{"id":"https://openalex.org/W7140144503","doi":"https://doi.org/10.48550/arxiv.2603.19251","title":"Enhancing Legal LLMs through Metadata-Enriched RAG Pipelines and Direct Preference Optimization","display_name":"Enhancing Legal LLMs through Metadata-Enriched RAG Pipelines and Direct Preference Optimization","publication_year":2026,"publication_date":"2026-02-25","ids":{"openalex":"https://openalex.org/W7140144503","doi":"https://doi.org/10.48550/arxiv.2603.19251"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2603.19251","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2603.19251","pdf_url":"https://arxiv.org/pdf/2603.19251","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2603.19251","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5119758308","display_name":"Suyash Maniyar","orcid":"https://orcid.org/0009-0000-5882-4377","observed_orcids":["https://orcid.org/0009-0000-5882-4377"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Maniyar, Suyash","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130402082","display_name":"Deepali Singh","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Singh, Deepali","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5130365720","display_name":"Rohith Reddy","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Reddy, Rohith","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":false,"primary_topic":{"id":"https://openalex.org/T13643","display_name":"Artificial Intelligence in Law","score":0.6714000105857849,"subfield":{"id":"https://openalex.org/subfields/3308","display_name":"Law"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13643","display_name":"Artificial Intelligence in Law","score":0.6714000105857849,"subfield":{"id":"https://openalex.org/subfields/3308","display_name":"Law"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.3041999936103821,"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":0.005900000222027302,"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/retrieval-augmented-generation","display_name":"retrieval-augmented generation","score":0.9869999885559082},{"id":"https://openalex.org/keywords/metadata-enrichment","display_name":"metadata enrichment","score":0.578000009059906},{"id":"https://openalex.org/keywords/direct-preference-optimization","display_name":"direct preference optimization","score":0.7260000109672546},{"id":"https://openalex.org/keywords/hallucination-mitigation","display_name":"hallucination mitigation","score":0.8629999756813049},{"id":"https://openalex.org/keywords/legal-document-retrieval","display_name":"legal document retrieval","score":0.7170000076293945},{"id":"https://openalex.org/keywords/lexical-redundancy","display_name":"lexical redundancy","score":0.6809999942779541},{"id":"https://openalex.org/keywords/local-llm-deployment","display_name":"local LLM deployment","score":0.5540000200271606},{"id":"https://openalex.org/keywords/data-privacy","display_name":"data privacy","score":0.7670000195503235}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6403999924659729},{"id":"https://openalex.org/C93518851","wikidata":"https://www.wikidata.org/wiki/Q180160","display_name":"Metadata","level":2,"score":0.6251999735832214},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.6205000281333923},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5564000010490417},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5085999965667725},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.4846999943256378},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.4058000147342682},{"id":"https://openalex.org/C2993995455","wikidata":"https://www.wikidata.org/wiki/Q3150005","display_name":"Legal document","level":2,"score":0.39320001006126404},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3402999937534332},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.325300008058548},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3075000047683716},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.30489999055862427},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.296999990940094},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2921000123023987},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2906999886035919},{"id":"https://openalex.org/C200635333","wikidata":"https://www.wikidata.org/wiki/Q158970","display_name":"Tort","level":3,"score":0.28940001130104065},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2827000021934509},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.27630001306533813},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.26109999418258667},{"id":"https://openalex.org/C2778137410","wikidata":"https://www.wikidata.org/wiki/Q2732820","display_name":"Government (linguistics)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.25110000371932983},{"id":"https://openalex.org/C171686336","wikidata":"https://www.wikidata.org/wiki/Q3532085","display_name":"Topic model","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2603.19251","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2603.19251","pdf_url":"https://arxiv.org/pdf/2603.19251","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2603.19251","doi":"https://doi.org/10.48550/arxiv.2603.19251","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.19251","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":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2603.19251","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2603.19251","pdf_url":"https://arxiv.org/pdf/2603.19251","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","score":0.666700005531311,"display_name":"Peace, justice, and strong institutions"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7140144503.pdf","grobid_xml":"https://content.openalex.org/works/W7140144503.grobid-xml"},"referenced_works_count":4,"referenced_works":["https://openalex.org/W4378465439","https://openalex.org/W4416033570","https://openalex.org/W4417210016","https://openalex.org/W6910723435"],"related_works":["https://openalex.org/W7204709710","https://openalex.org/W7160586684","https://openalex.org/W7209954475","https://openalex.org/W4416038826","https://openalex.org/W4413782328","https://openalex.org/W4408323770","https://openalex.org/W4414528059","https://openalex.org/W4415317938","https://openalex.org/W7162502051","https://openalex.org/W7125501920"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"perform":[4],"well":[5],"in":[6,48,72,121],"short":[7],"contexts":[8],"but":[9,45],"degrade":[10],"on":[11],"long":[12],"legal":[13,26,49,73,122],"documents,":[14],"often":[15],"producing":[16],"hallucinations":[17],"such":[18,32],"as":[19],"incorrect":[20],"clauses":[21],"or":[22],"precedents.":[23],"In":[24],"the":[25],"domain,":[27],"where":[28,78],"precision":[29],"is":[30,111],"critical,":[31],"errors":[33,67,77],"undermine":[34],"reliability":[35],"and":[36,75,99,119],"trust.":[37],"Retrieval":[38],"Augmented":[39],"Generation":[40],"(RAG)":[41],"helps":[42],"ground":[43],"outputs":[44],"remains":[46],"limited":[47],"settings,":[50],"especially":[51],"with":[52],"small,":[53],"locally":[54],"deployed":[55],"models":[56,79],"required":[57],"for":[58],"data":[59],"privacy.":[60],"We":[61],"identify":[62],"two":[63],"failure":[64],"modes:":[65],"retrieval":[66],"due":[68],"to":[69,94,105],"lexical":[70],"redundancy":[71],"corpora,":[74],"decoding":[76],"generate":[80],"answers":[81],"despite":[82],"insufficient":[83],"context.":[84],"To":[85],"address":[86],"this,":[87],"we":[88],"propose":[89],"Metadata":[90],"Enriched":[91],"Hybrid":[92],"RAG":[93],"improve":[95,116],"document":[96],"level":[97],"retrieval,":[98],"apply":[100],"Direct":[101],"Preference":[102],"Optimization":[103],"(DPO)":[104],"enforce":[106],"safe":[107],"refusal":[108],"when":[109],"context":[110],"inadequate.":[112],"Together,":[113],"these":[114],"methods":[115],"grounding,":[117],"reliability,":[118],"safety":[120],"language":[123],"models.":[124]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-03-24T00:00:00"}