{"id":"https://openalex.org/W7156990019","doi":"https://doi.org/10.48550/arxiv.2604.24008","title":"Coverage-Based Calibration for Post-Training Quantization via Weighted Set Cover over Outlier Channels","display_name":"Coverage-Based Calibration for Post-Training Quantization via Weighted Set Cover over Outlier Channels","publication_year":2026,"publication_date":"2026-04-27","ids":{"openalex":"https://openalex.org/W7156990019","doi":"https://doi.org/10.48550/arxiv.2604.24008"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.24008","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24008","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.24008","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134780877","display_name":"Ibne Farabi Shihab","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Shihab, Ibne Farabi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5119832027","display_name":"Sanjeda Akter","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Akter, Sanjeda","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134767452","display_name":"Anuj Sharma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sharma, Anuj","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5134780877"],"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/T10181","display_name":"Natural Language Processing Techniques","score":0.19130000472068787,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.19130000472068787,"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":0.14710000157356262,"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":0.09220000356435776,"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/outlier","display_name":"Outlier","score":0.7695000171661377},{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.6959999799728394},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5881999731063843},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.4016999900341034},{"id":"https://openalex.org/keywords/anomaly-detection","display_name":"Anomaly detection","score":0.36559998989105225},{"id":"https://openalex.org/keywords/rendering","display_name":"Rendering (computer graphics)","score":0.35899999737739563},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.3515999913215637},{"id":"https://openalex.org/keywords/cover","display_name":"Cover (algebra)","score":0.32690000534057617}],"concepts":[{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.7695000171661377},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.6959999799728394},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.6057999730110168},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5881999731063843},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4950999915599823},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.48350000381469727},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.4016999900341034},{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.36559998989105225},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3628000020980835},{"id":"https://openalex.org/C205711294","wikidata":"https://www.wikidata.org/wiki/Q176953","display_name":"Rendering (computer graphics)","level":2,"score":0.35899999737739563},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.3515999913215637},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.32690000534057617},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31850001215934753},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.30219998955726624},{"id":"https://openalex.org/C51823790","wikidata":"https://www.wikidata.org/wiki/Q504353","display_name":"Greedy algorithm","level":2,"score":0.299699991941452},{"id":"https://openalex.org/C30684385","wikidata":"https://www.wikidata.org/wiki/Q176509","display_name":"Ringing","level":3,"score":0.29670000076293945},{"id":"https://openalex.org/C2779346075","wikidata":"https://www.wikidata.org/wiki/Q7268763","display_name":"Quality Score","level":3,"score":0.29330000281333923},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.28630000352859497},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2815000116825104},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.2734000086784363},{"id":"https://openalex.org/C2776292839","wikidata":"https://www.wikidata.org/wiki/Q5179217","display_name":"Coverage probability","level":3,"score":0.2720000147819519},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.2671999931335449},{"id":"https://openalex.org/C2834757","wikidata":"https://www.wikidata.org/wiki/Q4925424","display_name":"Monotone polygon","level":2,"score":0.26579999923706055},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.257099986076355}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.24008","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24008","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.24008","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.24008","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Post-Training":[0],"Quantization":[1],"(PTQ)":[2],"compresses":[3],"large":[4,43],"language":[5],"models":[6],"to":[7,35,48,192,204],"low":[8],"bit-widths":[9],"using":[10],"a":[11,27,89,130,221,226,236,241],"small":[12,179],"calibration":[13,32,70,86,180,197,214,229],"set,":[14],"and":[15,53,84,102,112,154,158,161,171,198,240],"its":[16],"quality":[17,71],"depends":[18],"strongly":[19],"on":[20,108],"which":[21,31],"samples":[22,33],"are":[23],"chosen.":[24],"We":[25,119],"identify":[26],"failure":[28],"mode":[29],"in":[30],"fail":[34],"activate":[36],"outlier":[37,95,233],"channels,":[38],"hidden":[39],"dimensions":[40],"with":[41,174,184,206,235],"unusually":[42],"activations,":[44],"causing":[45],"the":[46,103,123,141,175],"quantizer":[47],"underestimate":[49],"their":[50],"dynamic":[51],"range":[52],"producing":[54],"per-channel":[55],"reconstruction":[56],"errors":[57],"that":[58,68,122],"dominate":[59],"layer-wise":[60],"loss.":[61],"Motivated":[62],"by":[63,75,80,190,202],"this":[64],"observation,":[65],"we":[66],"argue":[67],"PTQ":[69,223],"is":[72,99,126,219],"governed":[73],"more":[74],"weighted":[76,90,135,142,232],"outlier-channel":[77],"coverage":[78,136,143],"than":[79,148],"generic":[81],"sample":[82],"representativeness,":[83],"formulate":[85],"selection":[87,230],"as":[88,145,231],"set":[91],"cover":[92],"problem":[93],"over":[94,167,195],"channels.":[96],"The":[97,217],"objective":[98,144],"monotone":[100],"submodular,":[101],"greedy":[104],"algorithm,":[105],"COVERCAL,":[106],"operates":[107],"pre-computed":[109],"activation":[110],"statistics":[111],"requires":[113],"no":[114],"GPU":[115],"time":[116],"at":[117,178,215],"selection.":[118],"further":[120],"show":[121],"weight":[124],"choice":[125],"internally":[127],"consistent:":[128],"under":[129,156],"stylized":[131],"clipping":[132],"model,":[133],"missed":[134],"upper-bounds":[137],"surrogate":[138],"loss,":[139],"justifying":[140],"principled":[146],"rather":[147],"purely":[149],"empirical.":[150],"Across":[151],"LLaMA-2,":[152],"LLaMA-3,":[153],"Mistral,":[155],"AWQ":[157],"GPTQ":[159],"backends":[160],"five":[162],"downstream":[163],"evaluations,":[164],"COVERCAL":[165,187],"improves":[166,188],"random,":[168],"max-perplexity,":[169],"max-activation-variance,":[170],"stratified":[172],"baselines,":[173],"largest":[176],"gains":[177],"budgets.":[181],"At":[182],"INT4":[183],"128":[185],"samples,":[186,208],"MMLU":[189],"1.2":[191],"1.5":[193],"points":[194],"random":[196,213],"reduces":[199],"perplexity":[200],"degradation":[201],"15":[203],"30\\%;":[205],"64":[207],"it":[209],"matches":[210],"or":[211],"exceeds":[212],"256.":[216],"contribution":[218],"not":[220],"new":[222],"backend":[224],"but":[225],"formulation":[227],"of":[228],"coverage,":[234],"simple,":[237],"efficient":[238],"algorithm":[239],"surrogate-based":[242],"justification.":[243]},"counts_by_year":[],"updated_date":"2026-04-29T06:16:36.941037","created_date":"2026-04-29T00:00:00"}