{"id":"https://openalex.org/W4412944948","doi":"https://doi.org/10.18653/v1/2025.acl-long.1047","title":"Neural Incompatibility: The Unbridgeable Gap of Cross-Scale Parametric Knowledge Transfer in Large Language Models","display_name":"Neural Incompatibility: The Unbridgeable Gap of Cross-Scale Parametric Knowledge Transfer in Large Language Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412944948","doi":"https://doi.org/10.18653/v1/2025.acl-long.1047"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.1047","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1047","pdf_url":"https://aclanthology.org/2025.acl-long.1047.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.1047.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053871882","display_name":"Yuqiao Tan","orcid":"https://orcid.org/0009-0009-0693-5230"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuqiao Tan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110645828","display_name":"Shizhu He","orcid":"https://orcid.org/0009-0009-2565-1508"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shizhu He","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100389893","display_name":"K. Liu","orcid":"https://orcid.org/0000-0002-2639-0698"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kang Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100358009","display_name":"Jun Zhao","orcid":"https://orcid.org/0000-0002-3573-152X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jun Zhao","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":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"21586","last_page":"21601"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.986299991607666,"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":0.986299991607666,"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":0.9609000086784363,"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/computer-science","display_name":"Computer science","score":0.7185531258583069},{"id":"https://openalex.org/keywords/parametric-statistics","display_name":"Parametric statistics","score":0.5179706811904907},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.4865192770957947},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.38128840923309326},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3790382742881775},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.06715184450149536},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.04665812849998474}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7185531258583069},{"id":"https://openalex.org/C117251300","wikidata":"https://www.wikidata.org/wiki/Q1849855","display_name":"Parametric statistics","level":2,"score":0.5179706811904907},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.4865192770957947},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.38128840923309326},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3790382742881775},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.06715184450149536},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.04665812849998474},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.1047","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1047","pdf_url":"https://aclanthology.org/2025.acl-long.1047.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.1047","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.1047","pdf_url":"https://aclanthology.org/2025.acl-long.1047.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.44999998807907104}],"awards":[{"id":"https://openalex.org/G6346556542","display_name":null,"funder_award_id":"62376270","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412944948.pdf","grobid_xml":"https://content.openalex.org/works/W4412944948.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"offer":[4],"a":[5,22,113,121],"transparent":[6],"brain":[7],"with":[8],"accessible":[9],"parameters":[10,56,95],"that":[11,126,148],"encode":[12],"extensive":[13],"knowledge,":[14],"which":[15,92],"can":[16],"be":[17],"analyzed,":[18],"located":[19],"and":[20,37,60,99,119,151,169,195],"transferred.Consequently,":[21],"key":[23],"research":[24,62,201],"challenge":[25],"is":[26,73,206],"to":[27,77,105,181],"transcend":[28],"traditional":[29],"knowledge":[30,49,86],"transfer":[31,87],"paradigms":[32],"rooted":[33],"in":[34,70,155],"symbolic":[35],"language":[36],"achieve":[38,78],"genuine":[39],"Parametric":[40],"Knowledge":[41],"Transfer":[42],"(PKT).Significantly,":[43],"exploring":[44],"effective":[45,183],"methods":[46],"for":[47,96,103,108,199],"transferring":[48],"across":[50,133],"LLMs":[51,132,174,194],"of":[52,131,175,193],"different":[53],"scales":[54,134],"through":[55],"presents":[57],"an":[58],"intriguing":[59],"valuable":[61],"direction.In":[63],"this":[64],"paper,":[65],"we":[66,111,162],"first":[67],"demonstrate":[68,147],"Alignment":[69],"parametric":[71,129,170,191],"space":[72],"the":[74,83,128,167,190],"fundamental":[75,179],"prerequisite":[76],"successful":[79],"cross-scale":[80],"PKT.We":[81],"redefine":[82],"previously":[84],"explored":[85],"as":[88,166],"Post-Align":[89],"PKT":[90,116],"(PostPKT),":[91],"utilizes":[93],"extracted":[94],"LoRA":[97],"initialization":[98],"requires":[100],"subsequent":[101],"fine-tune":[102],"alignment.Hence,":[104],"reduce":[106],"cost":[107],"further":[109],"fine-tuning,":[110],"introduce":[112],"novel":[114],"Pre-Align":[115],"(PrePKT)":[117],"paradigm":[118],"propose":[120],"solution":[122],"called":[123],"LaTen":[124],"(Locate-Then-Align)":[125],"aligns":[127],"spaces":[130],"only":[135],"using":[136],"several":[137],"training":[138],"steps":[139],"without":[140],"following":[141],"training.Comprehensive":[142],"experiments":[143],"on":[144,202],"four":[145],"benchmarks":[146],"both":[149],"PostPKT":[150],"PrePKT":[152],"face":[153],"challenges":[154,180],"achieving":[156,182],"consistently":[157],"stable":[158],"transfer.Through":[159],"in-depth":[160],"analysis,":[161],"identify":[163],"Neural":[164],"Incompatibility":[165],"ethological":[168],"structural":[171],"differences":[172],"between":[173],"varying":[176],"scales,":[177],"presenting":[178],"PKT.These":[184],"findings":[185],"provide":[186],"fresh":[187],"insights":[188],"into":[189],"architectures":[192],"highlight":[196],"promising":[197],"directions":[198],"future":[200],"efficient":[203],"PKT.Our":[204],"code":[205],"available":[207],"at":[208],"https://github.com/":[209],"Trae1ounG/Neural_Incompatibility.":[210]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-17T09:13:05.818461","created_date":"2025-10-10T00:00:00"}