{"id":"https://openalex.org/W7163999470","doi":"https://doi.org/10.48550/arxiv.2606.08721","title":"A Geometric Measure of Linear Separability for Neural Representations","display_name":"A Geometric Measure of Linear Separability for Neural Representations","publication_year":2026,"publication_date":"2026-06-07","ids":{"openalex":"https://openalex.org/W7163999470","doi":"https://doi.org/10.48550/arxiv.2606.08721"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.08721","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.08721","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.08721","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138254441","display_name":"Yi Wei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei, Yi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138230589","display_name":"Xuan Qi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qi, Xuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138246417","display_name":"Furao Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Furao","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":false,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.5221999883651733,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.5221999883651733,"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":0.08330000191926956,"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":0.031599998474121094,"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/affine-transformation","display_name":"Affine transformation","score":0.8453999757766724},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.6985999941825867},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5228999853134155},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.47839999198913574},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.4316999912261963},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.4101000130176544},{"id":"https://openalex.org/keywords/linear-model","display_name":"Linear model","score":0.40549999475479126},{"id":"https://openalex.org/keywords/class","display_name":"Class (philosophy)","score":0.3928000032901764}],"concepts":[{"id":"https://openalex.org/C92757383","wikidata":"https://www.wikidata.org/wiki/Q382497","display_name":"Affine transformation","level":2,"score":0.8453999757766724},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.6985999941825867},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6146000027656555},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5228999853134155},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.47839999198913574},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4471000134944916},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.439300000667572},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4316999912261963},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.4101000130176544},{"id":"https://openalex.org/C163175372","wikidata":"https://www.wikidata.org/wiki/Q3339222","display_name":"Linear model","level":2,"score":0.40549999475479126},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.3928000032901764},{"id":"https://openalex.org/C162392398","wikidata":"https://www.wikidata.org/wiki/Q272404","display_name":"Finite set","level":2,"score":0.3856000006198883},{"id":"https://openalex.org/C158251709","wikidata":"https://www.wikidata.org/wiki/Q354025","display_name":"Intrusion","level":2,"score":0.38019999861717224},{"id":"https://openalex.org/C207221997","wikidata":"https://www.wikidata.org/wiki/Q938614","display_name":"Affine combination","level":3,"score":0.3560999929904938},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.321399986743927},{"id":"https://openalex.org/C18516315","wikidata":"https://www.wikidata.org/wiki/Q4688950","display_name":"Affine shape adaptation","level":4,"score":0.31459999084472656},{"id":"https://openalex.org/C112128483","wikidata":"https://www.wikidata.org/wiki/Q382504","display_name":"Affine hull","level":4,"score":0.3133000135421753},{"id":"https://openalex.org/C49766605","wikidata":"https://www.wikidata.org/wiki/Q207643","display_name":"Linear map","level":2,"score":0.3037000000476837},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.2987000048160553},{"id":"https://openalex.org/C173110770","wikidata":"https://www.wikidata.org/wiki/Q382698","display_name":"Affine space","level":3,"score":0.2922999858856201},{"id":"https://openalex.org/C6802819","wikidata":"https://www.wikidata.org/wiki/Q1072174","display_name":"Linear system","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.25699999928474426},{"id":"https://openalex.org/C20050597","wikidata":"https://www.wikidata.org/wiki/Q261527","display_name":"Linear form","level":2,"score":0.25369998812675476}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.08721","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.08721","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.08721","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.08721","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"neural":[1,92],"classifiers":[2],"commonly":[3],"rely":[4],"on":[5,21,71],"linear":[6,30,111,119],"readouts,":[7],"yet":[8],"predictive":[9],"metrics":[10],"alone":[11],"do":[12],"not":[13],"characterize":[14],"the":[15,19,28,64,72,150],"class-wise":[16,140,173],"geometry":[17],"of":[18],"representations":[20,89],"which":[22],"such":[23],"readouts":[24],"operate.":[25],"We":[26,94,131],"introduce":[27],"directional":[29],"separability":[31],"measure":[32],"(LSM),":[33],"a":[34,42,47,134],"finite-sample":[35,164],"diagnostic":[36],"for":[37,138],"one-sided":[38],"affine":[39,54,103,136],"separability.":[40],"For":[41],"target":[43,73],"class":[44],"A":[45,61],"and":[46,62,85,106,154,167,179],"competing":[48],"set":[49],"B,":[50],"LSM":[51,141,170],"searches":[52],"over":[53],"halfspaces":[55],"that":[56,68],"contain":[57],"all":[58],"samples":[59],"in":[60,142],"measures":[63],"smallest":[65],"competing-sample":[66],"intrusion":[67,174],"must":[69],"remain":[70],"side,":[74],"normalized":[75],"by":[76,118,124],"|A|.":[77],"The":[78],"resulting":[79],"quantity":[80],"is":[81],"asymmetric,":[82],"class-wise,":[83],"target-normalized,":[84],"applicable":[86],"to":[87,101,171],"finite":[88],"extracted":[90],"from":[91,121,149],"networks.":[93],"establish":[95],"its":[96],"supporting-hyperplane":[97],"characterization,":[98],"relate":[99],"it":[100],"optimal":[102],"classification":[104],"accuracy,":[105],"prove":[107],"invariance":[108],"under":[109],"full-rank":[110],"embeddings.":[112],"These":[113],"results":[114],"separate":[115],"changes":[116],"caused":[117,123],"reparameterization":[120],"those":[122],"information":[125],"loss":[126],"or":[127],"nonlinear":[128],"geometric":[129,165],"transformations.":[130],"also":[132],"give":[133],"penalty-based":[135],"search":[137],"estimating":[139],"high-dimensional":[143],"features,":[144],"with":[145],"reported":[146],"values":[147],"computed":[148],"original":[151],"discrete":[152],"preservation":[153],"violation":[155],"criterion.":[156],"Finally,":[157],"we":[158],"analyze":[159],"coordinatewise":[160],"gated":[161],"nonlinearities":[162],"as":[163],"operators":[166],"empirically":[168],"use":[169],"diagnose":[172],"across":[175],"common":[176],"deep-learning":[177],"components":[178],"architectures.":[180]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-10T00:00:00"}