{"id":"https://openalex.org/W4400518643","doi":"https://doi.org/10.48550/arxiv.2407.06322","title":"MagMax: Leveraging Model Merging for Seamless Continual Learning","display_name":"MagMax: Leveraging Model Merging for Seamless Continual Learning","publication_year":2024,"publication_date":"2024-07-08","ids":{"openalex":"https://openalex.org/W4400518643","doi":"https://doi.org/10.48550/arxiv.2407.06322"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2407.06322","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2407.06322","pdf_url":"https://arxiv.org/pdf/2407.06322","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","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2407.06322","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5078472796","display_name":"Daniel Marczak","orcid":"https://orcid.org/0000-0002-6352-9134","observed_orcids":["https://orcid.org/0000-0002-6352-9134"]},"institutions":[{"id":"https://openalex.org/I108403487","display_name":"Warsaw University of Technology","ror":"https://ror.org/00y0xnp53","country_code":"PL","type":"education","lineage":["https://openalex.org/I108403487"]}],"countries":["PL"],"is_corresponding":false,"raw_author_name":"Marczak, Daniel","raw_affiliation_strings":["Warsaw University of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Warsaw University of Technology","institution_ids":["https://openalex.org/I108403487"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017330235","display_name":"Bart\u0142omiej Twardowski","orcid":"https://orcid.org/0000-0003-2117-8679","observed_orcids":["https://orcid.org/0000-0003-2117-8679"]},"institutions":[{"id":"https://openalex.org/I123044942","display_name":"Universitat Aut\u00f2noma de Barcelona","ror":"https://ror.org/052g8jq94","country_code":"ES","type":"education","lineage":["https://openalex.org/I123044942"]},{"id":"https://openalex.org/I4387153096","display_name":"Computer Vision Center","ror":"https://ror.org/00s0nnj93","country_code":null,"type":"other","lineage":["https://openalex.org/I4387153040","https://openalex.org/I4387153096"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Twardowski, Bart\u0142omiej","raw_affiliation_strings":["Autonomous University of Barcelona","Computer Vision Center"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Autonomous University of Barcelona","institution_ids":["https://openalex.org/I123044942"]},{"raw_affiliation_string":"Computer Vision Center","institution_ids":["https://openalex.org/I4387153096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070627781","display_name":"T. P. Trzcinski","orcid":"https://orcid.org/0000-0002-1486-8906","observed_orcids":["https://orcid.org/0000-0002-1486-8906"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Trzci\u0144ski, Tomasz","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5069232913","display_name":"Sebastian Cygert","orcid":"https://orcid.org/0000-0002-4763-8381","observed_orcids":["https://orcid.org/0000-0002-4763-8381"]},"institutions":[{"id":"https://openalex.org/I169333911","display_name":"Gda\u0144sk University of Technology","ror":"https://ror.org/006x4sc24","country_code":"PL","type":"education","lineage":["https://openalex.org/I169333911"]}],"countries":["PL"],"is_corresponding":false,"raw_author_name":"Cygert, Sebastian","raw_affiliation_strings":["Gda\u0144sk University of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Gda\u0144sk University of Technology","institution_ids":["https://openalex.org/I169333911"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"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":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":0.9994999766349792,"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":0.9994999766349792,"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/T10036","display_name":"Advanced Neural Network Applications","score":9.999999747378752e-05,"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":0.0,"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/continual-learning","display_name":"continual learning","score":0.9950000047683716},{"id":"https://openalex.org/keywords/model-merging","display_name":"model merging","score":0.8870000243186951},{"id":"https://openalex.org/keywords/large-pre-trained-models","display_name":"large pre-trained models","score":0.8450000286102295},{"id":"https://openalex.org/keywords/catastrophic-forgetting","display_name":"catastrophic forgetting","score":0.6200000047683716},{"id":"https://openalex.org/keywords/weight-averaging","display_name":"weight averaging","score":0.7620000243186951},{"id":"https://openalex.org/keywords/sequential-fine-tuning","display_name":"sequential fine-tuning","score":0.8669999837875366},{"id":"https://openalex.org/keywords/class-incremental-learning","display_name":"class-incremental learning","score":0.9350000023841858},{"id":"https://openalex.org/keywords/domain-incremental-learning","display_name":"domain-incremental learning","score":0.9700000286102295},{"id":"https://openalex.org/keywords/knowledge-integration","display_name":"knowledge integration","score":0.8960000276565552}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6979033350944519},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3879278600215912}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2407.06322","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2407.06322","pdf_url":"https://arxiv.org/pdf/2407.06322","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.2407.06322","doi":"https://doi.org/10.48550/arxiv.2407.06322","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2407.06322","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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2407.06322","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2407.06322","pdf_url":"https://arxiv.org/pdf/2407.06322","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":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1349584921","display_name":null,"funder_award_id":"2022/45/B/ST6/02817","funder_id":"https://openalex.org/F4320322650","funder_display_name":"Narodowym Centrum Nauki"},{"id":"https://openalex.org/G3232057226","display_name":null,"funder_award_id":"101120237","funder_id":"https://openalex.org/F4320334322","funder_display_name":"HORIZON EUROPE Framework Programme"},{"id":"https://openalex.org/G8356777922","display_name":null,"funder_award_id":"PLG/2023/016393","funder_id":"https://openalex.org/F4320326815","funder_display_name":"Infrastruktura PL-Grid"}],"funders":[{"id":"https://openalex.org/F4320322650","display_name":"Narodowym Centrum Nauki","ror":"https://ror.org/03ha2q922"},{"id":"https://openalex.org/F4320326815","display_name":"Infrastruktura PL-Grid","ror":"https://ror.org/01m3qaz74"},{"id":"https://openalex.org/F4320334322","display_name":"HORIZON EUROPE Framework Programme","ror":null},{"id":"https://openalex.org/F4320337160","display_name":"Academic Computer Centre Cyfronet, AGH University of Science and Technology","ror":"https://ror.org/0575jwx13"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4400518643.pdf","grobid_xml":"https://content.openalex.org/works/W4400518643.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W2382290278","https://openalex.org/W4395014643"],"abstract_inverted_index":{"This":[0],"paper":[1],"introduces":[2],"a":[3,48,95],"continual":[4,32,87,101],"learning":[5,33,88,102,125],"approach":[6],"named":[7],"MagMax,":[8,94],"which":[9],"utilizes":[10],"model":[11,67],"merging":[12,68],"to":[13,18,37],"enable":[14],"large":[15,104],"pre-trained":[16,105],"models":[17,106],"continuously":[19],"learn":[20],"from":[21,30],"new":[22],"data":[23],"without":[24],"forgetting":[25,39],"previously":[26],"acquired":[27],"knowledge.":[28],"Distinct":[29],"traditional":[31],"methods":[34],"that":[35,71,99],"aim":[36],"reduce":[38],"during":[40],"task":[41],"training,":[42],"MagMax":[43,117],"combines":[44],"sequential":[45],"fine-tuning":[46],"with":[47],"maximum":[49],"magnitude":[50],"weight":[51,75,79],"selection":[52,80],"for":[53,107],"effective":[54],"knowledge":[55],"integration":[56],"across":[57],"tasks.":[58,109],"Our":[59,110],"initial":[60],"contribution":[61],"is":[62,129],"an":[63],"extensive":[64],"examination":[65],"of":[66,103,116],"techniques,":[69],"revealing":[70],"simple":[72],"approaches":[73],"like":[74],"averaging":[76],"and":[77,123],"random":[78],"surprisingly":[81],"hold":[82],"up":[83],"well":[84],"in":[85,118],"various":[86,119],"contexts.":[89],"More":[90],"importantly,":[91],"we":[92],"present":[93],"novel":[96],"model-merging":[97],"strategy":[98],"enables":[100],"successive":[108],"thorough":[111],"evaluation":[112],"demonstrates":[113],"the":[114],"superiority":[115],"scenarios,":[120],"including":[121],"class-":[122],"domain-incremental":[124],"settings.":[126],"The":[127],"code":[128],"available":[130],"at":[131],"this":[132],"URL:":[133],"https://github.com/danielm1405/magmax.":[134]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-09-30T09:27:27.532063","created_date":"2025-10-10T00:00:00"}