{"id":"https://openalex.org/W6986681499","doi":null,"title":"Quantization of Large Language Models and the Impact on Output Performance","display_name":"Quantization of Large Language Models and the Impact on Output Performance","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W6986681499"},"language":"en","primary_location":{"id":"pmh:oai:lup-student-papers.lub.lu.se:9178256","doi":null,"is_oa":true,"landing_page_url":"http://lup.lub.lu.se/student-papers/record/9178256","pdf_url":"http://lup.lub.lu.se/student-papers/record/9178256/file/9178257.pdf","source":{"id":"https://openalex.org/S4306402420","display_name":"Lund University Publications Student Papers (Lund University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I187531555","host_organization_name":"Lund University","host_organization_lineage":["https://openalex.org/I187531555"],"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":"M2"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"http://lup.lub.lu.se/student-papers/record/9178256/file/9178257.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Baki Davidsson, Robin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Baki Davidsson, Robin","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":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":true,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9282000064849854,"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":0.9282000064849854,"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":0.04320000112056732,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.003700000001117587,"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/large-language-models","display_name":"large language models","score":0.8799999952316284},{"id":"https://openalex.org/keywords/quantization","display_name":"quantization","score":0.859000027179718},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.7630000114440918},{"id":"https://openalex.org/keywords/bit-precision","display_name":"bit precision","score":0.9620000123977661},{"id":"https://openalex.org/keywords/low-bit-quantization","display_name":"low-bit quantization","score":0.5519999861717224},{"id":"https://openalex.org/keywords/model-size","display_name":"model size","score":0.5649999976158142},{"id":"https://openalex.org/keywords/resource-constrained-deployment","display_name":"resource-constrained deployment","score":0.7870000004768372},{"id":"https://openalex.org/keywords/program-comprehension","display_name":"program comprehension","score":0.9959999918937683},{"id":"https://openalex.org/keywords/reading-comprehension","display_name":"reading comprehension","score":0.9800000190734863},{"id":"https://openalex.org/keywords/textual-coherence","display_name":"text coherence","score":0.7459999918937683},{"id":"https://openalex.org/keywords/mmlu-pro","display_name":"MMLU-Pro","score":0.8619999885559082}],"concepts":[{"id":"https://openalex.org/C144251240","wikidata":"https://www.wikidata.org/wiki/Q7068229","display_name":"Nucleofection","level":4,"score":0.4636000096797943},{"id":"https://openalex.org/C2992336715","wikidata":"https://www.wikidata.org/wiki/Q63431143","display_name":"Gestational period","level":4,"score":0.39250001311302185},{"id":"https://openalex.org/C2778283623","wikidata":"https://www.wikidata.org/wiki/Q18032200","display_name":"TSG101","level":5,"score":0.3452000021934509},{"id":"https://openalex.org/C2777054765","wikidata":"https://www.wikidata.org/wiki/Q6402731","display_name":"Dysgeusia","level":3,"score":0.34119999408721924},{"id":"https://openalex.org/C180938184","wikidata":"https://www.wikidata.org/wiki/Q2142270","display_name":"Liquation","level":3,"score":0.335999995470047},{"id":"https://openalex.org/C18743360","wikidata":"https://www.wikidata.org/wiki/Q1208096","display_name":"Diafiltration","level":4,"score":0.3246999979019165},{"id":"https://openalex.org/C2777968768","wikidata":"https://www.wikidata.org/wiki/Q1280161","display_name":"Emperipolesis","level":4,"score":0.3199999928474426},{"id":"https://openalex.org/C2776781215","wikidata":"https://www.wikidata.org/wiki/Q83253","display_name":"Triacetin","level":2,"score":0.3158999979496002},{"id":"https://openalex.org/C104545631","wikidata":"https://www.wikidata.org/wiki/Q464858","display_name":"Demotion","level":3,"score":0.302700012922287},{"id":"https://openalex.org/C2777742743","wikidata":"https://www.wikidata.org/wiki/Q19904005","display_name":"Durvalumab","level":5,"score":0.30250000953674316},{"id":"https://openalex.org/C133074676","wikidata":"https://www.wikidata.org/wiki/Q428729","display_name":"Fusible alloy","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C2777935831","wikidata":"https://www.wikidata.org/wiki/Q3144949","display_name":"Hemopericardium","level":3,"score":0.296999990940094},{"id":"https://openalex.org/C145741570","wikidata":"https://www.wikidata.org/wiki/Q7251534","display_name":"Proteogenomics","level":5,"score":0.2946999967098236},{"id":"https://openalex.org/C2778843634","wikidata":"https://www.wikidata.org/wiki/Q2013780","display_name":"Ocrelizumab","level":4,"score":0.2937999963760376},{"id":"https://openalex.org/C2777158700","wikidata":"https://www.wikidata.org/wiki/Q1419356","display_name":"Hyporeflexia","level":3,"score":0.2912999987602234},{"id":"https://openalex.org/C135979968","wikidata":"https://www.wikidata.org/wiki/Q609809","display_name":"Protein isoform","level":5,"score":0.28700000047683716},{"id":"https://openalex.org/C2779645999","wikidata":"https://www.wikidata.org/wiki/Q1858477","display_name":"Frugality","level":2,"score":0.2858999967575073},{"id":"https://openalex.org/C2775917601","wikidata":"https://www.wikidata.org/wiki/Q1424050","display_name":"Sclerodactyly","level":4,"score":0.2786000072956085},{"id":"https://openalex.org/C2779627259","wikidata":"https://www.wikidata.org/wiki/Q779763","display_name":"Pretext","level":3,"score":0.27559998631477356},{"id":"https://openalex.org/C2909186138","wikidata":"https://www.wikidata.org/wiki/Q1500373","display_name":"Hyperlactatemia","level":2,"score":0.2736999988555908},{"id":"https://openalex.org/C2780904289","wikidata":"https://www.wikidata.org/wiki/Q426498","display_name":"Tantalum carbide","level":3,"score":0.26809999346733093},{"id":"https://openalex.org/C2781032047","wikidata":"https://www.wikidata.org/wiki/Q938793","display_name":"Articular cartilage damage","level":5,"score":0.2612000107765198},{"id":"https://openalex.org/C2778128915","wikidata":"https://www.wikidata.org/wiki/Q3782613","display_name":"Contractualism","level":2,"score":0.25920000672340393},{"id":"https://openalex.org/C2776714567","wikidata":"https://www.wikidata.org/wiki/Q7861081","display_name":"Typhoid vaccine","level":3,"score":0.257999986410141},{"id":"https://openalex.org/C2776356786","wikidata":"https://www.wikidata.org/wiki/Q1048573","display_name":"Tubulopathy","level":3,"score":0.25519999861717224}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:lup-student-papers.lub.lu.se:9178256","doi":null,"is_oa":true,"landing_page_url":"http://lup.lub.lu.se/student-papers/record/9178256","pdf_url":"http://lup.lub.lu.se/student-papers/record/9178256/file/9178257.pdf","source":{"id":"https://openalex.org/S4306402420","display_name":"Lund University Publications Student Papers (Lund University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I187531555","host_organization_name":"Lund University","host_organization_lineage":["https://openalex.org/I187531555"],"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":"M2"}],"best_oa_location":{"id":"pmh:oai:lup-student-papers.lub.lu.se:9178256","doi":null,"is_oa":true,"landing_page_url":"http://lup.lub.lu.se/student-papers/record/9178256","pdf_url":"http://lup.lub.lu.se/student-papers/record/9178256/file/9178257.pdf","source":{"id":"https://openalex.org/S4306402420","display_name":"Lund University Publications Student Papers (Lund University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I187531555","host_organization_name":"Lund University","host_organization_lineage":["https://openalex.org/I187531555"],"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":"M2"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W6986681499.pdf","grobid_xml":"https://content.openalex.org/works/W6986681499.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W2525332836","https://openalex.org/W2626778328","https://openalex.org/W2787214294","https://openalex.org/W2963695965","https://openalex.org/W3030163527","https://openalex.org/W3175306105","https://openalex.org/W3202388704","https://openalex.org/W4307934016","https://openalex.org/W4380714727","https://openalex.org/W4387561528","https://openalex.org/W4388979610","https://openalex.org/W4390602555","https://openalex.org/W4390722772","https://openalex.org/W4391670689","https://openalex.org/W4393147132","https://openalex.org/W4401306886","https://openalex.org/W4401306979","https://openalex.org/W4402531536","https://openalex.org/W4402672004","https://openalex.org/W4402683730","https://openalex.org/W4415796274","https://openalex.org/W7133224126"],"related_works":["https://openalex.org/W7161755584","https://openalex.org/W7075993908","https://openalex.org/W4414360219","https://openalex.org/W4396746969","https://openalex.org/W7214925610","https://openalex.org/W7131740665","https://openalex.org/W4415030951","https://openalex.org/W4396881553","https://openalex.org/W7155032781","https://openalex.org/W4392340571"],"abstract_inverted_index":{"Recent":[0],"advancements":[1],"in":[2,11,126,150],"large":[3,61,74],"language":[4,13,62,75],"models":[5,63,76,169,172,192],"(LLMs)":[6],"have":[7],"demonstrated":[8],"remarkable":[9],"capabilities":[10],"natural":[12],"processing":[14],"tasks.However,":[15],"their":[16],"substantial":[17,144],"size":[18,48],"often":[19],"poses":[20],"significant":[21,184],"challenges":[22],"for":[23,89,94,99,220],"deployment":[24],"on":[25,55,82],"resource":[26,201],"constrained":[27],"devices.This":[28],"necessitates":[29],"the":[30,56,70,87,120,163,208],"development":[31],"of":[32,60,72],"efficient":[33],"model":[34,47,211],"compression":[35],"techniques,":[36],"with":[37,123],"quantization":[38,44,79],"emerging":[39],"as":[40,86],"a":[41,65,106],"leading":[42],"solution.While":[43],"can":[45,181],"reduce":[46],"and":[49,58,91,97,102,146,170,200,214],"improve":[50],"computational":[51],"efficiency,":[52,199],"its":[53],"impact":[54,164],"performance":[57,71,145,185],"accuracy":[59],"remains":[64],"critical":[66],"question.This":[67],"thesis":[68],"investigates":[69],"eight":[73],"across":[77,167],"various":[78],"levels,":[80],"focusing":[81],"key":[83],"tasks":[84],"such":[85],"MMLU-Pro":[88],"knowledge":[90],"reasoning,":[92],"CRUXEval":[93],"code":[95],"comprehension,":[96],"MuSR":[98],"reading":[100],"comprehension":[101],"coherence.The":[103],"results":[104,203],"reveal":[105],"consistent":[107],"trend:":[108],"higher":[109],"bit":[110,129,158],"precision":[111,130,159,190],"(e.g.,":[112,131],"8-bit":[113],"Q8_0)":[114],"generally":[115],"yields":[116],"better":[117],"performance,":[118,138,162,216],"although":[119],"improvements":[121],"diminish":[122],"further":[124],"increases":[125],"precision.Contrarily,":[127],"lower":[128,135,157],"2-bit":[132],"Q2_K)":[133],"show":[134],"but":[136,139],"acceptable":[137],"there":[140],"are":[141],"instances":[142],"where":[143],"coherence":[147],"losses":[148],"occur":[149],"some":[151],"models.The":[152],"findings":[153],"indicate":[154],"that":[155],"while":[156],"typically":[160],"reduces":[161],"varies":[165],"significantly":[166],"different":[168],"tasks.Larger":[171],"exhibit":[173],"greater":[174],"resilience":[175],"to":[176],"aggressive":[177],"quantization,":[178,213],"though":[179],"they":[180],"still":[182],"suffer":[183],"drops":[186],"at":[187],"very":[188],"low":[189],"levels.Mid-sized":[191],"offer":[193],"an":[194],"optimal":[195],"balance":[196],"between":[197,210],"capability,":[198],"usage.These":[202],"provide":[204],"valuable":[205],"insights":[206],"into":[207],"trade-offs":[209],"size,":[212],"task":[215],"offering":[217],"practical":[218],"guidance":[219],"real-world":[221],"applications.":[222]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}