{"id":"https://openalex.org/W7167942866","doi":"https://doi.org/10.48550/arxiv.2607.08046","title":"What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness","display_name":"What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness","publication_year":2026,"publication_date":"2026-07-09","ids":{"openalex":"https://openalex.org/W7167942866","doi":"https://doi.org/10.48550/arxiv.2607.08046"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.08046","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.08046","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2607.08046","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5082174299","display_name":"Rapha\u00ebl Sarfati","orcid":"https://orcid.org/0000-0003-4944-0632"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sarfati, Rapha\u00ebl","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021358316","display_name":"Pratyush Ranjan Tiwari","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tiwari, Pratyush Ranjan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128261556","display_name":"Siddharth Boppana","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Boppana, Siddharth","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123866981","display_name":"Christopher J. Earls","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Earls, Christopher J.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007593475","display_name":"Srikar Varadaraj","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Varadaraj, Srikar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5121819360","display_name":"Eric Ho","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ho, Eric","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/T10028","display_name":"Topic Modeling","score":0.26159998774528503,"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":0.26159998774528503,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.094200000166893,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13629","display_name":"Text Readability and Simplification","score":0.07800000160932541,"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/trace","display_name":"TRACE (psycholinguistics)","score":0.5648000240325928},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.47929999232292175},{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.4756999909877777},{"id":"https://openalex.org/keywords/window","display_name":"Window (computing)","score":0.47440001368522644},{"id":"https://openalex.org/keywords/question-answering","display_name":"Question answering","score":0.42160001397132874},{"id":"https://openalex.org/keywords/ask-price","display_name":"Ask price","score":0.3709999918937683},{"id":"https://openalex.org/keywords/term","display_name":"Term (time)","score":0.367900013923645}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6122000217437744},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.5648000240325928},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.47929999232292175},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.4756999909877777},{"id":"https://openalex.org/C2778751112","wikidata":"https://www.wikidata.org/wiki/Q835016","display_name":"Window (computing)","level":2,"score":0.47440001368522644},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46399998664855957},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.42160001397132874},{"id":"https://openalex.org/C90329073","wikidata":"https://www.wikidata.org/wiki/Q914232","display_name":"Ask price","level":2,"score":0.3709999918937683},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.367900013923645},{"id":"https://openalex.org/C2776325391","wikidata":"https://www.wikidata.org/wiki/Q6917865","display_name":"Motivated reasoning","level":3,"score":0.33399999141693115},{"id":"https://openalex.org/C97364631","wikidata":"https://www.wikidata.org/wiki/Q484284","display_name":"Deductive reasoning","level":2,"score":0.31929999589920044},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.28949999809265137},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.2870999872684479},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.28619998693466187},{"id":"https://openalex.org/C103057564","wikidata":"https://www.wikidata.org/wiki/Q4751139","display_name":"Analytic reasoning","level":3,"score":0.27790001034736633},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2718999981880188},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.271699994802475},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C62360110","wikidata":"https://www.wikidata.org/wiki/Q96777007","display_name":"Circumscription","level":2,"score":0.26579999923706055}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.08046","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.08046","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2607.08046","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.08046","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,"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"},"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":[1,194],"models":[2,199],"fine-tuned":[3],"for":[4,63,189],"forecasting":[5],"can":[6],"be":[7],"accurate":[8],"yet":[9],"poorly":[10],"calibrated,":[11],"and":[12,51,65,75,115,155,157,192,197],"their":[13,103],"chain-of-thought":[14],"(CoT)":[15],"reasoning":[16,93,112,145,198],"may":[17],"not":[18],"faithfully":[19],"reflect":[20],"the":[21,83,87,92,111,119,129,132,152,161],"evidence":[22,73],"behind":[23],"a":[24,32,58,147,186],"forecast.":[25],"We":[26,67],"ask":[27],"whether":[28],"internal":[29,183],"representations":[30,184],"offer":[31],"more":[33,200],"direct":[34],"window":[35],"into":[36],"both.":[37],"Working":[38],"with":[39,173],"Eternis-Forecaster":[40],"8B":[41],"on":[42,48],"OpenForesight,":[43],"we":[44],"train":[45],"representation-pooling":[46],"probes":[47,98],"intermediate":[49],"activations":[50,104],"find":[52],"they":[53,116],"achieve":[54],"substantially":[55],"better":[56,109],"calibration;":[57],"result":[59],"that":[60,139],"also":[61,117],"holds":[62],"GLM-4.7-Flash":[64],"GLM-4.5-Air.":[66],"then":[68],"assess":[69],"CoT":[70,130],"faithfulness":[71],"through":[72],"ablation":[74],"diversionary":[76],"injection:":[77],"removing":[78],"an":[79],"influential":[80],"source":[81],"in":[82,123],"prompt":[84],"often":[85],"changes":[86],"model's":[88],"forecast":[89],"while":[90],"leaving":[91],"trace":[94,113],"untouched.":[95],"The":[96],"same":[97],"function":[99],"as":[100,185],"lie":[101],"detectors:":[102],"track":[105],"behavioral":[106],"shifts":[107],"far":[108],"than":[110],"does,":[114],"predict":[118],"direction":[120],"of":[121,125,163,170,176],"change":[122],"84%":[124],"cases,":[126],"including":[127],"when":[128],"conceals":[131],"perturbation's":[133],"influence.":[134],"Finally,":[135],"forced":[136],"answering":[137],"reveals":[138],"forecasts":[140],"are":[141],"largely":[142],"fixed":[143],"before":[144],"begins:":[146],"single":[148],"pre-reasoning":[149],"pass":[150],"recovers":[151],"committed":[153],"answer":[154,166],"confidence,":[156],"routing":[158],"questions":[159],"by":[160],"spread":[162],"this":[164],"pre-set":[165],"distribution":[167],"saves":[168],"30-47%":[169],"generated":[171],"tokens,":[172],"no":[174],"loss":[175],"accuracy.":[177],"Together,":[178],"these":[179],"results":[180],"establish":[181],"probing":[182],"practical":[187],"tool":[188],"calibrating,":[190],"auditing,":[191],"triaging":[193],"model":[195],"forecasters":[196],"broadly.":[201]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-11T00:00:00"}