{"id":"https://openalex.org/W7102392873","doi":"https://doi.org/10.48550/arxiv.2510.24427","title":"SynthWorlds: Controlled Parallel Worlds for Disentangling Reasoning and Knowledge in Language Models","display_name":"SynthWorlds: Controlled Parallel Worlds for Disentangling Reasoning and Knowledge in Language Models","publication_year":2025,"publication_date":"2025-10-28","ids":{"openalex":"https://openalex.org/W7102392873","doi":"https://doi.org/10.48550/arxiv.2510.24427"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2510.24427","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.24427","pdf_url":"https://arxiv.org/pdf/2510.24427","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":"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/2510.24427","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Gu, Ken","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I201448701","display_name":"University of Washington","ror":"https://ror.org/00cvxb145","country_code":"US","type":"education","lineage":["https://openalex.org/I201448701"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gu, Ken","raw_affiliation_strings":["University of Washington ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Washington ,","institution_ids":["https://openalex.org/I201448701"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047190193","display_name":"Advait Bhat","orcid":"https://orcid.org/0000-0002-5524-2387","observed_orcids":["https://orcid.org/0000-0002-5524-2387"]},"institutions":[{"id":"https://openalex.org/I201448701","display_name":"University of Washington","ror":"https://ror.org/00cvxb145","country_code":"US","type":"education","lineage":["https://openalex.org/I201448701"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bhat, Advait","raw_affiliation_strings":["University of Washington ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Washington ,","institution_ids":["https://openalex.org/I201448701"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009772512","display_name":"Mike A. Merrill","orcid":"https://orcid.org/0000-0003-2994-0058","observed_orcids":["https://orcid.org/0000-0003-2994-0058"]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Merrill, Mike A","raw_affiliation_strings":["Stanford University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University ,","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101446790","display_name":"Robert West","orcid":"https://orcid.org/0000-0002-3984-1232","observed_orcids":["https://orcid.org/0000-0002-3984-1232"]},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"West, Robert","raw_affiliation_strings":["EPFL ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"EPFL ,","institution_ids":["https://openalex.org/I5124864"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100352325","display_name":"Xin Liu","orcid":"https://orcid.org/0000-0003-3275-4125","observed_orcids":["https://orcid.org/0000-0003-3275-4125"]},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Xin","raw_affiliation_strings":["Google Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Research","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100681741","display_name":"Daniel McDuff","orcid":"https://orcid.org/0000-0001-7313-0082","observed_orcids":["https://orcid.org/0000-0001-7313-0082"]},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"McDuff, Daniel","raw_affiliation_strings":["Google Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Research","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5033639139","display_name":"Tim Althoff","orcid":"https://orcid.org/0000-0003-4793-2289","observed_orcids":["https://orcid.org/0000-0003-4793-2289"]},"institutions":[{"id":"https://openalex.org/I201448701","display_name":"University of Washington","ror":"https://ror.org/00cvxb145","country_code":"US","type":"education","lineage":["https://openalex.org/I201448701"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Althoff, Tim","raw_affiliation_strings":["University of Washington ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Washington ,","institution_ids":["https://openalex.org/I201448701"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"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":true,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9312000274658203,"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/T10028","display_name":"Topic Modeling","score":0.9312000274658203,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.02759999968111515,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.01979999989271164,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/language-model-reasoning","display_name":"language model reasoning","score":0.6610000133514404},{"id":"https://openalex.org/keywords/multiverse","display_name":"multiverse","score":0.503000020980835},{"id":"https://openalex.org/keywords/multi-hop-question-answering","display_name":"multi-hop question answering","score":0.8889999985694885},{"id":"https://openalex.org/keywords/page-navigation","display_name":"page navigation","score":0.8339999914169312},{"id":"https://openalex.org/keywords/retrieval-augmented-generation","display_name":"retrieval-augmented generation","score":0.890999972820282},{"id":"https://openalex.org/keywords/closed-book-question-answering","display_name":"closed-book question answering","score":0.9089999794960022}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7817000150680542},{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.6646000146865845},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.6643000245094299},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.6394000053405762},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5892000198364258},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.569100022315979},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5340999960899353},{"id":"https://openalex.org/C117251300","wikidata":"https://www.wikidata.org/wiki/Q1849855","display_name":"Parametric statistics","level":2,"score":0.4975999891757965},{"id":"https://openalex.org/C100660578","wikidata":"https://www.wikidata.org/wiki/Q18733","display_name":"Recall","level":2,"score":0.44440001249313354},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.39070001244544983},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.38269999623298645},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3806999921798706},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36880001425743103},{"id":"https://openalex.org/C2777220311","wikidata":"https://www.wikidata.org/wiki/Q6423340","display_name":"Knowledge acquisition","level":2,"score":0.364300012588501},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.3587000072002411},{"id":"https://openalex.org/C24574437","wikidata":"https://www.wikidata.org/wiki/Q7135228","display_name":"Parametric model","level":3,"score":0.33869999647140503},{"id":"https://openalex.org/C161301231","wikidata":"https://www.wikidata.org/wiki/Q3478658","display_name":"Knowledge representation and reasoning","level":2,"score":0.33649998903274536},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.30959999561309814},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.3068999946117401},{"id":"https://openalex.org/C66024118","wikidata":"https://www.wikidata.org/wiki/Q1122506","display_name":"Computational model","level":2,"score":0.29260000586509705},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.2777999937534332},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.2549000084400177},{"id":"https://openalex.org/C115925183","wikidata":"https://www.wikidata.org/wiki/Q1412694","display_name":"Knowledge-based systems","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2510.24427","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.24427","pdf_url":"https://arxiv.org/pdf/2510.24427","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2510.24427","doi":"https://doi.org/10.48550/arxiv.2510.24427","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.24427","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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2510.24427","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.24427","pdf_url":"https://arxiv.org/pdf/2510.24427","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"score":0.5759949684143066,"display_name":"Quality education","id":"https://metadata.un.org/sdg/4"}],"study_designs":[],"awards":[{"id":"https://openalex.org/G1465050376","display_name":"CAREER: Realizing the Potential of Behavioral Data Science for Population Health","funder_award_id":"2142794","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6671297155","display_name":null,"funder_award_id":"CAREER","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G8529959211","display_name":"III: Large: Collaborative Research: Analysis Engineering for Robust End-to-End Data Science","funder_award_id":"1901386","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7102392873.pdf","grobid_xml":"https://content.openalex.org/works/W7102392873.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Evaluating":[0],"the":[1,40,132],"reasoning":[2,50,108,183],"ability":[3],"of":[4,88,182],"language":[5],"models":[6,72,135],"(LMs)":[7],"is":[8,84],"complicated":[9],"by":[10],"their":[11],"extensive":[12],"parametric":[13,75,139],"world":[14,140],"knowledge,":[15,76],"where":[16,71,81],"benchmark":[17],"performance":[18,133],"often":[19],"reflects":[20],"factual":[21,53],"recall":[22],"rather":[23],"than":[24],"genuine":[25],"reasoning.":[26],"Existing":[27],"datasets":[28],"and":[29,77,102,118,144,161,179,184],"approaches":[30],"(e.g.,":[31,115,120],"temporal":[32],"filtering,":[33],"paraphrasing,":[34],"adversarial":[35],"substitution)":[36],"cannot":[37],"cleanly":[38],"separate":[39],"two.":[41],"We":[42],"present":[43],"SynthWorlds,":[44,56],"a":[45,68,78,125,165],"framework":[46],"that":[47,173],"disentangles":[48],"task":[49],"complexity":[51],"from":[52,137],"knowledge.":[54,141],"In":[55],"we":[57,91],"construct":[58],"parallel":[59],"corpora":[60],"representing":[61],"two":[62,93],"worlds":[63],"with":[64],"identical":[65],"interconnected":[66],"structure:":[67],"real-mapped":[69],"world,":[70,80],"may":[73],"exploit":[74],"synthetic-mapped":[79],"such":[82],"knowledge":[83,127],"meaningless.":[85],"On":[86],"top":[87],"these":[89],"corpora,":[90],"design":[92],"mirrored":[94],"tasks":[95],"as":[96,131],"case":[97],"studies:":[98],"multi-hop":[99],"question":[100],"answering":[101],"page":[103],"navigation,":[104],"which":[105],"maintain":[106],"equal":[107],"difficulty":[109],"across":[110],"worlds.":[111],"Experiments":[112],"in":[113,171],"parametric-only":[114],"closed-book":[116],"QA)":[117],"knowledge-augmented":[119],"retrieval-augmented)":[121],"LM":[122],"settings":[123],"reveal":[124],"persistent":[126],"advantage":[128],"gap,":[129,153],"defined":[130],"boost":[134],"gain":[136],"memorized":[138],"Knowledge":[142],"acquisition":[143],"integration":[145],"mechanisms":[146],"reduce":[147],"but":[148],"do":[149],"not":[150],"eliminate":[151],"this":[152],"highlighting":[154],"opportunities":[155],"for":[156,168],"system":[157],"improvements.":[158],"Fully":[159],"automatic":[160],"scalable,":[162],"SynthWorlds":[163],"provides":[164],"controlled":[166],"environment":[167],"evaluating":[169],"LMs":[170],"ways":[172],"were":[174],"previously":[175],"challenging,":[176],"enabling":[177],"precise":[178],"testable":[180],"comparisons":[181],"memorization.":[185]},"counts_by_year":[],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2025-10-30T00:00:00"}