"was":[233,252],"given;":[234],"elsewhere":[235],"factors":[236],"rarely":[237],"hold":[238],"them.":[239],"Sampled":[240],"tests":[241],"overstate:":[242],"pass":[244],"networks":[245],"refutes.":[248],"How":[249],"this":[250,428,448,525],"work":[251,526],"done.":[253],"David":[254],"Barkhuizen":[255],"posed":[256],"question,":[258],"learning":[259],"from":[262],"behaviour":[264],"known":[266],"structure,":[267],"with":[268,322],"proof,":[269],"directed":[271],"research":[273],"programme":[274],"throughout,":[275],"steering":[276],"results":[279],"came":[280],"in:":[281],"he":[282],"set":[283],"goal":[285],"rules":[288],"evidence":[290],"(claims":[291],"registered":[292],"before":[293,394],"data,":[295,396],"acting":[297],"early":[299],"looks,":[300],"cell":[302],"able":[303],"succeed),":[305],"redirected":[306],"when":[308],"findings":[309],"called":[310],"for":[311,425,500],"it,":[312],"decided":[313,326],"each":[315],"decision":[316],"point":[317],"among":[318],"options":[319],"laid":[320],"out":[321],"costs,":[324],"paper":[329],"claims":[330],"how":[332,476],"published.":[335],"Claude":[336,346,413],"(Opus":[337],"5.5,":[338],"model":[340],"made":[341],"by":[342],"Anthropic),":[343],"working":[344],"through":[345],"Code,":[347],"agentic":[349],"harness":[350],"gave":[352],"shell,":[355],"repositories":[357],"used,":[363],"rest,":[366],"stage:":[369],"literature":[371],"search":[372],"reading;":[374],"method's":[376],"design;":[377],"code,":[379],"training":[381,498],"certification":[383],"library":[384],"native":[387],"crate;":[388],"registration":[390],"test":[393],"runs;":[399],"analysis;":[401],"drafting":[404],"paper,":[407,488],"tables":[409],"figures.":[412],"listed":[416],"creator":[419],"because":[420],"cannot":[422],"accountable":[424],"record;":[427],"note":[429],"says":[430,475],"who":[431],"what.":[433],"Not":[434],"yet":[435,451],"independently":[436],"checked.":[437],"This":[438],"preprint,":[441],"peer":[443],"reviewed.":[444],"No":[445],"outside":[447],"collaboration":[449],"has":[450,520,528],"checked":[452],"findings;":[454],"we":[455],"invite":[456],"check.":[458],"The":[459],"networks,":[460],"result":[462],"files":[463],"code":[466],"are":[467],"released":[468],"datapack":[471],"(doi:10.5281/zenodo.23733575),":[472],"whose":[473],"README":[474],"check":[478],"files,":[480],"regenerate":[481],"number":[483],"table":[485],"rerun":[489,497],"bit":[499],"bit.":[501],"fault":[503],"found":[504],"will":[505],"corrected":[507],"new":[510],"version,":[511],"claim":[514],"withdrawn.":[515],"Anthropic":[516],"organisation":[519],"been":[522],"involved":[523],"consented":[530],"to,":[531],"reviewed":[532],"endorsed":[534]},"counts_by_year":[],"updated_date":"2026-11-02T09:45:03.842371","created_date":"2026-11-01T00:00:00"}