{"id":"https://openalex.org/W7203705824","doi":"https://doi.org/10.48550/arxiv.2608.14906","title":"Optimal Watermark Localization in Mixed-Source Large Language Model Texts","display_name":"Optimal Watermark Localization in Mixed-Source Large Language Model Texts","publication_year":2026,"publication_date":"2026-08-14","ids":{"openalex":"https://openalex.org/W7203705824","doi":"https://doi.org/10.48550/arxiv.2608.14906"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.14906","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14906","pdf_url":"https://arxiv.org/pdf/2608.14906","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.14906","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011147039","display_name":"Jos\u00e9 Blanchet","orcid":"https://orcid.org/0000-0001-5895-0912","observed_orcids":["https://orcid.org/0000-0001-5895-0912"]},"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":"Blanchet, Jose H.","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/A5122786770","display_name":"T. Tony Cai","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Cai, T. Tony","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147414908","display_name":"Xiang Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li, Xiang","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147443536","display_name":"Hao Liu","orcid":"https://orcid.org/0000-0002-1718-1448","observed_orcids":["https://orcid.org/0000-0002-1718-1448"]},"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":"Liu, Hao","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/A5147403058","display_name":"Qi Long","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Long, Qi","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101342734","display_name":"Weijie J. Su","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Su, Weijie J.","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]}],"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":3,"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/T10388","display_name":"Advanced Steganography and Watermarking Techniques","score":0.41499999165534973,"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"}},"topics":[{"id":"https://openalex.org/T10388","display_name":"Advanced Steganography and Watermarking Techniques","score":0.41499999165534973,"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"}},{"id":"https://openalex.org/T12380","display_name":"Authorship Attribution and Profiling","score":0.08739999681711197,"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/T10136","display_name":"Statistical Methods and Inference","score":0.050700001418590546,"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/watermarking","display_name":"watermarking","score":0.597000002861023},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9490000009536743},{"id":"https://openalex.org/keywords/phase-transitions","display_name":"phase transitions","score":0.9010000228881836},{"id":"https://openalex.org/keywords/adaptive-thresholding","display_name":"adaptive thresholding","score":0.8399999737739563},{"id":"https://openalex.org/keywords/signal-sparsity","display_name":"signal sparsity","score":0.6439999938011169},{"id":"https://openalex.org/keywords/paraphrasing","display_name":"paraphrase","score":0.38199999928474426},{"id":"https://openalex.org/keywords/text-editing","display_name":"text editing","score":0.4519999921321869},{"id":"https://openalex.org/keywords/watermark-detection","display_name":"watermark detection","score":0.31700000166893005}],"concepts":[{"id":"https://openalex.org/C164112704","wikidata":"https://www.wikidata.org/wiki/Q7974348","display_name":"Watermark","level":3,"score":0.7843000292778015},{"id":"https://openalex.org/C191178318","wikidata":"https://www.wikidata.org/wiki/Q2256906","display_name":"Thresholding","level":3,"score":0.6606000065803528},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6226999759674072},{"id":"https://openalex.org/C150817343","wikidata":"https://www.wikidata.org/wiki/Q875932","display_name":"Digital watermarking","level":3,"score":0.5983999967575073},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5720999836921692},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.5699999928474426},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5092999935150146},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.47099998593330383},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.43630000948905945},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.430400013923645},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.41530001163482666},{"id":"https://openalex.org/C66882249","wikidata":"https://www.wikidata.org/wiki/Q169336","display_name":"Homogeneous","level":2,"score":0.3901999890804291},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.38670000433921814},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3578999936580658},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.34279999136924744},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.313400000333786},{"id":"https://openalex.org/C2221639","wikidata":"https://www.wikidata.org/wiki/Q2877","display_name":"Discrete cosine transform","level":3,"score":0.3093000054359436},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.3043999969959259},{"id":"https://openalex.org/C44280652","wikidata":"https://www.wikidata.org/wiki/Q104837","display_name":"Phase (matter)","level":2,"score":0.29910001158714294},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2824000120162964},{"id":"https://openalex.org/C137270730","wikidata":"https://www.wikidata.org/wiki/Q120811","display_name":"Detection theory","level":3,"score":0.28130000829696655},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.27730000019073486},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.25130000710487366},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.25130000710487366}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.14906","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14906","pdf_url":"https://arxiv.org/pdf/2608.14906","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.2608.14906","doi":"https://doi.org/10.48550/arxiv.2608.14906","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.14906","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:2608.14906","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14906","pdf_url":"https://arxiv.org/pdf/2608.14906","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":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G4650965343","display_name":"Collaborative Research: CIF: Medium: Statistical and Algorithmic Foundations of Distributionally Robust Policy Learning","funder_award_id":"2312204","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4931875759","display_name":"Collaborative Research: FET: Medium: Quantum Monte Carlo Speed Ups for Multilevel Computations and Other Statistical Algorithms","funder_award_id":"2403007","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G895949175","display_name":"Geometrization Approaches toward Understanding Deep Learning","funder_award_id":"2310679","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"},{"id":"https://openalex.org/F4320306078","display_name":"U.S. Department of Defense","ror":"https://ror.org/0447fe631"},{"id":"https://openalex.org/F4320332161","display_name":"National Institutes of Health","ror":"https://ror.org/01cwqze88"},{"id":"https://openalex.org/F4320337345","display_name":"Office of Naval Research","ror":"https://ror.org/00rk2pe57"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7203705824.pdf","grobid_xml":"https://content.openalex.org/works/W7203705824.grobid-xml"},"referenced_works_count":10,"referenced_works":["https://openalex.org/W1607372231","https://openalex.org/W4229005866","https://openalex.org/W4322718191","https://openalex.org/W4403680773","https://openalex.org/W4403884264","https://openalex.org/W4412945560","https://openalex.org/W4414939355","https://openalex.org/W4416261448","https://openalex.org/W7119523234","https://openalex.org/W7130689510"],"related_works":["https://openalex.org/W4414790673","https://openalex.org/W4380033094","https://openalex.org/W4411112989","https://openalex.org/W7161386412","https://openalex.org/W7105242699","https://openalex.org/W7166594550","https://openalex.org/W4414453530","https://openalex.org/W4407557405","https://openalex.org/W4412945560","https://openalex.org/W7117121472"],"abstract_inverted_index":{"Watermarking":[0],"provides":[1],"a":[2,29,63,72,100,164],"principled":[3],"way":[4],"to":[5,184],"authenticate":[6],"text":[7,19],"generated":[8],"by":[9,88],"large":[10],"language":[11],"models":[12],"(LLMs).":[13],"In":[14],"practice,":[15],"however,":[16],"the":[17,114,137,156,168,175,190],"final":[18],"may":[20],"be":[21,54],"mixed-source,":[22],"with":[23,71],"watermark":[24,48,60,77,170],"evidence":[25],"surviving":[26,169],"at":[27,80],"only":[28],"subset":[30],"of":[31,47,116,155,167],"token":[32],"positions":[33],"after":[34],"rewriting,":[35],"insertion,":[36],"deletion,":[37],"or":[38,158],"paraphrasing.":[39],"Although":[40],"prior":[41],"work":[42],"has":[43],"studied":[44],"global":[45,104],"detection":[46,105,129],"signals,":[49],"when":[50],"such":[51],"signals":[52],"can":[53],"localized":[55],"remains":[56],"unclear.":[57],"We":[58,121,143],"formulate":[59],"localization":[61,119,201],"as":[62],"token-level":[64],"multiple-testing":[65],"problem":[66],"based":[67],"on":[68,196],"pivotal":[69],"statistics,":[70],"latent":[73],"indicator":[74],"recording":[75],"whether":[76],"dependence":[78],"survives":[79],"each":[81],"position.":[82],"Under":[83],"an":[84,146],"asymptotic":[85],"regime":[86,139],"indexed":[87],"exponents":[89,157],"for":[90,103,109],"signal":[91],"sparsity,":[92],"next-token":[93,160],"concentration,":[94],"and":[95,106,111,130,179],"effective-vocabulary":[96],"growth,":[97],"we":[98],"derive":[99],"sharp":[101],"boundary":[102,178],"phase":[107,192],"transitions":[108],"discovery":[110,124,177,181],"classification":[112,133],"within":[113,140],"class":[115],"coordinatewise":[117],"pivot-based":[118,186],"rules.":[120,187],"show":[122],"that":[123,131,150],"is":[125,134],"strictly":[126],"harder":[127],"than":[128],"consistent":[132],"impossible":[135],"across":[136],"parameter":[138],"this":[141],"class.":[142],"then":[144],"develop":[145],"adaptive":[147],"thresholding":[148],"method":[149,173],"does":[151],"not":[152],"require":[153],"knowledge":[154],"time-varying":[159],"distributions,":[161],"but":[162],"uses":[163],"data-driven":[165],"estimate":[166],"fraction.":[171],"The":[172],"attains":[174],"optimal":[176],"near-optimal":[180],"power":[182],"relative":[183],"homogeneous":[185],"Simulations":[188],"support":[189],"theoretical":[191],"transitions,":[193],"while":[194],"experiments":[195],"model-generated":[197],"texts":[198],"demonstrate":[199],"practical":[200],"performance":[202],"under":[203],"common":[204],"edit":[205],"mechanisms.":[206]},"counts_by_year":[{"year":2026,"cited_by_count":3}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-19T00:00:00"}