{"id":"https://openalex.org/W7166880640","doi":"https://doi.org/10.18653/v1/2026.acl-long.1862","title":"Learning Uncertainty from Sequential Internal Dispersion in Large Language Models","display_name":"Learning Uncertainty from Sequential Internal Dispersion in Large Language Models","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166880640","doi":"https://doi.org/10.18653/v1/2026.acl-long.1862"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.1862","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1862","pdf_url":"https://aclanthology.org/2026.acl-long.1862.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.1862.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5106011068","display_name":"Ponhvoan Srey","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ponhvoan Srey","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139715013","display_name":"Xiaobao Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaobao Wu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5025645183","display_name":"Cong-Duy Nguyen","orcid":"https://orcid.org/0000-0002-0931-460X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cong-Duy T Nguyen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139773614","display_name":"Anh Tuan Luu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Anh Tuan Luu","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.86894882,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"40088","last_page":"40106"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.37560001015663147,"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":0.37560001015663147,"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":0.04179999977350235,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.04039999842643738,"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/dispersion","display_name":"Dispersion (optics)","score":0.3596000075340271},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.3149999976158142},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.28999999165534973},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.27489998936653137},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2644999921321869}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5583000183105469},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4140999913215637},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.398499995470047},{"id":"https://openalex.org/C177562468","wikidata":"https://www.wikidata.org/wiki/Q182893","display_name":"Dispersion (optics)","level":2,"score":0.3596000075340271},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3149999976158142},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.3149999976158142},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.28999999165534973},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.27489998936653137},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2644999921321869},{"id":"https://openalex.org/C2778067643","wikidata":"https://www.wikidata.org/wiki/Q166507","display_name":"Interval (graph theory)","level":2,"score":0.25940001010894775},{"id":"https://openalex.org/C159886148","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov process","level":2,"score":0.25609999895095825}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.1862","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1862","pdf_url":"https://aclanthology.org/2026.acl-long.1862.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.1862","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1862","pdf_url":"https://aclanthology.org/2026.acl-long.1862.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5220339298248291}],"awards":[],"funders":[{"id":"https://openalex.org/F4320320671","display_name":"National Research Foundation","ror":"https://ror.org/05s0g1g46"},{"id":"https://openalex.org/F4320320709","display_name":"National Research Foundation Singapore","ror":"https://ror.org/03cpyc314"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166880640.pdf","grobid_xml":"https://content.openalex.org/works/W7166880640.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Uncertainty":[0],"estimation":[1],"is":[2],"a":[3,59,74],"promising":[4],"approach":[5],"to":[6,21],"detect":[7],"hallucinations":[8],"in":[9,81],"large":[10,144],"language":[11],"models":[12],"(LLMs).Recent":[13],"approaches":[14],"commonly":[15],"depend":[16],"on":[17,29,44,96,143],"model":[18,103],"internal":[19,89],"states":[20,32],"estimate":[22],"uncertainty.However,":[23],"they":[24],"suffer":[25],"from":[26,38,70],"strict":[27],"assumptions":[28],"how":[30],"hidden":[31,71],"should":[33],"evolve":[34],"across":[35,91],"layers,":[36,92],"and":[37,104,123,140],"information":[39,126],"loss":[40],"by":[41],"solely":[42],"focusing":[43],"last":[45],"or":[46,86],"mean":[47],"tokens.To":[48],"address":[49],"these":[50],"issues,":[51],"we":[52],"present":[53],"Sequential":[54],"Internal":[55],"Variance":[56],"Representation":[57],"(SIVR),":[58],"supervised":[60],"hallucination":[61],"detection":[62],"framework":[63],"that":[64,78],"leverages":[65],"token-wise,":[66],"layer-wise":[67],"features":[68],"derived":[69],"states.SIVR":[72],"adopts":[73],"more":[75],"basic":[76],"assumption":[77],"uncertainty":[79],"manifests":[80],"the":[82,101,109,148],"degree":[83],"of":[84,88,112,120],"dispersion":[85],"variance":[87,114],"representations":[90],"rather":[93],"than":[94],"relying":[95,142],"specific":[97],"assumptions,":[98],"which":[99],"makes":[100],"method":[102],"task":[105],"agnostic.It":[106],"additionally":[107],"aggregates":[108],"full":[110],"sequence":[111],"per-token":[113],"features,":[115],"learning":[116],"temporal":[117],"patterns":[118],"indicative":[119],"factual":[121],"errors":[122],"thereby":[124],"preventing":[125],"loss.Experimental":[127],"results":[128],"demonstrate":[129],"SIVR":[130,136],"consistently":[131],"outperforms":[132],"strong":[133],"baselines.Most":[134],"importantly,":[135],"enjoys":[137],"stronger":[138],"generalisation":[139],"avoids":[141],"training":[145],"sets,":[146],"highlighting":[147],"potential":[149],"for":[150],"practical":[151],"deployment.":[152]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}