{"id":"https://openalex.org/W7006461566","doi":null,"title":"Towards Robust and Scalable Evaluation for Large Language Models","display_name":"Towards Robust and Scalable Evaluation for Large Language Models","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W7006461566"},"language":"en","primary_location":{"id":"pmh:oai:escholarship.org:ark:/13030/qt51r72944","doi":null,"is_oa":true,"landing_page_url":"https://escholarship.org/uc/item/51r72944","pdf_url":"https://escholarship.org/content/qt51r72944/qt51r72944.pdf","source":{"id":"https://openalex.org/S4306400115","display_name":"eScholarship (California Digital Library)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I2801248553","host_organization_name":"California Digital Library","host_organization_lineage":["https://openalex.org/I2801248553"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"etd"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://escholarship.org/content/qt51r72944/qt51r72944.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Chiang, Wei-Lin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Chiang, Wei-Lin","raw_affiliation_strings":["UC Berkeley"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UC Berkeley","institution_ids":["https://openalex.org/I95457486"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I95457486"],"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/T10028","display_name":"Topic Modeling","score":0.6538000106811523,"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":0.6538000106811523,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.2551000118255615,"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":0.013700000010430813,"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/large-language-models","display_name":"large language models","score":0.9110000133514404},{"id":"https://openalex.org/keywords/benchmark-contamination","display_name":"benchmark contamination","score":0.8080000281333923},{"id":"https://openalex.org/keywords/human-evaluation","display_name":"human evaluation","score":0.46399998664855957},{"id":"https://openalex.org/keywords/chatbot-arena","display_name":"Chatbot Arena","score":0.890999972820282},{"id":"https://openalex.org/keywords/human-preferences","display_name":"human preferences","score":0.5600000023841858},{"id":"https://openalex.org/keywords/llm-as-a-judge","display_name":"LLM-as-a-judge","score":0.9409999847412109},{"id":"https://openalex.org/keywords/conversational-ai","display_name":"conversational AI","score":0.6380000114440918},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.48399999737739563}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7488999962806702},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6456999778747559},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5315999984741211},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.4964999854564667},{"id":"https://openalex.org/C2779041454","wikidata":"https://www.wikidata.org/wiki/Q870780","display_name":"Chatbot","level":2,"score":0.4763999879360199},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46880000829696655},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.45010000467300415},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4189000129699707},{"id":"https://openalex.org/C62230096","wikidata":"https://www.wikidata.org/wiki/Q275969","display_name":"Crowdsourcing","level":2,"score":0.4043999910354614},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.39719998836517334},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.3253999948501587},{"id":"https://openalex.org/C153701036","wikidata":"https://www.wikidata.org/wiki/Q659974","display_name":"Trustworthiness","level":2,"score":0.31220000982284546},{"id":"https://openalex.org/C168065819","wikidata":"https://www.wikidata.org/wiki/Q845566","display_name":"Debugging","level":2,"score":0.2913999855518341},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.28279998898506165},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2793000042438507},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2624000012874603},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2515000104904175}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:escholarship.org:ark:/13030/qt51r72944","doi":null,"is_oa":true,"landing_page_url":"https://escholarship.org/uc/item/51r72944","pdf_url":"https://escholarship.org/content/qt51r72944/qt51r72944.pdf","source":{"id":"https://openalex.org/S4306400115","display_name":"eScholarship (California Digital Library)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I2801248553","host_organization_name":"California Digital Library","host_organization_lineage":["https://openalex.org/I2801248553"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"etd"}],"best_oa_location":{"id":"pmh:oai:escholarship.org:ark:/13030/qt51r72944","doi":null,"is_oa":true,"landing_page_url":"https://escholarship.org/uc/item/51r72944","pdf_url":"https://escholarship.org/content/qt51r72944/qt51r72944.pdf","source":{"id":"https://openalex.org/S4306400115","display_name":"eScholarship (California Digital Library)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I2801248553","host_organization_name":"California Digital Library","host_organization_lineage":["https://openalex.org/I2801248553"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"etd"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7006461566.pdf","grobid_xml":"https://content.openalex.org/works/W7006461566.grobid-xml"},"referenced_works_count":78,"referenced_works":["https://openalex.org/W1840435438","https://openalex.org/W1973196098","https://openalex.org/W2160637580","https://openalex.org/W2170341113","https://openalex.org/W2597603852","https://openalex.org/W2786672974","https://openalex.org/W2794325560","https://openalex.org/W2950220847","https://openalex.org/W2950700477","https://openalex.org/W2951365061","https://openalex.org/W2953937638","https://openalex.org/W3030163527","https://openalex.org/W3103621845","https://openalex.org/W3126216366","https://openalex.org/W3177813494","https://openalex.org/W3198002980","https://openalex.org/W3199246459","https://openalex.org/W3201174429","https://openalex.org/W3204096139","https://openalex.org/W3210904070","https://openalex.org/W4221142221","https://openalex.org/W4221159672","https://openalex.org/W4223908421","https://openalex.org/W4281690148","https://openalex.org/W4310428868","https://openalex.org/W4318908031","https://openalex.org/W4320003957","https://openalex.org/W4327810158","https://openalex.org/W4327811957","https://openalex.org/W4360836968","https://openalex.org/W4362707064","https://openalex.org/W4366341216","https://openalex.org/W4367000491","https://openalex.org/W4376167329","https://openalex.org/W4377121468","https://openalex.org/W4378499145","https://openalex.org/W4379539933","https://openalex.org/W4380136478","https://openalex.org/W4381586770","https://openalex.org/W4381714872","https://openalex.org/W4384662964","https://openalex.org/W4384918448","https://openalex.org/W4385571232","https://openalex.org/W4385571466","https://openalex.org/W4385572634","https://openalex.org/W4385682136","https://openalex.org/W4385849424","https://openalex.org/W4386908094","https://openalex.org/W4386977707","https://openalex.org/W4387322976","https://openalex.org/W4387355767","https://openalex.org/W4387985331","https://openalex.org/W4387994765","https://openalex.org/W4388555916","https://openalex.org/W4388718089","https://openalex.org/W4388890574","https://openalex.org/W4388926373","https://openalex.org/W4389470007","https://openalex.org/W4389518953","https://openalex.org/W4389518965","https://openalex.org/W4390041933","https://openalex.org/W4390723197","https://openalex.org/W4391136507","https://openalex.org/W4392616695","https://openalex.org/W4392617597","https://openalex.org/W4392679398","https://openalex.org/W4392822190","https://openalex.org/W4394708144","https://openalex.org/W4399836654","https://openalex.org/W4401042689","https://openalex.org/W4402671800","https://openalex.org/W4415795978","https://openalex.org/W7039287486","https://openalex.org/W7133189643","https://openalex.org/W7133220558","https://openalex.org/W7133221203","https://openalex.org/W7133224126","https://openalex.org/W7133233472"],"related_works":["https://openalex.org/W4410342726","https://openalex.org/W4410392463","https://openalex.org/W4391421129","https://openalex.org/W4411113256","https://openalex.org/W7161065936","https://openalex.org/W4404149281","https://openalex.org/W6966752457","https://openalex.org/W7104267478","https://openalex.org/W7161820043","https://openalex.org/W4393161118"],"abstract_inverted_index":{"The":[0],"rapid":[1],"advancement":[2],"of":[3,120,126,213],"Large":[4],"Language":[5],"Models":[6],"(LLMs),":[7],"driven":[8],"by":[9,55,76],"scaling":[10],"laws":[11],"and":[12,30,38,62,68,79,89,95,196,206,224],"substantial":[13],"computational":[14],"investments,":[15],"has":[16],"revolutionized":[17],"artificial":[18],"intelligence,":[19],"enabling":[20,197],"significant":[21],"breakthroughs":[22],"across":[23],"domains":[24],"such":[25],"as":[26,137],"conversational":[27,146],"AI,":[28],"programming,":[29],"complex":[31],"reasoning.":[32],"However,":[33],"traditional":[34],"benchmarks,":[35],"being":[36],"static":[37],"narrowly":[39],"scoped,":[40],"struggle":[41],"to":[42,87,160],"comprehensively":[43],"evaluate":[44],"LLM":[45],"capabilities":[46],"in":[47,145],"real-world":[48,102,199],"scenarios.":[49],"These":[50,201],"limitations":[51],"are":[52,217],"further":[53],"compounded":[54],"potential":[56],"contamination,":[57,92],"where":[58],"overlaps":[59],"between":[60],"training":[61],"test":[63],"data":[64],"inflate":[65],"performance":[66],"metrics":[67],"obscure":[69],"true":[70],"generalization.This":[71],"dissertation":[72],"tackles":[73],"these":[74],"challenges":[75],"introducing":[77],"robust":[78],"scalable":[80],"evaluation":[81,167,208],"methodologies.":[82],"First,":[83],"we":[84,104,150,177],"develop":[85],"methods":[86],"detect":[88],"mitigate":[90],"benchmark":[91],"ensuring":[93],"reliable":[94],"trustworthy":[96],"assessments.":[97],"To":[98],"align":[99],"evaluations":[100],"with":[101,173,226],"applications,":[103],"present":[105],"Chatbot":[106,128],"Arena,":[107],"a":[108],"live,":[109],"crowdsourced":[110,191],"platform":[111,140],"that":[112,156,216],"gathers":[113],"real-time":[114],"human":[115,143,148,162,174,227],"feedback":[116],"via":[117],"pairwise":[118],"comparisons":[119],"model":[121,134],"responses.":[122],"By":[123],"analyzing":[124],"millions":[125],"interactions,":[127],"Arena":[129],"offers":[130],"dynamic":[131],"insights":[132],"into":[133],"performance,":[135],"emerging":[136],"an":[138,153,182],"industry-standard":[139],"for":[141,185],"evaluating":[142],"preference":[144],"AI.Complementing":[147],"evaluation,":[149],"introduce":[151],"LLM-as-a-Judge,":[152],"automated":[154,183],"framework":[155],"leverages":[157],"advanced":[158],"LLMs":[159],"approximate":[161],"judgments,":[163],"significantly":[164],"reducing":[165],"manual":[166],"costs":[168],"while":[169],"achieving":[170],"strong":[171],"agreement":[172],"preferences.":[175],"Furthermore,":[176],"propose":[178],"the":[179,211],"BenchBuilder":[180],"pipeline,":[181],"system":[184],"creating":[186],"diverse,":[187],"high-quality":[188],"benchmarks":[189],"from":[190],"prompts,":[192],"addressing":[193],"contamination":[194],"issues":[195],"robust,":[198],"evaluations.":[200],"contributions":[202],"establish":[203],"scalable,":[204],"dynamic,":[205],"human-centric":[207],"systems,":[209],"advancing":[210],"development":[212],"AI":[214],"systems":[215],"not":[218],"only":[219],"powerful":[220],"but":[221],"also":[222],"reliable,":[223],"aligned":[225],"values.":[228]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}