{"id":"https://openalex.org/W7127079066","doi":"https://doi.org/10.48550/arxiv.2601.22531","title":"Learn from A Rationalist: Distilling Intermediate Interpretable Rationales","display_name":"Learn from A Rationalist: Distilling Intermediate Interpretable Rationales","publication_year":2026,"publication_date":"2026-01-30","ids":{"openalex":"https://openalex.org/W7127079066","doi":"https://doi.org/10.48550/arxiv.2601.22531"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2601.22531","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2601.22531","pdf_url":"https://arxiv.org/pdf/2601.22531","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","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2601.22531","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121288306","display_name":"Jiayi Dai","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154425047","display_name":"University of Alberta","ror":"https://ror.org/0160cpw27","country_code":"CA","type":"education","lineage":["https://openalex.org/I154425047"]},{"id":"https://openalex.org/I283854653","display_name":"Machine Intelligence Research Institute","ror":"https://ror.org/01h4ass90","country_code":"US","type":"nonprofit","lineage":["https://openalex.org/I283854653"]},{"id":"https://openalex.org/I4405260365","display_name":"Alberta Machine Intelligence Institute","ror":"https://ror.org/03msd8560","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4405260365"]}],"countries":["CA","US"],"is_corresponding":false,"raw_author_name":"Dai, Jiayi","raw_affiliation_strings":["Alberta Machine Intelligence Institute , Edmon- ton , Canada","Department of Computing Science , University of Alberta , Ed- monton , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Alberta Machine Intelligence Institute , Edmon- ton , Canada","institution_ids":["https://openalex.org/I283854653","https://openalex.org/I4405260365"]},{"raw_affiliation_string":"Department of Computing Science , University of Alberta , Ed- monton , Canada","institution_ids":["https://openalex.org/I154425047"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5121293192","display_name":"Randy Goebel","orcid":"https://orcid.org/0000-0002-0739-2946","observed_orcids":["https://orcid.org/0000-0002-0739-2946"]},"institutions":[{"id":"https://openalex.org/I154425047","display_name":"University of Alberta","ror":"https://ror.org/0160cpw27","country_code":"CA","type":"education","lineage":["https://openalex.org/I154425047"]},{"id":"https://openalex.org/I283854653","display_name":"Machine Intelligence Research Institute","ror":"https://ror.org/01h4ass90","country_code":"US","type":"nonprofit","lineage":["https://openalex.org/I283854653"]},{"id":"https://openalex.org/I4405260365","display_name":"Alberta Machine Intelligence Institute","ror":"https://ror.org/03msd8560","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4405260365"]}],"countries":["CA","US"],"is_corresponding":false,"raw_author_name":"Goebel, Randy","raw_affiliation_strings":["Alberta Machine Intelligence Institute , Edmon- ton , Canada","Department of Computing Science , University of Alberta , Ed- monton , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Alberta Machine Intelligence Institute , Edmon- ton , Canada","institution_ids":["https://openalex.org/I283854653","https://openalex.org/I4405260365"]},{"raw_affiliation_string":"Department of Computing Science , University of Alberta , Ed- monton , Canada","institution_ids":["https://openalex.org/I154425047"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"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":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":1.0,"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":1.0,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.0,"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/rationale-extraction","display_name":"rationale extraction","score":0.972000002861023},{"id":"https://openalex.org/keywords/knowledge-distillation","display_name":"knowledge distillation","score":0.9070000052452087},{"id":"https://openalex.org/keywords/interpretability","display_name":"interpretability","score":0.4699999988079071},{"id":"https://openalex.org/keywords/deep-neural-network","display_name":"deep neural network","score":0.8610000014305115},{"id":"https://openalex.org/keywords/feature-selection","display_name":"feature selection","score":0.5870000123977661},{"id":"https://openalex.org/keywords/teacher-student-learning","display_name":"teacher-student learning","score":0.7170000076293945},{"id":"https://openalex.org/keywords/bert","display_name":"BERT","score":0.6320000290870667},{"id":"https://openalex.org/keywords/vision-transformer","display_name":"vision transformer","score":0.8320000171661377}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.9143000245094299},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7775999903678894},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7031000256538391},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6481000185012817},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5737000107765198},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.5419999957084656},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4724999964237213},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.42750000953674316},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.42399999499320984},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4147999882698059},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4104999899864197},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.3808000087738037},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.37290000915527344},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.37279999256134033},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.36010000109672546},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3564000129699707},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.3089999854564667},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.30730000138282776},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3068999946117401},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.30239999294281006},{"id":"https://openalex.org/C193415008","wikidata":"https://www.wikidata.org/wiki/Q639681","display_name":"Network architecture","level":2,"score":0.2989000082015991},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.28999999165534973},{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.28999999165534973},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.2896000146865845},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.28290000557899475},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.27630001306533813},{"id":"https://openalex.org/C157170001","wikidata":"https://www.wikidata.org/wiki/Q4781507","display_name":"Applications of artificial intelligence","level":2,"score":0.27489998936653137},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.267300009727478},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2639999985694885},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.2565000057220459},{"id":"https://openalex.org/C2779808786","wikidata":"https://www.wikidata.org/wiki/Q6664603","display_name":"Locality","level":2,"score":0.2547999918460846},{"id":"https://openalex.org/C2779808786","wikidata":"https://www.wikidata.org/wiki/Q6664603","display_name":"Locality","level":2,"score":0.2547000050544739},{"id":"https://openalex.org/C111030470","wikidata":"https://www.wikidata.org/wiki/Q1430460","display_name":"Curse of dimensionality","level":2,"score":0.25119999051094055}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2601.22531","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2601.22531","pdf_url":"https://arxiv.org/pdf/2601.22531","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.2601.22531","doi":"https://doi.org/10.48550/arxiv.2601.22531","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.22531","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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2601.22531","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2601.22531","pdf_url":"https://arxiv.org/pdf/2601.22531","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":[],"sustainable_development_goals_aurora":[{"id":"https://metadata.un.org/sdg/4","score":0.47356653213500977,"display_name":"Quality education"}],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320314212","display_name":"Alberta Machine Intelligence Institute","ror":"https://ror.org/03msd8560"},{"id":"https://openalex.org/F4320334593","display_name":"Natural Sciences and Engineering Research Council of Canada","ror":"https://ror.org/01h531d29"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7127079066.pdf","grobid_xml":"https://content.openalex.org/works/W7127079066.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Because":[0,173],"of":[1,5,16,25,66,71,80,106,141,174,179,197,205,237],"the":[2,14,56,60,64,78,93,103,120,137,150,175,180,195,234,238],"pervasive":[3],"use":[4],"deep":[6],"neural":[7,43,95,117,184],"networks":[8,44,96,118],"(DNNs),":[9],"especially":[10],"in":[11,77,147],"high-stakes":[12],"domains,":[13],"interpretability":[15],"DNNs":[17,36],"has":[18],"received":[19],"increased":[20],"attention.":[21],"The":[22],"general":[23],"idea":[24],"rationale":[26],"extraction":[27],"(RE)":[28],"is":[29,86],"to":[30,47,68,149,158],"provide":[31],"an":[32],"interpretable-by-design":[33],"framework":[34],"for":[35],"via":[37],"a":[38,131,142,145,190],"select-predict":[39],"architecture":[40],"where":[41,130],"two":[42],"learn":[45,166],"jointly":[46],"perform":[48],"feature":[49,83],"selection":[50],"and":[51,89,139,170,207,216,226],"prediction,":[52,63],"respectively.":[53],"Given":[54],"only":[55],"remote":[57],"supervision":[58],"from":[59,136,168],"final":[61],"task":[62],"process":[65],"learning":[67],"select":[69],"subsets":[70],"features":[72],"(or":[73],"rationales)":[74],"requires":[75],"searching":[76],"space":[79],"all":[81],"possible":[82],"combinations,":[84],"which":[85],"computationally":[87],"challenging":[88],"even":[90],"harder":[91],"when":[92],"base":[94],"are":[97,110],"not":[98],"sufficiently":[99],"capable.":[100],"To":[101,193],"improve":[102],"predictive":[104,235],"performance":[105,236],"RE":[107,133,153,159,240],"models":[108],"that":[109,230],"based":[111],"on":[112],"less":[113],"capable":[114],"or":[115],"smaller":[116],"(i.e.,":[119,144,220],"students),":[121],"we":[122,199],"propose":[123],"REKD":[124,231],"(Rationale":[125],"Extraction":[126],"with":[127,162,202],"Knowledge":[128],"Distillation)":[129],"student":[132,239],"model":[134],"learns":[135],"rationales":[138],"predictions":[140],"teacher":[143],"rationalist)":[146],"addition":[148],"student's":[151],"own":[152],"optimization.":[154],"This":[155],"structural":[156],"adjustment":[157],"aligns":[160],"well":[161],"how":[163],"humans":[164],"could":[165,186],"effectively":[167],"interpretable":[169],"verifiable":[171],"knowledge.":[172],"neural-model":[176],"agnostic":[177],"nature":[178],"method,":[181],"any":[182],"black-box":[183],"network":[185],"be":[187],"integrated":[188],"as":[189],"backbone":[191],"model.":[192],"demonstrate":[194],"viability":[196],"REKD,":[198],"conduct":[200],"experiments":[201,213],"multiple":[203],"variants":[204],"BERT":[206],"vision":[208,217],"transformer":[209],"(ViT)":[210],"models.":[211,241],"Our":[212],"across":[214],"language":[215],"classification":[218],"datasets":[219],"IMDB":[221],"movie":[222],"reviews,":[223],"CIFAR":[224,227],"10":[225],"100)":[228],"show":[229],"significantly":[232],"improves":[233]},"counts_by_year":[],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2026-02-03T00:00:00"}