{"id":"https://openalex.org/W7127316264","doi":"https://doi.org/10.48550/arxiv.2602.00158","title":"RAPTOR: Ridge-Adaptive Logistic Probes","display_name":"RAPTOR: Ridge-Adaptive Logistic Probes","publication_year":2026,"publication_date":"2026-01-29","ids":{"openalex":"https://openalex.org/W7127316264","doi":"https://doi.org/10.48550/arxiv.2602.00158"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2602.00158","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.00158","pdf_url":"https://arxiv.org/pdf/2602.00158","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":"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/2602.00158","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124925736","display_name":"Ziqi Gao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gao, Ziqi","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062466225","display_name":"Yaotian Zhu","orcid":"https://orcid.org/0009-0004-8174-6839","observed_orcids":["https://orcid.org/0009-0004-8174-6839"]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhu, Yaotian","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134461697","display_name":"Qingcheng Zeng","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zeng, Qingcheng","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124919454","display_name":"Xu Zhao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhao, Xu","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124966335","display_name":"Ziqing Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wang, Ziqing","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101587259","display_name":"Feng Ruan","orcid":"https://orcid.org/0000-0003-1336-8229","observed_orcids":["https://orcid.org/0000-0003-1336-8229"]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ruan, Feng","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5124947248","display_name":"Kaize Ding","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I111979921","display_name":"Northwestern University","ror":"https://ror.org/000e0be47","country_code":"US","type":"education","lineage":["https://openalex.org/I111979921"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ding, Kaize","raw_affiliation_strings":["Northwestern University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Northwestern University","institution_ids":["https://openalex.org/I111979921"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I111979921"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":2,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9603999853134155,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9603999853134155,"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":0.01590000092983246,"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":0.011599999852478504,"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/ridge-logistic-regression","display_name":"ridge logistic regression","score":0.7990000247955322},{"id":"https://openalex.org/keywords/concept-vectors","display_name":"concept vectors","score":0.7360000014305115},{"id":"https://openalex.org/keywords/activation-steering","display_name":"activation steering","score":0.9139999747276306},{"id":"https://openalex.org/keywords/directional-stability","display_name":"directional stability","score":0.5600000023841858},{"id":"https://openalex.org/keywords/tikhonov-regularization","display_name":"Tikhonov regularization","score":0.8659999966621399},{"id":"https://openalex.org/keywords/teacher-student-models","display_name":"teacher-student models","score":0.8999999761581421},{"id":"https://openalex.org/keywords/instruction-tuned-llms","display_name":"instruction-tuned LLMs","score":0.7279999852180481}],"concepts":[{"id":"https://openalex.org/C151956035","wikidata":"https://www.wikidata.org/wiki/Q1132755","display_name":"Logistic regression","level":2,"score":0.6226999759674072},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.6212999820709229},{"id":"https://openalex.org/C32277403","wikidata":"https://www.wikidata.org/wiki/Q740445","display_name":"Ridge","level":2,"score":0.5480999946594238},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5084999799728394},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.49390000104904175},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.492000013589859},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4902999997138977},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4666000008583069},{"id":"https://openalex.org/C39891107","wikidata":"https://www.wikidata.org/wiki/Q5767098","display_name":"Hinge loss","level":3,"score":0.4481000006198883},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.42309999465942383},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4228000044822693},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.40540000796318054},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4020000100135803},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3968000113964081},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.38679999113082886},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.36419999599456787},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.3531999886035919},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.35100001096725464},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3359000086784363},{"id":"https://openalex.org/C180802074","wikidata":"https://www.wikidata.org/wiki/Q1052379","display_name":"Logistic function","level":2,"score":0.3345000147819519},{"id":"https://openalex.org/C44000306","wikidata":"https://www.wikidata.org/wiki/Q244330","display_name":"Hinge","level":2,"score":0.3100999891757965},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.30889999866485596},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.3043999969959259},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.3019999861717224},{"id":"https://openalex.org/C166550679","wikidata":"https://www.wikidata.org/wiki/Q263400","display_name":"Gaussian network model","level":3,"score":0.3005000054836273},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.29910001158714294},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.2919999957084656},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.2892000079154968},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.2831999957561493},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2815999984741211},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2603999972343445}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2602.00158","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.00158","pdf_url":"https://arxiv.org/pdf/2602.00158","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2602.00158","doi":"https://doi.org/10.48550/arxiv.2602.00158","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.00158","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:2602.00158","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.00158","pdf_url":"https://arxiv.org/pdf/2602.00158","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7127316264.pdf","grobid_xml":"https://content.openalex.org/works/W7127316264.grobid-xml"},"referenced_works_count":3,"referenced_works":["https://openalex.org/W2529194139","https://openalex.org/W4404986049","https://openalex.org/W7130610581"],"related_works":["https://openalex.org/W4403818765","https://openalex.org/W7220648582","https://openalex.org/W7213414758","https://openalex.org/W7223003689","https://openalex.org/W7125133517","https://openalex.org/W7141371174","https://openalex.org/W4417046140","https://openalex.org/W3198672493","https://openalex.org/W7161695532","https://openalex.org/W7160552723"],"abstract_inverted_index":{"Probing":[0],"studies":[1],"what":[2],"information":[3],"is":[4,35],"encoded":[5],"in":[6,28,119,158,164],"a":[7,14,31,38,50,88,151],"frozen":[8],"LLM's":[9],"layer":[10,51],"representations":[11],"by":[12,46,79,136],"training":[13,129],"lightweight":[15],"predictor":[16],"on":[17,63,106,189],"top":[18],"of":[19,59,154],"them.":[20],"Beyond":[21],"analysis,":[22],"probes":[23],"are":[24,68,134],"often":[25],"used":[26],"operationally":[27],"probe-then-steer":[29],"pipelines:":[30],"learned":[32],"concept":[33,65,98,111],"vector":[34],"extracted":[36],"from":[37,100],"probe":[39,92,174],"and":[40,74,109,126,176,179],"injected":[41],"via":[42],"additive":[43],"activation":[44],"steering":[45,139],"adding":[47],"it":[48],"to":[49,76],"representation":[52],"during":[53],"the":[54,143,165],"forward":[55],"pass.":[56],"The":[57],"effectiveness":[58],"this":[60],"pipeline":[61],"hinges":[62],"estimating":[64],"vectors":[66,99],"that":[67,183],"accurate,":[69],"directionally":[70],"stable":[71],"under":[72],"ablation,":[73],"inexpensive":[75],"obtain.":[77],"Motivated":[78],"these":[80,131],"desiderata,":[81],"we":[82,149],"propose":[83],"RAPTOR":[84,113],"(Ridge-Adaptive":[85],"Logistic":[86],"Probe),":[87],"simple":[89],"L2-regularized":[90],"logistic":[91,156],"whose":[93],"validation-tuned":[94],"ridge":[95,155],"strength":[96,172],"yields":[97],"normalized":[101],"weights.":[102],"Across":[103],"extensive":[104],"experiments":[105],"instruction-tuned":[107],"LLMs":[108],"human-written":[110],"datasets,":[112],"matches":[114],"or":[115],"exceeds":[116],"strong":[117],"baselines":[118],"accuracy":[120,175],"while":[121],"achieving":[122],"competitive":[123],"directional":[124],"stability":[125,178],"substantially":[127],"lower":[128],"cost;":[130],"quantitative":[132],"results":[133],"supported":[135],"qualitative":[137],"downstream":[138],"demonstrations.":[140],"Finally,":[141],"using":[142],"Convex":[144],"Gaussian":[145,161],"Min-max":[146],"Theorem":[147],"(CGMT),":[148],"provide":[150],"mechanistic":[152],"characterization":[153],"regression":[157],"an":[159],"idealized":[160],"teacher-student":[162],"model":[163],"high-dimensional":[166],"few-shot":[167],"regime,":[168],"explaining":[169],"how":[170],"penalty":[171],"mediates":[173],"concept-vector":[177],"yielding":[180],"structural":[181],"predictions":[182],"qualitatively":[184],"align":[185],"with":[186],"trends":[187],"observed":[188],"real":[190],"LLM":[191],"embeddings.":[192]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-11-12T12:28:57.059296","created_date":"2026-02-04T00:00:00"}