{"id":"https://openalex.org/W7127635518","doi":"https://doi.org/10.48550/arxiv.2602.02726","title":"Vector Quantized Latent Concepts: A Scalable Alternative to Clustering-Based Concept Discovery","display_name":"Vector Quantized Latent Concepts: A Scalable Alternative to Clustering-Based Concept Discovery","publication_year":2026,"publication_date":"2026-02-02","ids":{"openalex":"https://openalex.org/W7127635518","doi":"https://doi.org/10.48550/arxiv.2602.02726"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2602.02726","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.02726","pdf_url":"https://arxiv.org/pdf/2602.02726","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/2602.02726","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5109707656","display_name":"Xuemin Yu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I129902397","display_name":"Dalhousie University","ror":"https://ror.org/01e6qks80","country_code":"CA","type":"education","lineage":["https://openalex.org/I129902397"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Yu, Xuemin","raw_affiliation_strings":["Dalhousie University , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dalhousie University , Canada","institution_ids":["https://openalex.org/I129902397"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125011452","display_name":"Ankur Garg","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I168635309","display_name":"University of Calgary","ror":"https://ror.org/03yjb2x39","country_code":"CA","type":"education","lineage":["https://openalex.org/I168635309"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Garg, Ankur","raw_affiliation_strings":["University of Calgary , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Calgary , Canada","institution_ids":["https://openalex.org/I168635309"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032466547","display_name":"Samira Ebrahimi Kahou","orcid":"https://orcid.org/0009-0005-6283-3811","observed_orcids":["https://orcid.org/0009-0005-6283-3811"]},"institutions":[{"id":"https://openalex.org/I168635309","display_name":"University of Calgary","ror":"https://ror.org/03yjb2x39","country_code":"CA","type":"education","lineage":["https://openalex.org/I168635309"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Kahou, Samira Ebrahimi","raw_affiliation_strings":["University of Calgary , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Calgary , Canada","institution_ids":["https://openalex.org/I168635309"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5125073311","display_name":"Hassan Sajjad","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I129902397","display_name":"Dalhousie University","ror":"https://ror.org/01e6qks80","country_code":"CA","type":"education","lineage":["https://openalex.org/I129902397"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Sajjad, Hassan","raw_affiliation_strings":["Dalhousie University , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dalhousie University , Canada","institution_ids":["https://openalex.org/I129902397"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"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":0.9833999872207642,"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":0.9833999872207642,"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":0.01510000042617321,"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":9.999999747378752e-05,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/vector-quantization","display_name":"vector quantization","score":0.8199999928474426},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.847000002861023},{"id":"https://openalex.org/keywords/interpretability","display_name":"interpretability","score":0.7120000123977661},{"id":"https://openalex.org/keywords/hidden-state-representations","display_name":"hidden state representations","score":0.5320000052452087},{"id":"https://openalex.org/keywords/hierarchical-clustering","display_name":"hierarchical clustering","score":0.6259999871253967},{"id":"https://openalex.org/keywords/k-means-clustering","display_name":"K-means clustering","score":0.8379999995231628},{"id":"https://openalex.org/keywords/codebook-learning","display_name":"codebook learning","score":0.7450000047683716},{"id":"https://openalex.org/keywords/sparse-autoencoders","display_name":"sparse autoencoders","score":0.8309999704360962},{"id":"https://openalex.org/keywords/decoder-only-models","display_name":"decoder-only models","score":0.6359999775886536}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7419000267982483},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.7156999707221985},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.6723999977111816},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.618399977684021},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6157000064849854},{"id":"https://openalex.org/C127759330","wikidata":"https://www.wikidata.org/wiki/Q637416","display_name":"Codebook","level":2,"score":0.5785999894142151},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.5389999747276306},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.5153999924659729},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4050000011920929},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.3813999891281128},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.37229999899864197},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3718999922275543},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.36579999327659607},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3578999936580658},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.35760000348091125},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.31139999628067017},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3111000061035156},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2874000072479248},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.2840999960899353},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2727999985218048},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.2653999924659729},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.2653000056743622},{"id":"https://openalex.org/C39920170","wikidata":"https://www.wikidata.org/wiki/Q693083","display_name":"Soundness","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.26010000705718994},{"id":"https://openalex.org/C170133592","wikidata":"https://www.wikidata.org/wiki/Q1806883","display_name":"Latent semantic analysis","level":2,"score":0.257099986076355},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.2538999915122986}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2602.02726","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.02726","pdf_url":"https://arxiv.org/pdf/2602.02726","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.2602.02726","doi":"https://doi.org/10.48550/arxiv.2602.02726","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.02726","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:2602.02726","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2602.02726","pdf_url":"https://arxiv.org/pdf/2602.02726","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/W7127635518.pdf","grobid_xml":"https://content.openalex.org/works/W7127635518.grobid-xml"},"referenced_works_count":18,"referenced_works":["https://openalex.org/W2626639386","https://openalex.org/W2789543585","https://openalex.org/W2951384661","https://openalex.org/W2965373594","https://openalex.org/W3034346995","https://openalex.org/W3093205559","https://openalex.org/W3206395542","https://openalex.org/W4376312331","https://openalex.org/W4384918448","https://openalex.org/W4386076713","https://openalex.org/W4386839891","https://openalex.org/W4387985331","https://openalex.org/W4403345627","https://openalex.org/W4404403015","https://openalex.org/W4405253355","https://openalex.org/W4405903187","https://openalex.org/W7084403599","https://openalex.org/W7161063287"],"related_works":["https://openalex.org/W4411630325","https://openalex.org/W4415308934","https://openalex.org/W7131883130","https://openalex.org/W7126113542","https://openalex.org/W4403818765","https://openalex.org/W4416085688","https://openalex.org/W7216282081","https://openalex.org/W4417539449","https://openalex.org/W7140309328","https://openalex.org/W7171767714"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"encode":[4],"rich":[5],"semantic":[6],"information":[7,19],"in":[8,104,115],"their":[9],"hidden":[10,28,93],"states,":[11],"yet":[12],"it":[13],"remains":[14,113],"difficult":[15],"to":[16,52,56,102],"understand":[17],"what":[18],"these":[20],"internal":[21],"representations":[22],"capture.":[23],"Latent":[24,76],"concepts":[25,48,90,138],"extracted":[26],"from":[27],"states":[29],"offer":[30],"a":[31,42,79,86,129],"promising":[32],"direction":[33],"for":[34],"interpreting":[35],"LLMs,":[36],"but":[37,49,65],"existing":[38],"clustering-based":[39],"methods":[40],"face":[41],"trade-off:":[43],"hierarchical":[44,110],"clustering":[45],"produces":[46],"coherent":[47,70],"is":[50],"limited":[51],"small":[53],"datasets":[54],"due":[55],"its":[57],"quadratic":[58],"memory":[59],"cost,":[60,106],"while":[61],"K-Means":[62,103],"scales":[63,107],"efficiently":[64],"may":[66],"yield":[67],"less":[68],"semantically":[69],"concepts.":[71],"We":[72],"propose":[73],"Vector":[74],"Quantized":[75],"Concept":[77],"(VQLC),":[78],"discrete":[80],"concept":[81],"learning":[82],"framework":[83],"that":[84,135],"learns":[85],"codebook":[87],"of":[88],"latent":[89],"on":[91,121],"frozen":[92],"states.":[94],"Across":[95],"12":[96],"dataset-model":[97],"settings,":[98],"VQLC":[99],"stays":[100],"close":[101],"computational":[105],"better":[108],"than":[109],"clustering,":[111],"and":[112,128,141],"competitive":[114],"faithfulness,":[116],"with":[117],"the":[118,136],"clearest":[119],"gains":[120],"decoder-only":[122],"models.":[123],"LLMs-based":[124],"evaluation,":[125],"qualitative":[126],"analysis,":[127],"Sparse":[130],"Autoencoder":[131],"(SAE)":[132],"comparison":[133],"demonstrate":[134],"learned":[137],"are":[139],"interpretable":[140],"task-relevant.":[142]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-02-06T00:00:00"}