{"id":"https://openalex.org/W7213402146","doi":"https://doi.org/10.48550/arxiv.2609.16391","title":"Where Post-Training Quantization Breaks Text Embedders: A Measured Map Across Four Embedder Families","display_name":"Where Post-Training Quantization Breaks Text Embedders: A Measured Map Across Four Embedder Families","publication_year":2026,"publication_date":"2026-09-14","ids":{"openalex":"https://openalex.org/W7213402146","doi":"https://doi.org/10.48550/arxiv.2609.16391"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2609.16391","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16391","pdf_url":"https://arxiv.org/pdf/2609.16391","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":null,"license_id":null,"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/2609.16391","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5080589950","display_name":"HyoJung Han","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Han, Hyojung","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5080589950"],"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9023000001907349,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9023000001907349,"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/T10028","display_name":"Topic Modeling","score":0.05620000138878822,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.005799999926239252,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/post-training-quantization","display_name":"post-training quantization","score":0.972000002861023},{"id":"https://openalex.org/keywords/weight-only-quantization","display_name":"weight-only quantization","score":0.8679999709129333},{"id":"https://openalex.org/keywords/text-embeddings","display_name":"text embeddings","score":0.6439999938011169},{"id":"https://openalex.org/keywords/bit-width-allocation","display_name":"bit-width allocation","score":0.6499999761581421},{"id":"https://openalex.org/keywords/group-size","display_name":"group size","score":0.503000020980835},{"id":"https://openalex.org/keywords/knowledge-distillation","display_name":"knowledge distillation","score":0.9319999814033508},{"id":"https://openalex.org/keywords/ndcg","display_name":"NDCG","score":0.4099999964237213},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.35199999809265137}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6668000221252441},{"id":"https://openalex.org/C43364308","wikidata":"https://www.wikidata.org/wiki/Q8799","display_name":"Byte","level":2,"score":0.6419000029563904},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5286999940872192},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5254999995231628},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5123000144958496},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.45019999146461487},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.4449999928474426},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4180999994277954},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.388700008392334},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.33390000462532043},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.33180001378059387},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.3310999870300293},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3303999900817871},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.29670000076293945},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2865000069141388},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.274399995803833},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.2685000002384186},{"id":"https://openalex.org/C93518851","wikidata":"https://www.wikidata.org/wiki/Q180160","display_name":"Metadata","level":2,"score":0.26809999346733093},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26739999651908875},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.2590999901294708},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2567000091075897},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.2535000145435333},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.25209999084472656},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.25189998745918274},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.16391","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16391","pdf_url":"https://arxiv.org/pdf/2609.16391","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.16391","doi":"https://doi.org/10.48550/arxiv.2609.16391","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.16391","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:2609.16391","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16391","pdf_url":"https://arxiv.org/pdf/2609.16391","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":null,"license_id":null,"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":0.5037262439727783,"display_name":"Quality education"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7213402146.pdf","grobid_xml":"https://content.openalex.org/works/W7213402146.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Weight-only":[0],"post-training":[1],"quantization":[2,41],"is":[3,124,167,191],"the":[4,13,21,70,92,102,120,132,141,188,209,230,249,253],"cheapest":[5],"way":[6],"to":[7,82,127,157,182],"shrink":[8],"a":[9,30,60,115,192],"retrieval":[10,49],"embedder,":[11],"and":[12,65,68,73,136,146,207,224,248],"received":[14],"advice":[15,47],"for":[16,169,178,256],"applying":[17],"it":[18,232],"--":[19,36,226],"protect":[20],"embedding":[22,87],"table,":[23],"allocate":[24,128],"bits":[25],"by":[26],"module":[27],"sensitivity,":[28],"prefer":[29],"ranking-aware":[31],"objective":[32],"over":[33],"weight":[34],"reconstruction":[35,150,165],"was":[37,233],"carried":[38],"into":[39],"LLM":[40],"largely":[42],"intact.":[43],"We":[44],"test":[45],"that":[46,242],"on":[48,221],"embedders":[50],"directly,":[51],"quantizing":[52],"five":[53],"checkpoints":[54],"from":[55,155],"four":[56],"architecture":[57],"families":[58],"across":[59,187],"grid":[61,190],"of":[62,107,143,160,212,240,259],"bit":[63,172],"widths":[64,173],"group":[66],"sizes,":[67],"isolating":[69],"embedding,":[71],"attention":[72],"feed-forward":[74],"blocks":[75],"at":[76,118,130,147,199,219],"each":[77],"width.":[78],"Every":[79],"heuristic":[80],"fails":[81],"transfer":[83],"as":[84,91,114],"stated.":[85],"The":[86],"table":[88],"never":[89],"emerges":[90],"dominant":[93],"isolated":[94],"protection":[95],"priority":[96],"in":[97,105,204],"any":[98],"family,":[99],"despite":[100],"being":[101,140],"largest":[103],"tensor":[104],"several":[106],"them.":[108,260],"Module":[109],"sensitivity":[110],"does":[111],"not":[112],"survive":[113],"transferable":[116],"ordering:":[117],"INT4/g16":[119],"spread":[121],"between":[122],"modules":[123],"too":[125],"small":[126],"against,":[129],"INT3":[131,200],"ordering":[133],"becomes":[134],"family-dependent":[135],"joint":[137],"damage":[138],"stops":[139],"sum":[142],"its":[144,184,213],"parts,":[145],"INT2":[148],"comparable":[149],"error":[151],"accompanies":[152],"retention":[153],"ranging":[154],"1.3":[156],"65.9":[158],"percent":[159],"full":[161],"precision.":[162],"A":[163,195],"cheap":[164],"proxy":[166],"useful":[168],"screening":[170],"uniform":[171],"but":[174,227],"substantially":[175],"less":[176],"reliable":[177],"choosing":[179],"which":[180],"tensors":[181],"protect;":[183],"apparent":[185],"strength":[186],"whole":[189],"range-extension":[193],"artifact.":[194],"distilled":[196,234],"109M":[197],"student":[198],"holds":[201],"78.04":[202],"NDCG@10":[203],"68.4":[205],"MB":[206,218],"dominates":[208],"extreme-PTQ":[210],"arm":[211],"own":[214],"0.6B":[215],"teacher,":[216],"297.9":[217],"64.46,":[220],"both":[222],"size":[223],"quality":[225],"only":[228],"inside":[229],"task":[231],"for.":[235],"Sizes":[236],"are":[237],"byte":[238,254],"counts":[239],"files":[241],"exist":[243],"rather":[244],"than":[245],"arithmetic":[246],"estimates,":[247],"measurement":[250],"repository":[251],"carries":[252],"provenance":[255],"every":[257],"one":[258]},"counts_by_year":[],"updated_date":"2026-09-22T17:25:34.749102","created_date":"2026-09-17T00:00:00"}