{"id":"https://openalex.org/W4302013344","doi":"https://doi.org/10.48550/arxiv.2210.01117","title":"Omnigrok: Grokking Beyond Algorithmic Data","display_name":"Omnigrok: Grokking Beyond Algorithmic Data","publication_year":2022,"publication_date":"2022-10-03","ids":{"openalex":"https://openalex.org/W4302013344","doi":"https://doi.org/10.48550/arxiv.2210.01117"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2210.01117","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2210.01117","pdf_url":"https://arxiv.org/pdf/2210.01117","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":null,"license_id":null,"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/2210.01117","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101924541","display_name":"Ziming Liu","orcid":"https://orcid.org/0000-0003-3581-7494","observed_orcids":["https://orcid.org/0000-0003-3581-7494","https://orcid.org/0000-0002-7090-8853"]},"institutions":[{"id":"https://openalex.org/I4210160315","display_name":"The NSF AI Institute for Artificial Intelligence and Fundamental Interactions","ror":"https://ror.org/04pvzz946","country_code":"US","type":"facility","lineage":["https://openalex.org/I121934306","https://openalex.org/I12912129","https://openalex.org/I136199984","https://openalex.org/I4210160315","https://openalex.org/I63966007"]},{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Ziming","raw_affiliation_strings":["Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology","institution_ids":["https://openalex.org/I4210160315","https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071048251","display_name":"Eric J. Michaud","orcid":"https://orcid.org/0000-0001-7912-1953","observed_orcids":["https://orcid.org/0000-0001-7912-1953"]},"institutions":[{"id":"https://openalex.org/I4210160315","display_name":"The NSF AI Institute for Artificial Intelligence and Fundamental Interactions","ror":"https://ror.org/04pvzz946","country_code":"US","type":"facility","lineage":["https://openalex.org/I121934306","https://openalex.org/I12912129","https://openalex.org/I136199984","https://openalex.org/I4210160315","https://openalex.org/I63966007"]},{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Michaud, Eric J.","raw_affiliation_strings":["Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology","institution_ids":["https://openalex.org/I4210160315","https://openalex.org/I63966007"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5091601455","display_name":"Max Tegmark","orcid":"https://orcid.org/0000-0001-7670-7190","observed_orcids":["https://orcid.org/0000-0001-7670-7190"]},"institutions":[{"id":"https://openalex.org/I4210160315","display_name":"The NSF AI Institute for Artificial Intelligence and Fundamental Interactions","ror":"https://ror.org/04pvzz946","country_code":"US","type":"facility","lineage":["https://openalex.org/I121934306","https://openalex.org/I12912129","https://openalex.org/I136199984","https://openalex.org/I4210160315","https://openalex.org/I63966007"]},{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tegmark, Max","raw_affiliation_strings":["Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Physics , Institute for AI and Fundamental Interactions Massachusetts Institute of Technology","institution_ids":["https://openalex.org/I4210160315","https://openalex.org/I63966007"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":8,"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":0.48829999566078186,"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":0.48829999566078186,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.25780001282691956,"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/T10320","display_name":"Neural Networks and Applications","score":0.14650000631809235,"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/grokking","display_name":"grokking","score":0.9950000047683716},{"id":"https://openalex.org/keywords/loss-landscape","display_name":"loss landscape","score":0.8399999737739563},{"id":"https://openalex.org/keywords/weight-norm","display_name":"weight norm","score":0.8059999942779541},{"id":"https://openalex.org/keywords/weight-decay","display_name":"weight decay","score":0.6340000033378601},{"id":"https://openalex.org/keywords/representation-learning","display_name":"representation learning","score":0.6759999990463257},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.5889999866485596},{"id":"https://openalex.org/keywords/overfitting","display_name":"overfitting","score":0.7440000176429749}],"concepts":[{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.9623633623123169},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7095657587051392},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6904700398445129},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.6281209588050842},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.5720227360725403},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.532774806022644},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.5033926367759705},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4785069525241852},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4738274812698364},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.45000672340393066},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.4109785854816437},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.4012221693992615},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3979566991329193},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.20658046007156372},{"id":"https://openalex.org/C111472728","wikidata":"https://www.wikidata.org/wiki/Q9471","display_name":"Epistemology","level":1,"score":0.12329968810081482},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2210.01117","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2210.01117","pdf_url":"https://arxiv.org/pdf/2210.01117","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2210.01117","doi":"https://doi.org/10.48550/arxiv.2210.01117","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2210.01117","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:2210.01117","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2210.01117","pdf_url":"https://arxiv.org/pdf/2210.01117","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G2366492356","display_name":null,"funder_award_id":"PHY-2\u2009019\u2009786","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G3509299210","display_name":"AI Institute: AI Research Institute for Fundamental Interactions","funder_award_id":"2019786","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G539115378","display_name":"Graduate Research Fellowship Program (GRFP)","funder_award_id":"2141064","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4387995264","https://openalex.org/W4226434736","https://openalex.org/W4387560732","https://openalex.org/W4382141734","https://openalex.org/W7133037975","https://openalex.org/W4388684804","https://openalex.org/W4407184859","https://openalex.org/W7162512523","https://openalex.org/W7216013745","https://openalex.org/W4399151187"],"abstract_inverted_index":{"Grokking,":[0],"the":[1,13,26,33,41,50,84,91,108,122],"unusual":[2],"phenomenon":[3],"for":[4,43,117,127],"algorithmic":[5,118,128],"datasets":[6,129],"where":[7],"generalization":[8],"happens":[9],"long":[10],"after":[11],"overfitting":[12],"training":[14,36,54],"data,":[15],"has":[16],"remained":[17],"elusive.":[18],"We":[19,45,120],"aim":[20],"to":[21,47,97,114,130],"understand":[22],"grokking":[23,99,116,126],"by":[24,90],"analyzing":[25],"loss":[27],"landscapes":[28],"of":[29,76,86,125],"neural":[30],"networks,":[31],"identifying":[32],"mismatch":[34],"between":[35],"and":[37,55,65,105],"test":[38,56],"losses":[39,57],"as":[40,49],"cause":[42],"grokking.":[44],"refer":[46],"this":[48],"\"LU":[51],"mechanism\"":[52],"because":[53],"(against":[58],"model":[59],"weight":[60,81],"norm)":[61],"typically":[62],"resemble":[63],"\"L\"":[64],"\"U\",":[66],"respectively.":[67],"This":[68],"simple":[69],"mechanism":[70],"can":[71],"nicely":[72],"explain":[73],"many":[74],"aspects":[75],"grokking:":[77],"data":[78],"size":[79],"dependence,":[80,83],"decay":[82],"emergence":[85],"representations,":[87],"etc.":[88],"Guided":[89],"intuitive":[92],"picture,":[93],"we":[94,111],"are":[95,112],"able":[96,113],"induce":[98],"on":[100],"tasks":[101],"involving":[102],"images,":[103],"language":[104],"molecules.":[106],"In":[107],"reverse":[109],"direction,":[110],"eliminate":[115],"datasets.":[119],"attribute":[121],"dramatic":[123],"nature":[124],"representation":[131],"learning.":[132]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":3}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2022-10-06T00:00:00"}