{"id":"https://openalex.org/W7140319612","doi":"https://doi.org/10.48550/arxiv.2603.22299","title":"Between the Layers Lies the Truth: Uncertainty Estimation in LLMs Using Intra-Layer Local Information Scores","display_name":"Between the Layers Lies the Truth: Uncertainty Estimation in LLMs Using Intra-Layer Local Information Scores","publication_year":2026,"publication_date":"2026-03-17","ids":{"openalex":"https://openalex.org/W7140319612","doi":"https://doi.org/10.48550/arxiv.2603.22299"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.22299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.22299","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.22299","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5115983654","display_name":"Zvi Badash","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Badash, Zvi N.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130619117","display_name":"Yonatan Belinkov","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Belinkov, Yonatan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5130574330","display_name":"Moti Freiman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Freiman, Moti","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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3375000059604645,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3375000059604645,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.13600000739097595,"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/T10028","display_name":"Topic Modeling","score":0.11479999870061874,"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/brier-score","display_name":"Brier score","score":0.8481000065803528},{"id":"https://openalex.org/keywords/heuristics","display_name":"Heuristics","score":0.6524999737739563},{"id":"https://openalex.org/keywords/estimation","display_name":"Estimation","score":0.4424000084400177},{"id":"https://openalex.org/keywords/point-estimation","display_name":"Point estimation","score":0.2937000095844269},{"id":"https://openalex.org/keywords/encode","display_name":"ENCODE","score":0.2921000123023987}],"concepts":[{"id":"https://openalex.org/C35405484","wikidata":"https://www.wikidata.org/wiki/Q4967066","display_name":"Brier score","level":2,"score":0.8481000065803528},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.6524999737739563},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5450000166893005},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.46470001339912415},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.4426000118255615},{"id":"https://openalex.org/C96250715","wikidata":"https://www.wikidata.org/wiki/Q965330","display_name":"Estimation","level":2,"score":0.4424000084400177},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.39739999175071716},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.38600000739097595},{"id":"https://openalex.org/C41426520","wikidata":"https://www.wikidata.org/wiki/Q1192065","display_name":"Point estimation","level":2,"score":0.2937000095844269},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.2921000123023987},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.28929999470710754},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.28760001063346863},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.2777999937534332},{"id":"https://openalex.org/C27438332","wikidata":"https://www.wikidata.org/wiki/Q2873","display_name":"Principal component analysis","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.2533000111579895}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.22299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.22299","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.22299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.22299","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":{"Large":[0],"language":[1],"models":[2,125],"(LLMs)":[3],"are":[4,16],"often":[5],"confidently":[6],"wrong,":[7],"making":[8],"reliable":[9],"uncertainty":[10,140],"estimation":[11],"(UE)":[12],"essential.":[13],"Output-based":[14],"heuristics":[15],"cheap":[17],"but":[18],"brittle,":[19],"while":[20],"probing":[21,58,103],"internal":[22,45],"representations":[23,46],"is":[24],"effective":[25],"yet":[26],"high-dimensional":[27],"and":[28,71,90,108],"hard":[29],"to":[30,87,137],"transfer.":[31],"We":[32],"propose":[33],"a":[34,48,133],"compact,":[35],"per-instance":[36],"UE":[37,130],"method":[38,56,131],"that":[39],"scores":[40],"cross-layer":[41],"agreement":[42],"patterns":[43],"in":[44,122,141],"using":[47],"single":[49],"forward":[50],"pass.":[51],"Across":[52],"three":[53],"models,":[54],"our":[55,129],"matches":[57],"in-distribution,":[59],"with":[60],"mean":[61],"diagonal":[62],"differences":[63,121],"of":[64],"at":[65],"most":[66],"$-1.8$":[67],"AUPRC":[68,89,106],"percentage":[69],"points":[70,107,111],"$+4.9$":[72],"Brier":[73,92,110],"score":[74],"points.":[75,93],"Under":[76,94],"cross-dataset":[77],"transfer,":[78],"it":[79,98],"consistently":[80],"outperforms":[81],"probing,":[82],"achieving":[83],"off-diagonal":[84],"gains":[85],"up":[86],"$+2.86$":[88],"$+21.02$":[91],"4-bit":[95],"weight-only":[96],"quantization,":[97],"remains":[99],"robust,":[100],"improving":[101],"over":[102],"by":[104],"$+1.94$":[105],"$+5.33$":[109],"on":[112],"average.":[113],"Beyond":[114],"performance,":[115],"examining":[116],"specific":[117],"layer--layer":[118],"interactions":[119],"reveals":[120],"how":[123],"disparate":[124],"encode":[126],"uncertainty.":[127],"Altogether,":[128],"offers":[132],"lightweight,":[134],"compact":[135],"means":[136],"capture":[138],"transferable":[139],"LLMs.":[142]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-26T00:00:00"}