{"id":"https://openalex.org/W4416036012","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.1108","title":"Mitigating Catastrophic Forgetting in Large Language Models with Forgetting-aware Pruning","display_name":"Mitigating Catastrophic Forgetting in Large Language Models with Forgetting-aware Pruning","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416036012","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.1108"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.emnlp-main.1108","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.1108","pdf_url":"https://aclanthology.org/2025.emnlp-main.1108.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.emnlp-main.1108.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100352892","display_name":"Wei Huang","orcid":"https://orcid.org/0000-0003-1418-6267"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Huang","raw_affiliation_strings":["Ant Group , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group , China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5038610148","display_name":"Anda Cheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Anda Cheng","raw_affiliation_strings":["Ant Group , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group , China","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5028637336","display_name":"Yinggui Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yinggui Wang","raw_affiliation_strings":["Ant Group , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group , China","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.35349829,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"21853","last_page":"21867"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.35019999742507935,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.35019999742507935,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.1145000010728836,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.1005999967455864,"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/pruning","display_name":"Pruning","score":0.5378999710083008},{"id":"https://openalex.org/keywords/forgetting","display_name":"Forgetting","score":0.508899986743927},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4731999933719635},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.26460000872612},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.26269999146461487},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.2572000026702881}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6060000061988831},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.5378999710083008},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5339999794960022},{"id":"https://openalex.org/C7149132","wikidata":"https://www.wikidata.org/wiki/Q1377840","display_name":"Forgetting","level":2,"score":0.508899986743927},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4731999933719635},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3799999952316284},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3610999882221222},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.26269999146461487},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.2572000026702881},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25290000438690186},{"id":"https://openalex.org/C141218545","wikidata":"https://www.wikidata.org/wiki/Q7521336","display_name":"Simulation language","level":2,"score":0.2524000108242035},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2517000138759613},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.2506999969482422}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.emnlp-main.1108","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.1108","pdf_url":"https://aclanthology.org/2025.emnlp-main.1108.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.emnlp-main.1108","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.1108","pdf_url":"https://aclanthology.org/2025.emnlp-main.1108.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416036012.pdf","grobid_xml":"https://content.openalex.org/works/W4416036012.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"advancements":[1],"in":[2,11],"large":[3],"language":[4,131],"models":[5],"(LLMs)":[6],"have":[7],"shown":[8],"impressive":[9],"capabilities":[10],"various":[12],"downstream":[13,40,63,158],"tasks":[14],"but":[15],"typically":[16],"face":[17],"Catastrophic":[18],"Forgetting":[19],"(CF)":[20],"during":[21],"fine-tuning.In":[22],"this":[23,77,98],"paper,":[24],"we":[25],"propose":[26],"the":[27,46,53,59,81,84,101,110,161],"Forgetting-Aware":[28],"Pruning":[29],"Metric":[30],"(FAPM),":[31],"a":[32,71,92],"novel":[33],"pruning-based":[34],"approach":[35],"to":[36,48,87,94,109,150,163],"balance":[37],"CF":[38,149],"and":[39,141],"task":[41,50,85],"performance.Our":[42],"investigation":[43],"reveals":[44],"that":[45,146],"degree":[47],"which":[49],"vectors":[51],"(i.e.,":[52],"subtraction":[54],"of":[55,83],"pre-trained":[56,67,88],"weights":[57,60],"from":[58],"fine-tuned":[61],"on":[62,76,157],"tasks)":[64],"overlap":[65],"with":[66],"model":[68,89,114],"parameters":[69,90],"is":[70],"critical":[72],"factor":[73],"for":[74],"CF.Based":[75],"finding,":[78],"FAPM":[79,104,147],"employs":[80],"ratio":[82],"vector":[86],"as":[91],"metric":[93],"quantify":[95],"CF,":[96],"integrating":[97],"measure":[99],"into":[100],"pruning":[102],"criteria.Importantly,":[103],"does":[105,117],"not":[106],"necessitate":[107],"modifications":[108],"training":[111],"process":[112],"or":[113],"architecture,":[115],"nor":[116],"it":[118],"require":[119],"any":[120],"auxiliary":[121],"data.We":[122],"conducted":[123],"extensive":[124],"experiments":[125],"across":[126],"eight":[127],"datasets,":[128],"covering":[129],"natural":[130],"inference,":[132],"General":[133],"Q&A,":[134,136,138],"Medical":[135],"Math":[137],"reading":[139],"comprehension,":[140],"cloze":[142],"tests.The":[143],"results":[144],"demonstrate":[145],"limits":[148],"just":[151],"0.25%":[152],"while":[153],"maintaining":[154],"99.67%":[155],"accuracy":[156],"tasks.We":[159],"provide":[160],"code":[162],"reproduce":[164],"our":[165],"results.1":[166],".":[167]},"counts_by_year":[],"updated_date":"2026-08-21T09:56:20.448147","created_date":"2025-11-08T00:00:00"}