{"id":"https://openalex.org/W7214948618","doi":"https://doi.org/10.48550/arxiv.2609.36230","title":"The Role of Feed-Forward Layers in Transformer Dynamics","display_name":"The Role of Feed-Forward Layers in Transformer Dynamics","publication_year":2026,"publication_date":"2026-09-28","ids":{"openalex":"https://openalex.org/W7214948618","doi":"https://doi.org/10.48550/arxiv.2609.36230"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.36230","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36230","pdf_url":"https://arxiv.org/pdf/2609.36230","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.36230","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5056734483","display_name":"Thomas Jacob Maranzatto","orcid":"https://orcid.org/0000-0002-6105-2758","observed_orcids":["https://orcid.org/0000-0002-6105-2758"]},"institutions":[{"id":"https://openalex.org/I66946132","display_name":"University of Maryland, College Park","ror":"https://ror.org/047s2c258","country_code":"US","type":"education","lineage":["https://openalex.org/I66946132"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Maranzatto, Thomas Jacob","raw_affiliation_strings":["Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA","institution_ids":["https://openalex.org/I66946132"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114535869","display_name":"Semih Akko\u00e7","orcid":"https://orcid.org/0009-0007-0743-7220","observed_orcids":["https://orcid.org/0009-0007-0743-7220"]},"institutions":[{"id":"https://openalex.org/I66946132","display_name":"University of Maryland, College Park","ror":"https://ror.org/047s2c258","country_code":"US","type":"education","lineage":["https://openalex.org/I66946132"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Akkoc, Semih","raw_affiliation_strings":["Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA","institution_ids":["https://openalex.org/I66946132"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5153603548","display_name":"Sennur Ulukus","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I66946132","display_name":"University of Maryland, College Park","ror":"https://ror.org/047s2c258","country_code":"US","type":"education","lineage":["https://openalex.org/I66946132"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ulukus, Sennur","raw_affiliation_strings":["Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering University of Maryland College Park , MD 20742 , USA","institution_ids":["https://openalex.org/I66946132"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I66946132"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10028","display_name":"Topic Modeling","score":0.28540000319480896,"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":0.28540000319480896,"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":0.2847000062465668,"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/T12592","display_name":"Opinion Dynamics and Social Influence","score":0.10409999638795853,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/self-attention","display_name":"self-attention","score":0.9459999799728394},{"id":"https://openalex.org/keywords/interacting-particle-systems","display_name":"interacting particle systems","score":0.9789999723434448},{"id":"https://openalex.org/keywords/consensus-dynamics","display_name":"consensus dynamics","score":0.6309999823570251},{"id":"https://openalex.org/keywords/multi-head-attention","display_name":"multi-head attention","score":0.8019999861717224},{"id":"https://openalex.org/keywords/control-theoretic-analysis","display_name":"control-theoretic analysis","score":0.628000020980835},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.6840000152587891}],"concepts":[{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.5590000152587891},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.48989999294281006},{"id":"https://openalex.org/C191178318","wikidata":"https://www.wikidata.org/wiki/Q2256906","display_name":"Thresholding","level":3,"score":0.421999990940094},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.40720000863075256},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.40230000019073486},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.3928999900817871},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.38109999895095825},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.365200012922287},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.34060001373291016},{"id":"https://openalex.org/C500300565","wikidata":"https://www.wikidata.org/wiki/Q925667","display_name":"Computer simulation","level":2,"score":0.319599986076355},{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.3041999936103821},{"id":"https://openalex.org/C145912823","wikidata":"https://www.wikidata.org/wiki/Q113558","display_name":"Dynamics (music)","level":2,"score":0.28850001096725464},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.27090001106262207},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.2669000029563904},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.2628999948501587},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.25940001010894775},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.2583000063896179},{"id":"https://openalex.org/C79379906","wikidata":"https://www.wikidata.org/wiki/Q3174497","display_name":"Dynamical systems theory","level":2,"score":0.25459998846054077},{"id":"https://openalex.org/C57879066","wikidata":"https://www.wikidata.org/wiki/Q41217","display_name":"Mechanics","level":1,"score":0.25450000166893005}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.36230","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36230","pdf_url":"https://arxiv.org/pdf/2609.36230","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.2609.36230","doi":"https://doi.org/10.48550/arxiv.2609.36230","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.36230","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:2609.36230","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.36230","pdf_url":"https://arxiv.org/pdf/2609.36230","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":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214948618.pdf","grobid_xml":"https://content.openalex.org/works/W7214948618.grobid-xml"},"referenced_works_count":18,"referenced_works":["https://openalex.org/W2137983211","https://openalex.org/W2333059519","https://openalex.org/W2963341956","https://openalex.org/W4390041933","https://openalex.org/W4399039217","https://openalex.org/W4401306886","https://openalex.org/W4401506292","https://openalex.org/W4403345493","https://openalex.org/W4403853555","https://openalex.org/W4407012314","https://openalex.org/W4411881822","https://openalex.org/W4414942720","https://openalex.org/W4415337592","https://openalex.org/W4416138639","https://openalex.org/W7106580379","https://openalex.org/W7126250235","https://openalex.org/W7160930630","https://openalex.org/W7170037525"],"related_works":["https://openalex.org/W4405057535","https://openalex.org/W4407385206","https://openalex.org/W4417024991","https://openalex.org/W7172419188","https://openalex.org/W7126046217","https://openalex.org/W4376123079","https://openalex.org/W4416331459","https://openalex.org/W6999270437","https://openalex.org/W4416968619","https://openalex.org/W7202213297"],"abstract_inverted_index":{"We":[0,81],"study":[1],"the":[2,16,21,25,35,48,53,61],"dynamical":[3],"behavior":[4,92],"of":[5,60],"tokens":[6,54],"in":[7],"transformers":[8],"from":[9,93],"a":[10],"control-theoretic":[11],"perspective.":[12],"Our":[13,42,67],"model":[14],"includes":[15],"feed-forward":[17,36,49],"layer":[18,37],"present":[19],"after":[20],"self-attention":[22,26],"mechanism,":[23],"with":[24],"mechanism":[27],"interpreted":[28],"as":[29,38],"an":[30,39],"interacting":[31],"particle":[32],"system":[33],"and":[34,64,77,89],"independent":[40],"control.":[41],"main":[43],"theoretical":[44],"result":[45,68],"establishes":[46],"that":[47],"network":[50],"can":[51],"steer":[52],"arbitrarily":[55],"close":[56],"to":[57,72,74,78,85,96],"consensus":[58],"regardless":[59],"key,":[62],"query,":[63],"value":[65],"matrices.":[66],"are":[69],"easily":[70],"extended":[71],"convergence":[73],"many":[75],"clusters":[76],"multi-head":[79],"attention.":[80],"conduct":[82],"numerical":[83],"experiments":[84],"verify":[86],"our":[87,94],"results,":[88],"compare":[90],"thresholding":[91],"theory":[95],"real-world":[97],"LLMs.":[98]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-01T00:00:00"}