{"id":"https://openalex.org/W7140221701","doi":"https://doi.org/10.48550/arxiv.2603.20882","title":"RubricRAG: Towards Interpretable and Reliable LLM Evaluation via Domain Knowledge Retrieval for Rubric Generation","display_name":"RubricRAG: Towards Interpretable and Reliable LLM Evaluation via Domain Knowledge Retrieval for Rubric Generation","publication_year":2026,"publication_date":"2026-03-21","ids":{"openalex":"https://openalex.org/W7140221701","doi":"https://doi.org/10.48550/arxiv.2603.20882"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.20882","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20882","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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.20882","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Dhole, Kaustubh D.","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dhole, Kaustubh D.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Agichtein, Eugene","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Agichtein, Eugene","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":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/T10028","display_name":"Topic Modeling","score":0.28790000081062317,"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":0.28790000081062317,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.10909999907016754,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.09130000323057175,"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/rubric","display_name":"Rubric","score":0.9868000149726868},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.6757000088691711},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5322999954223633},{"id":"https://openalex.org/keywords/domain-knowledge","display_name":"Domain knowledge","score":0.4569999873638153},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.436599999666214},{"id":"https://openalex.org/keywords/quality","display_name":"Quality (philosophy)","score":0.4336000084877014}],"concepts":[{"id":"https://openalex.org/C111640148","wikidata":"https://www.wikidata.org/wiki/Q847349","display_name":"Rubric","level":2,"score":0.9868000149726868},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7017999887466431},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.6757000088691711},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5322999954223633},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46619999408721924},{"id":"https://openalex.org/C207685749","wikidata":"https://www.wikidata.org/wiki/Q2088941","display_name":"Domain knowledge","level":2,"score":0.4569999873638153},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.436599999666214},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.4336000084877014},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4117000102996826},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3785000145435333},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.3319999873638153},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.3215999901294708},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2727999985218048},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.26750001311302185},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.2605000138282776},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.25380000472068787}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.20882","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20882","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":"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":"doi:10.48550/arxiv.2603.20882","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.20882","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":"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"},"sustainable_development_goals":[{"display_name":"Decent work and economic growth","score":0.6147629618644714,"id":"https://metadata.un.org/sdg/8"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"are":[4,26,117,173],"increasingly":[5],"evaluated":[6],"and":[7,64,90,93,120,158,162,212,224],"sometimes":[8],"trained":[9],"using":[10],"automated":[11,108,233],"graders":[12],"such":[13],"as":[14,137],"LLM-as-judges":[15],"that":[16,167,172,199],"output":[17],"scalar":[18],"scores":[19],"or":[20,39,45],"preferences.":[21],"While":[22,98],"convenient,":[23],"these":[24,115],"approaches":[25,100],"often":[27],"opaque:":[28],"a":[29,47,71,181,225],"single":[30],"score":[31],"rarely":[32],"explains":[33],"why":[34],"an":[35],"answer":[36],"is":[37,88,112],"good":[38,148],"bad,":[40],"which":[41,185],"requirements":[42],"were":[43],"missed,":[44],"how":[46],"system":[48],"should":[49],"be":[50],"improved.":[51],"This":[52],"lack":[53],"of":[54,228],"interpretability":[55],"limits":[56],"their":[57],"usefulness":[58],"for":[59,96,107,122,146,207,213],"model":[60],"development,":[61],"dataset":[62],"curation,":[63],"high-stakes":[65],"deployment.":[66,97],"Query-specific":[67],"rubric-based":[68],"evaluation":[69,216,231],"offers":[70],"more":[72,203],"transparent":[73],"alternative":[74],"by":[75],"decomposing":[76],"quality":[77],"into":[78],"explicit,":[79],"checkable":[80],"criteria.":[81],"However,":[82],"manually":[83],"designing":[84],"high-quality,":[85],"query-specific":[86],"rubrics":[87,106,116,136,171,190,205],"labor-intensive":[89],"cognitively":[91],"demanding":[92],"not":[94],"feasible":[95],"previous":[99],"have":[101],"focused":[102],"on":[103,154,159],"generating":[104],"intermediate":[105],"downstream":[109,215],"evaluation,":[110],"it":[111],"unclear":[113],"if":[114],"both":[118,206,221],"interpretable":[119,204,230],"effective":[121],"human":[123],"users.":[124],"In":[125],"this":[126],"work,":[127],"we":[128,165],"investigate":[129],"whether":[130],"LLMs":[131,169],"can":[132,201],"generate":[133,202],"useful,":[134],"instance-specific":[135],"compared":[138],"to":[139,209],"human-authored":[140,177,210],"rubrics,":[141,211],"while":[142],"also":[143],"improving":[144],"effectiveness":[145],"identifying":[147],"responses.":[149],"Through":[150],"our":[151],"systematic":[152],"study":[153],"two":[155],"rubric":[156,234],"benchmarks,":[157],"multiple":[160],"few-shot":[161],"post-training":[163],"strategies,":[164],"find":[166],"off-the-shelf":[168],"produce":[170],"poorly":[174],"aligned":[175],"with":[176],"ones.":[178],"We":[179,197],"introduce":[180],"simple":[182],"strategy,":[183],"RubricRAG,":[184],"retrieves":[186],"domain":[187],"knowledge":[188],"via":[189],"at":[191],"inference":[192],"time":[193],"from":[194],"related":[195],"queries.":[196],"demonstrate":[198],"RubricRAG":[200],"similarity":[208],"improved":[214],"effectiveness.":[217],"Our":[218],"results":[219],"highlight":[220],"the":[222],"challenges":[223],"promising":[226],"approach":[227],"scalable,":[229],"through":[232],"generation.":[235]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-25T00:00:00"}