{"id":"https://openalex.org/W4416445181","doi":"https://doi.org/10.48550/arxiv.2503.01917","title":"Steer LLM Latents for Hallucination Detection","display_name":"Steer LLM Latents for Hallucination Detection","publication_year":2025,"publication_date":"2025-03-01","ids":{"openalex":"https://openalex.org/W4416445181","doi":"https://doi.org/10.48550/arxiv.2503.01917"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2503.01917","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.01917","pdf_url":"https://arxiv.org/pdf/2503.01917","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":null,"license_id":null,"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/2503.01917","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5041702782","display_name":"Seongheon Park","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Park, Seongheon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001001983","display_name":"Xuefeng Du","orcid":"https://orcid.org/0000-0002-8080-7861","observed_orcids":["https://orcid.org/0000-0002-8080-7861"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Du, Xuefeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029426380","display_name":"Min-Hsuan Yeh","orcid":"https://orcid.org/0009-0001-4945-712X","observed_orcids":["https://orcid.org/0009-0001-4945-712X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yeh, Min-Hsuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049707744","display_name":"Haobo Wang","orcid":"https://orcid.org/0000-0001-8586-3048","observed_orcids":["https://orcid.org/0000-0001-8586-3048"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Haobo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100443469","display_name":"Yixuan Li","orcid":"https://orcid.org/0000-0002-0685-2875","observed_orcids":["https://orcid.org/0000-0002-0685-2875","https://orcid.org/0000-0001-8283-5885","https://orcid.org/0009-0000-3770-7622"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yixuan","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":true,"cited_by_count":32,"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/T10028","display_name":"Topic Modeling","score":0.5224999785423279,"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":0.5224999785423279,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.2003999948501587,"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/T11147","display_name":"Misinformation and Its Impacts","score":0.179299995303154,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/hallucination-detection","display_name":"hallucination detection","score":0.9739999771118164},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8920000195503235},{"id":"https://openalex.org/keywords/pseudo-labeling","display_name":"pseudo-labeling","score":0.7329999804496765},{"id":"https://openalex.org/keywords/optimal-transport","display_name":"optimal transport","score":0.9789999723434448},{"id":"https://openalex.org/keywords/confidence-based-filtering","display_name":"confidence-based filtering","score":0.699999988079071},{"id":"https://openalex.org/keywords/factual-accuracy","display_name":"factual accuracy","score":0.6010000109672546}],"concepts":[{"id":"https://openalex.org/C2911011789","wikidata":"https://www.wikidata.org/wiki/Q130741","display_name":"Hallucinating","level":2,"score":0.9405999779701233},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6219000220298767},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5710999965667725},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5236999988555908},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5102999806404114},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.459199994802475},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.4564000070095062},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.4065999984741211},{"id":"https://openalex.org/C17137986","wikidata":"https://www.wikidata.org/wiki/Q215067","display_name":"Orthogonality","level":2,"score":0.3871000111103058},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3799999952316284},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.34380000829696655},{"id":"https://openalex.org/C2781181686","wikidata":"https://www.wikidata.org/wiki/Q4226068","display_name":"Coherence (philosophical gambling strategy)","level":2,"score":0.33889999985694885},{"id":"https://openalex.org/C122770356","wikidata":"https://www.wikidata.org/wiki/Q1656753","display_name":"Identifiability","level":2,"score":0.3287999927997589},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3091999888420105},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.30550000071525574},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.2946999967098236},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.2865000069141388},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.2718999981880188},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2630999982357025},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2556999921798706}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2503.01917","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.01917","pdf_url":"https://arxiv.org/pdf/2503.01917","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2503.01917","doi":"https://doi.org/10.48550/arxiv.2503.01917","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2503.01917","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:2503.01917","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.01917","pdf_url":"https://arxiv.org/pdf/2503.01917","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":null,"license_id":null,"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/G6108913550","display_name":null,"funder_award_id":"FA9550-23-1-0184","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G8760265284","display_name":"CAREER: Foundations of Human-Centered Machine Learning in the Wild","funder_award_id":"2237037","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/F4320338279","display_name":"Air Force Office of Scientific Research","ror":"https://ror.org/011e9bt93"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416445181.pdf","grobid_xml":"https://content.openalex.org/works/W4416445181.grobid-xml"},"referenced_works_count":13,"referenced_works":["https://openalex.org/W2738674657","https://openalex.org/W4285429195","https://openalex.org/W4362515116","https://openalex.org/W4366850747","https://openalex.org/W4384918448","https://openalex.org/W4400023100","https://openalex.org/W4401306886","https://openalex.org/W4404330874","https://openalex.org/W4405655184","https://openalex.org/W4407208939","https://openalex.org/W4407758946","https://openalex.org/W4412158322","https://openalex.org/W4416040343"],"related_works":["https://openalex.org/W4414888143","https://openalex.org/W4412404236","https://openalex.org/W4415796039","https://openalex.org/W4402683996","https://openalex.org/W7151395083","https://openalex.org/W4415058864","https://openalex.org/W4415088582","https://openalex.org/W4415134368","https://openalex.org/W6948148045","https://openalex.org/W4394007010"],"abstract_inverted_index":{"Hallucinations":[0],"in":[1,11],"LLMs":[2,22],"pose":[3],"a":[4,56,90,121,144],"significant":[5],"concern":[6],"to":[7,39,70,96],"their":[8,27],"safe":[9],"deployment":[10],"real-world":[12,148],"applications.":[13,150],"Recent":[14],"approaches":[15],"have":[16],"leveraged":[17],"the":[18,51,64,72,105],"latent":[19],"space":[20,67],"of":[21,93],"for":[23,30,117,147],"hallucination":[24],"detection,":[25],"but":[26],"embeddings,":[28],"optimized":[29],"linguistic":[31],"coherence":[32],"rather":[33],"than":[34],"factual":[35],"accuracy,":[36],"often":[37],"fail":[38],"clearly":[40],"separate":[41],"truthful":[42,75],"and":[43,58,76,99,142],"hallucinated":[44,77],"content.":[45],"To":[46],"this":[47],"end,":[48],"we":[49],"propose":[50],"Truthfulness":[52],"Separator":[53],"Vector":[54],"(TSV),":[55],"lightweight":[57],"flexible":[59],"steering":[60],"vector":[61],"that":[62,128],"reshapes":[63],"LLM's":[65],"representation":[66],"during":[68],"inference":[69],"enhance":[71],"separation":[73],"between":[74],"outputs,":[78],"without":[79],"altering":[80],"model":[81],"parameters.":[82],"Our":[83],"two-stage":[84],"framework":[85],"first":[86],"trains":[87],"TSV":[88,129],"on":[89],"small":[91],"set":[92,107],"labeled":[94,135],"exemplars":[95],"form":[97],"compact":[98],"well-separated":[100],"clusters.":[101],"It":[102],"then":[103],"augments":[104],"exemplar":[106],"with":[108,120,133],"unlabeled":[109],"LLM":[110,149],"generations,":[111],"employing":[112],"an":[113],"optimal":[114],"transport-based":[115],"algorithm":[116],"pseudo-labeling":[118],"combined":[119],"confidence-based":[122],"filtering":[123],"process.":[124],"Extensive":[125],"experiments":[126],"demonstrate":[127],"achieves":[130],"state-of-the-art":[131],"performance":[132],"minimal":[134],"data,":[136],"exhibiting":[137],"strong":[138],"generalization":[139],"across":[140],"datasets":[141],"providing":[143],"practical":[145],"solution":[146]},"counts_by_year":[{"year":2026,"cited_by_count":23},{"year":2025,"cited_by_count":9}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}