{"id":"https://openalex.org/W7053377175","doi":null,"title":"Truthfulness in Large Language Models","display_name":"Truthfulness in Large Language Models","publication_year":2023,"publication_date":"2023-06-01","ids":{"openalex":"https://openalex.org/W7053377175"},"language":"en","primary_location":{"id":"pmh:oai:dspace.mit.edu:1721.1/151345","doi":null,"is_oa":true,"landing_page_url":"https://hdl.handle.net/1721.1/151345","pdf_url":"https://dspace.mit.edu/bitstream/1721.1/151345/1/liu-kevliu-meng-eecs-2023-thesis.pdf","source":{"id":"https://openalex.org/S4306400425","display_name":"DSpace@MIT (Massachusetts Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I63966007","host_organization_name":"Massachusetts Institute of Technology","host_organization_lineage":["https://openalex.org/I63966007"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Thesis"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://dspace.mit.edu/bitstream/1721.1/151345/1/liu-kevliu-meng-eecs-2023-thesis.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100658205","display_name":"Kevin Liu","orcid":"https://orcid.org/0000-0001-7688-5819","observed_orcids":["https://orcid.org/0000-0001-7688-5819"]},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Liu, Kevin","raw_affiliation_strings":["Mathematics and Computer Science and Engineering , Massachusetts Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mathematics and Computer Science and Engineering , Massachusetts Institute of Technology","institution_ids":["https://openalex.org/I63966007"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5100658205"],"corresponding_institution_ids":["https://openalex.org/I63966007"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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":true,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6743000149726868,"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":0.6743000149726868,"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":0.10760000348091125,"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":0.08510000258684158,"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/honesty","display_name":"honesty","score":0.7990000247955322},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9829999804496765},{"id":"https://openalex.org/keywords/linear-probing","display_name":"linear probing","score":0.7049999833106995},{"id":"https://openalex.org/keywords/internal-representations","display_name":"internal representations","score":0.6880000233650208},{"id":"https://openalex.org/keywords/confidence-calibration","display_name":"confidence calibration","score":0.6980000138282776},{"id":"https://openalex.org/keywords/model-perturbation","display_name":"model perturbation","score":0.6990000009536743},{"id":"https://openalex.org/keywords/factual-accuracy","display_name":"factual accuracy","score":0.5099999904632568}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.642799973487854},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.6413999795913696},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5175999999046326},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5005999803543091},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4747999906539917},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.3785000145435333},{"id":"https://openalex.org/C163175372","wikidata":"https://www.wikidata.org/wiki/Q3339222","display_name":"Linear model","level":2,"score":0.3472999930381775},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.3440999984741211},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.33559998869895935},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2797999978065491},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2612000107765198},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2524000108242035}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:dspace.mit.edu:1721.1/151345","doi":null,"is_oa":true,"landing_page_url":"https://hdl.handle.net/1721.1/151345","pdf_url":"https://dspace.mit.edu/bitstream/1721.1/151345/1/liu-kevliu-meng-eecs-2023-thesis.pdf","source":{"id":"https://openalex.org/S4306400425","display_name":"DSpace@MIT (Massachusetts Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I63966007","host_organization_name":"Massachusetts Institute of Technology","host_organization_lineage":["https://openalex.org/I63966007"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Thesis"}],"best_oa_location":{"id":"pmh:oai:dspace.mit.edu:1721.1/151345","doi":null,"is_oa":true,"landing_page_url":"https://hdl.handle.net/1721.1/151345","pdf_url":"https://dspace.mit.edu/bitstream/1721.1/151345/1/liu-kevliu-meng-eecs-2023-thesis.pdf","source":{"id":"https://openalex.org/S4306400425","display_name":"DSpace@MIT (Massachusetts Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I63966007","host_organization_name":"Massachusetts Institute of Technology","host_organization_lineage":["https://openalex.org/I63966007"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Thesis"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7053377175.pdf","grobid_xml":"https://content.openalex.org/works/W7053377175.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4393146683","https://openalex.org/W4402670137","https://openalex.org/W4387561538","https://openalex.org/W7221401661","https://openalex.org/W4402683996","https://openalex.org/W4412888889","https://openalex.org/W4411875187","https://openalex.org/W4391212039","https://openalex.org/W4388032052","https://openalex.org/W4409671508"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4],"been":[5],"experiencing":[6],"a":[7,50,63],"rapid":[8],"rise":[9],"in":[10,21,70],"utility,":[11],"accessibility,":[12],"and":[13,87],"popularity,":[14],"but":[15],"there":[16],"are":[17],"still":[18],"many":[19],"areas":[20],"which":[22],"they":[23],"can":[24],"improve.":[25],"One":[26],"such":[27],"area":[28],"for":[29],"improvement":[30],"is":[31,59,76],"their":[32,44],"truthfulness.":[33],"We":[34,47],"seek":[35],"to":[36,61,83],"improve":[37,62],"the":[38,54,81,89],"truthfulness":[39],"of":[40],"LLMs":[41],"by":[42,66],"probing":[43],"internal":[45],"representations.":[46],"find":[48],"that":[49],"linear":[51],"probe":[52,75],"on":[53],"last":[55],"hidden":[56],"layer":[57],"representation":[58],"able":[60],"model\u2019s":[64],"accuracy":[65],"reducing":[67],"its":[68,85],"confidence":[69],"incorrect":[71],"answers.":[72,93],"However,":[73],"this":[74],"less":[77],"effective":[78],"at":[79],"perturbing":[80],"model":[82,90],"change":[84],"behavior":[86],"driving":[88],"towards":[91],"correct":[92]},"counts_by_year":[],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2025-10-10T00:00:00"}