{"id":"https://openalex.org/W7164398789","doi":"https://doi.org/10.48550/arxiv.2606.11499","title":"Hubs or Fringes: Pretraining Data Selection via Web Graph Centrality","display_name":"Hubs or Fringes: Pretraining Data Selection via Web Graph Centrality","publication_year":2026,"publication_date":"2026-06-09","ids":{"openalex":"https://openalex.org/W7164398789","doi":"https://doi.org/10.48550/arxiv.2606.11499"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.11499","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.11499","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.11499","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138442236","display_name":"Vedant Badoni","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Badoni, Vedant","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138420598","display_name":"Danqi Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Danqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138433134","display_name":"Xinyi Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xinyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.42730000615119934,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.42730000615119934,"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/T12016","display_name":"Web Data Mining and Analysis","score":0.11670000106096268,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10028","display_name":"Topic Modeling","score":0.09350000321865082,"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/encode","display_name":"ENCODE","score":0.7419000267982483},{"id":"https://openalex.org/keywords/centrality","display_name":"Centrality","score":0.7242000102996826},{"id":"https://openalex.org/keywords/classifier","display_name":"Classifier (UML)","score":0.6082000136375427},{"id":"https://openalex.org/keywords/graph","display_name":"Graph","score":0.5157999992370605},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.4269999861717224},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.3366999924182892},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.3174000084400177},{"id":"https://openalex.org/keywords/model-selection","display_name":"Model selection","score":0.31279999017715454}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.748199999332428},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.7419000267982483},{"id":"https://openalex.org/C53811970","wikidata":"https://www.wikidata.org/wiki/Q5062194","display_name":"Centrality","level":2,"score":0.7242000102996826},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.6082000136375427},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.5157999992370605},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.4846000075340271},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.453900009393692},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4269999861717224},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40540000796318054},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.3366999924182892},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.3174000084400177},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.31279999017715454},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.2992999851703644},{"id":"https://openalex.org/C2987255567","wikidata":"https://www.wikidata.org/wiki/Q33002955","display_name":"Knowledge graph","level":2,"score":0.2833000123500824},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.27799999713897705},{"id":"https://openalex.org/C2776145971","wikidata":"https://www.wikidata.org/wiki/Q30673951","display_name":"Labeled data","level":2,"score":0.27309998869895935},{"id":"https://openalex.org/C2993807640","wikidata":"https://www.wikidata.org/wiki/Q103709453","display_name":"Attention network","level":2,"score":0.27059999108314514},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2615000009536743},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.2605000138282776},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25839999318122864},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2551000118255615},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2540999948978424},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.25189998745918274},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.11499","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.11499","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.11499","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.11499","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,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0],"performance":[1,178],"of":[2,62,156],"modern":[3],"language":[4],"models":[5,77,114],"depends":[6],"critically":[7],"on":[8,18,32,128,160],"pretraining":[9,69,193],"data":[10,14,40,194],"composition.":[11],"Yet":[12],"existing":[13,203],"selection":[15,41],"methods":[16],"rely":[17],"auxiliary":[19],"classifiers":[20],"for":[21,165,192],"document":[22],"scoring":[23],"or":[24,102],"mixture":[25],"optimization,":[26],"adding":[27],"computational":[28],"overhead":[29],"and":[30,55,112,117,123,143],"dependence":[31],"labeled":[33,100],"data.":[34],"We":[35,71,105],"propose":[36],"WebGraphMix,":[37],"a":[38,154,189],"lightweight":[39],"framework":[42],"that":[43,73,141,184,198],"computes":[44,89],"structural":[45,169],"centrality":[46,90],"scores":[47,91,170,175],"over":[48],"the":[49,60,68,109],"Common":[50],"Crawl":[51],"host-level":[52],"web":[53,94,145,185],"graph":[54,186],"uses":[56],"them":[57],"to":[58,78,135,163,179,202],"vary":[59],"proportion":[61],"central":[63,74,142],"versus":[64],"peripheral":[65,82,144],"documents":[66],"in":[67],"mixture.":[70],"hypothesize":[72],"hosts":[75,83],"expose":[76],"reusable":[79],"abstractions,":[80],"while":[81],"encode":[84,147],"specialized,":[85],"long-tail":[86],"knowledge.":[87],"WebGraphMix":[88,107],"efficiently":[92],"at":[93,115,153],"scale,":[95],"requiring":[96],"no":[97],"model":[98],"training,":[99],"data,":[101],"downstream":[103],"supervision.":[104],"integrate":[106],"into":[108],"DataComp-LM":[110],"pipeline":[111],"train":[113],"400M":[116],"1B":[118],"parameter":[119],"scales":[120],"with":[121,171],"8B":[122],"28B":[124],"tokens":[125],"respectively,":[126],"evaluating":[127],"23":[129],"tasks":[130],"ranging":[131],"from":[132],"factual":[133],"knowledge":[134],"symbolic":[136],"reasoning.":[137],"Our":[138],"experiments":[139],"show":[140],"regions":[146],"complementary":[148],"capabilities.":[149],"Mixture":[150],"combining":[151],"both":[152],"ratio":[155],"1:1":[157],"achieves":[158],"41.4%":[159],"average,":[161],"compared":[162],"39.8%":[164],"uniform":[166],"sampling.":[167],"Combining":[168],"document-level":[172],"quality":[173],"classifier":[174],"further":[176],"improves":[177],"43.8%.":[180],"These":[181],"findings":[182],"demonstrate":[183],"topology":[187],"is":[188,199],"meaningful":[190],"axis":[191],"curation,":[195],"capturing":[196],"information":[197],"largely":[200],"orthogonal":[201],"content-based":[204],"approaches.":[205]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-12T00:00:00"}