{"id":"https://openalex.org/W4391555547","doi":"https://doi.org/10.48550/arxiv.2402.00912","title":"Can we Constrain Concept Bottleneck Models to Learn Semantically Meaningful Input Features?","display_name":"Can we Constrain Concept Bottleneck Models to Learn Semantically Meaningful Input Features?","publication_year":2024,"publication_date":"2024-02-01","ids":{"openalex":"https://openalex.org/W4391555547","doi":"https://doi.org/10.48550/arxiv.2402.00912"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2402.00912","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.00912","pdf_url":"https://arxiv.org/pdf/2402.00912","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2402.00912","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5023953318","display_name":"Jack Furby","orcid":"https://orcid.org/0000-0002-2348-8091"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Furby, Jack","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037831458","display_name":"Daniel Cunnington","orcid":"https://orcid.org/0000-0003-0715-964X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cunnington, Daniel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012388184","display_name":"Dave Braines","orcid":"https://orcid.org/0000-0003-3296-0842"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Braines, Dave","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5005253980","display_name":"Alun Preece","orcid":"https://orcid.org/0000-0003-0349-9057"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Preece, Alun","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":0.9574999809265137,"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":0.9574999809265137,"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/T11303","display_name":"Bayesian Modeling and Causal Inference","score":0.9559000134468079,"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/bottleneck","display_name":"Bottleneck","score":0.8695513010025024},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6160032153129578},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4615018963813782},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3683493733406067}],"concepts":[{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.8695513010025024},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6160032153129578},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4615018963813782},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3683493733406067},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2402.00912","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.00912","pdf_url":"https://arxiv.org/pdf/2402.00912","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},{"id":"doi:10.48550/arxiv.2402.00912","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2402.00912","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2402.00912","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.00912","pdf_url":"https://arxiv.org/pdf/2402.00912","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2748952813","https://openalex.org/W1657880117","https://openalex.org/W2595172197","https://openalex.org/W2127970246","https://openalex.org/W2084856301","https://openalex.org/W1001352512","https://openalex.org/W4382618745","https://openalex.org/W2885125400","https://openalex.org/W1989889224","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Concept":[0],"Bottleneck":[1],"Models":[2],"(CBMs)":[3],"are":[4,19],"regarded":[5],"as":[6],"inherently":[7],"interpretable":[8],"because":[9],"they":[10],"first":[11],"predict":[12,22],"a":[13,23,37,58,65,137,165],"set":[14],"of":[15,106],"human-defined":[16],"concepts":[17,99,127,157],"which":[18],"used":[20],"to":[21,29,45,63,125,128,172,201,207],"task":[24],"label.":[25],"For":[26,51],"inherent":[27],"interpretability":[28],"be":[30],"fully":[31],"realised,":[32],"and":[33,98,144,188,192],"ensure":[34],"trust":[35],"in":[36,53],"model's":[38],"output,":[39],"it's":[40],"desirable":[41],"for":[42,152,169,178],"concept":[43,72,90,110,147,210],"predictions":[44,73],"use":[46],"semantically":[47,129,203],"meaningful":[48,130,204],"input":[49,78,96,131,142,176,205],"features.":[50,79],"instance,":[52,153],"an":[54,113],"image,":[55],"pixels":[56],"representing":[57],"broken":[59],"bone":[60],"should":[61],"contribute":[62],"predicting":[64],"fracture.":[66],"However,":[67],"current":[68],"literature":[69],"suggests":[70],"that":[71,82,121],"often":[74],"rely":[75],"on":[76,109,185],"irrelevant":[77],"We":[80,181],"hypothesise":[81],"this":[83,117],"occurs":[84],"when":[85],"dataset":[86,107],"labels":[87],"include":[88],"inaccurate":[89],"annotations,":[91],"or":[92],"the":[93,104,141,145,170,174,195,208],"relationship":[94],"between":[95,140],"features":[97,143,177,206],"is":[100,150],"unclear.":[101],"In":[102,116],"general,":[103],"effect":[105],"labelling":[108],"representations":[111],"remains":[112],"understudied":[114],"area.":[115],"paper,":[118],"we":[119],"demonstrate":[120,193],"CBMs":[122,198],"can":[123,199],"learn":[124,200],"map":[126],"features,":[132],"by":[133,154],"utilising":[134],"datasets":[135],"with":[136],"clear":[138,166],"link":[139],"desired":[146],"predictions.":[148,211],"This":[149],"achieved,":[151],"ensuring":[155],"multiple":[156],"do":[158],"not":[159],"always":[160],"co-occur":[161],"and,":[162],"therefore":[163],"provide":[164],"training":[167],"signal":[168],"CBM":[171],"distinguish":[173],"relevant":[175],"each":[179],"concept.":[180],"validate":[182],"our":[183],"hypothesis":[184],"both":[186],"synthetic":[187],"real-world":[189],"image":[190],"datasets,":[191],"under":[194],"correct":[196,209],"conditions,":[197],"attribute":[202]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2025-10-10T00:00:00"}