{"id":"https://openalex.org/W7162637510","doi":"https://doi.org/10.48550/arxiv.2605.27786","title":"Locality-Aware Redundancy Pruning for LLM Depth Compression","display_name":"Locality-Aware Redundancy Pruning for LLM Depth Compression","publication_year":2026,"publication_date":"2026-05-27","ids":{"openalex":"https://openalex.org/W7162637510","doi":"https://doi.org/10.48550/arxiv.2605.27786"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.27786","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.27786","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.2605.27786","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137226861","display_name":"Vincent-Daniel Yun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yun, Vincent-Daniel","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137235976","display_name":"Youngrae Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Youngrae","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089920562","display_name":"Woosang Lim","orcid":"https://orcid.org/0000-0002-6615-5801"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lim, Woosang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137277278","display_name":"YoungJin Heo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Heo, YoungJin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137201933","display_name":"Minkyu Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Minkyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137294862","display_name":"Sunwoo Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Sunwoo","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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.2313999980688095,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.2313999980688095,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.11469999700784683,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.07079999893903732,"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/redundancy","display_name":"Redundancy (engineering)","score":0.7829999923706055},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5604000091552734},{"id":"https://openalex.org/keywords/perplexity","display_name":"Perplexity","score":0.535099983215332},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.5223000049591064},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.48910000920295715},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.4456999897956848}],"concepts":[{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.7829999923706055},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6068000197410583},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5604000091552734},{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.535099983215332},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.5223000049591064},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.49380001425743103},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.48910000920295715},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.4456999897956848},{"id":"https://openalex.org/C2779808786","wikidata":"https://www.wikidata.org/wiki/Q6664603","display_name":"Locality","level":2,"score":0.3944000005722046},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3935999870300293},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.35850000381469727},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.33059999346733093},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.296099990606308},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2913999855518341},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2883000075817108},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.27619999647140503}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.27786","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.27786","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.2605.27786","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.27786","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"are":[3],"known":[4],"to":[5,105],"contain":[6],"representational":[7,99],"redundancy":[8,33,57],"across":[9,35,110],"network":[10],"depth,":[11],"making":[12],"depth":[13,46],"pruning":[14,24,47,103],"an":[15],"effective":[16],"approach":[17],"for":[18],"improving":[19],"inference":[20],"efficiency.":[21],"Existing":[22],"one-shot":[23,45],"methods":[25],"rely":[26],"on":[27,66],"local":[28],"layer":[29,94],"importance":[30],"or":[31,62],"fixed":[32],"assumptions":[34],"architectures.":[36],"We":[37,53],"propose":[38],"Locality-Aware":[39],"Redundancy":[40],"Pruning":[41],"(LoRP),":[42],"a":[43,87],"training-free":[44],"framework":[48],"guided":[49],"by":[50,98],"representation":[51],"locality.":[52],"show":[54,114],"that":[55],"inter-layer":[56,83],"can":[58],"be":[59],"either":[60],"localized":[61],"globally":[63],"distributed":[64],"depending":[65],"the":[67],"LLM":[68,112],"architecture.":[69],"To":[70],"characterize":[71],"this":[72],"phenomenon,":[73],"we":[74],"introduce":[75],"Representation":[76],"Locality":[77],"Score":[78],"(RLS),":[79],"derived":[80],"from":[81],"global":[82],"hidden-state":[84],"similarity.":[85],"Using":[86],"small":[88],"calibration":[89],"set,":[90],"LoRP":[91],"computes":[92],"pairwise":[93],"similarity,":[95,100],"clusters":[96],"layers":[97],"and":[101,119],"allocates":[102],"according":[104],"residual":[106],"intra-cluster":[107],"redundancy.":[108],"Experiments":[109],"diverse":[111],"families":[113],"improvements":[115],"in":[116],"both":[117],"perplexity":[118],"downstream":[120],"task":[121],"accuracy.":[122],"Official":[123],"github":[124],"repository:":[125],"https://github.com/daniel-eai/LoRP-Locality-Aware-Redundancy-Pruning/":[126]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-29T00:00:00"}