{"id":"https://openalex.org/W7159673651","doi":"https://doi.org/10.48550/arxiv.2604.28122","title":"Beyond Gaussian Bottlenecks: Topologically Aligned Encoding of Vision-Transformer Feature Spaces","display_name":"Beyond Gaussian Bottlenecks: Topologically Aligned Encoding of Vision-Transformer Feature Spaces","publication_year":2026,"publication_date":"2026-04-30","ids":{"openalex":"https://openalex.org/W7159673651","doi":"https://doi.org/10.48550/arxiv.2604.28122"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.28122","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.28122","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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.28122","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134946189","display_name":"Andrew Bond","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bond, Andrew","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134929704","display_name":"Ilkin Umut Melanlioglu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Melanlioglu, Ilkin Umut","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134932938","display_name":"Erkut Erdem","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Erdem, Erkut","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5000080119","display_name":"Aykut Erdem","orcid":"https://orcid.org/0000-0002-6280-8422"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Erdem, Aykut","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/T10531","display_name":"Advanced Vision and Imaging","score":0.6245999932289124,"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/T10531","display_name":"Advanced Vision and Imaging","score":0.6245999932289124,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.11729999631643295,"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/T10719","display_name":"3D Shape Modeling and Analysis","score":0.11309999972581863,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/autoencoder","display_name":"Autoencoder","score":0.5525000095367432},{"id":"https://openalex.org/keywords/gaussian","display_name":"Gaussian","score":0.45829999446868896},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.3937999904155731},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.37040001153945923},{"id":"https://openalex.org/keywords/point-cloud","display_name":"Point cloud","score":0.3646000027656555},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.34950000047683716},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.33230000734329224},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.323199987411499}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5641000270843506},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.5525000095367432},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.541100025177002},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.45829999446868896},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.4262000024318695},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.3937999904155731},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.37040001153945923},{"id":"https://openalex.org/C131979681","wikidata":"https://www.wikidata.org/wiki/Q1899648","display_name":"Point cloud","level":2,"score":0.3646000027656555},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.34950000047683716},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3467000126838684},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.33649998903274536},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.33230000734329224},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.323199987411499},{"id":"https://openalex.org/C104065381","wikidata":"https://www.wikidata.org/wiki/Q1002535","display_name":"Geometric modeling","level":2,"score":0.3228999972343445},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3163999915122986},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.30140000581741333},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.29899999499320984},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.29109999537467957},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2838999927043915},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.28299999237060547},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.28040000796318054},{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.27810001373291016},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2700999975204468},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.26809999346733093},{"id":"https://openalex.org/C108882727","wikidata":"https://www.wikidata.org/wiki/Q2991685","display_name":"Solid modeling","level":2,"score":0.2655999958515167},{"id":"https://openalex.org/C56435381","wikidata":"https://www.wikidata.org/wiki/Q1196371","display_name":"Geometric transformation","level":3,"score":0.25540000200271606}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.28122","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.28122","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":"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":"doi:10.48550/arxiv.2604.28122","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.28122","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":"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"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/11","display_name":"Sustainable cities and communities","score":0.4975453019142151}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"visual":[1,163],"world":[2,165],"modeling":[3,80],"systems":[4],"increasingly":[5],"rely":[6],"on":[7,60,84],"high-capacity":[8],"architectures":[9],"and":[10,62,75,119,131,164],"large-scale":[11],"data":[12],"to":[13,21,46,116],"produce":[14],"plausible":[15],"motion,":[16,73],"yet":[17],"they":[18],"often":[19],"fail":[20],"preserve":[22,117],"underlying":[23],"3D":[24,66],"geometry":[25,154],"or":[26],"physically":[27,161],"consistent":[28],"camera":[29,72,128],"dynamics.":[30],"A":[31],"key":[32],"limitation":[33],"lies":[34],"not":[35],"only":[36],"in":[37,41,113,147],"model":[38],"capacity,":[39],"but":[40],"the":[42,64,114],"latent":[43,55,65,107,153],"representations":[44,85],"used":[45],"encode":[47],"geometric":[48,120],"structure.":[49],"We":[50],"propose":[51],"S$^2$VAE,":[52],"a":[53,69,87,95,102,156],"geometry-first":[54],"learning":[56],"framework":[57],"that":[58,137],"focuses":[59],"compressing":[61],"representing":[63],"state":[67],"of":[68,98,104],"scene,":[70],"including":[71],"depth,":[74],"point-level":[76],"structure,":[77],"rather":[78],"than":[79],"appearance":[81],"alone.":[82],"Building":[83],"from":[86],"Visual":[88],"Geometry":[89],"Grounded":[90],"Transformer":[91],"(VGGT),":[92],"we":[93,135],"introduce":[94],"novel":[96],"type":[97],"variational":[99],"autoencoder":[100],"using":[101],"product":[103],"Power":[105],"Spherical":[106],"distributions,":[108],"explicitly":[109],"enforcing":[110],"hyperspherical":[111,139],"structure":[112],"bottleneck":[115],"directional":[118],"semantics":[121],"under":[122],"strong":[123],"compression.":[124],"Across":[125],"depth":[126],"estimation,":[127],"pose":[129],"recovery,":[130],"point":[132],"cloud":[133],"reconstruction,":[134],"show":[136],"geometry-aligned":[138],"latents":[140],"consistently":[141],"outperform":[142],"conventional":[143],"Gaussian":[144],"bottlenecks,":[145],"particularly":[146],"high-compression":[148],"regimes.":[149],"Our":[150],"results":[151],"highlight":[152],"as":[155],"first-class":[157],"design":[158],"choice":[159],"for":[160],"grounded":[162],"models.":[166]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-02T00:00:00"}