{"id":"https://openalex.org/W7167072006","doi":"https://doi.org/10.48550/arxiv.2607.00428","title":"HyFL-CLIP: Hyperbolic Fine-Tuning of CLIP for Robust Long-Context Understanding","display_name":"HyFL-CLIP: Hyperbolic Fine-Tuning of CLIP for Robust Long-Context Understanding","publication_year":2026,"publication_date":"2026-07-01","ids":{"openalex":"https://openalex.org/W7167072006","doi":"https://doi.org/10.48550/arxiv.2607.00428"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.00428","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.00428","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,"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.2607.00428","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5086590375","display_name":"Ji Ha Jang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jang, Ji Ha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139914049","display_name":"Hayeon Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Hayeon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034680280","display_name":"Chulwon \uc774\ucca0\uc6d0Lee","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Chulwon","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083424087","display_name":"Junghun James Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Junghun James","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139918466","display_name":"Se Young Chun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chun, Se Young","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9750999808311462,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9750999808311462,"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":0.006399999838322401,"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":0.004100000020116568,"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/context","display_name":"Context (archaeology)","score":0.629800021648407},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.5467000007629395},{"id":"https://openalex.org/keywords/euclidean-geometry","display_name":"Euclidean geometry","score":0.5221999883651733},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.51910001039505},{"id":"https://openalex.org/keywords/similarity","display_name":"Similarity (geometry)","score":0.49079999327659607},{"id":"https://openalex.org/keywords/logical-consequence","display_name":"Logical consequence","score":0.4140999913215637},{"id":"https://openalex.org/keywords/hyperbolic-space","display_name":"Hyperbolic space","score":0.383899986743927},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.37119999527931213},{"id":"https://openalex.org/keywords/a-priori-and-a-posteriori","display_name":"A priori and a posteriori","score":0.3675000071525574}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6675000190734863},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.629800021648407},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5467000007629395},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5418000221252441},{"id":"https://openalex.org/C129782007","wikidata":"https://www.wikidata.org/wiki/Q162886","display_name":"Euclidean geometry","level":2,"score":0.5221999883651733},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.51910001039505},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.49079999327659607},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.427700012922287},{"id":"https://openalex.org/C134752490","wikidata":"https://www.wikidata.org/wiki/Q374182","display_name":"Logical consequence","level":2,"score":0.4140999913215637},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3855000138282776},{"id":"https://openalex.org/C83677898","wikidata":"https://www.wikidata.org/wiki/Q1878538","display_name":"Hyperbolic space","level":2,"score":0.383899986743927},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.37119999527931213},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.3675000071525574},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.3564999997615814},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35100001096725464},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3294000029563904},{"id":"https://openalex.org/C120174047","wikidata":"https://www.wikidata.org/wiki/Q847073","display_name":"Euclidean distance","level":2,"score":0.32710000872612},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3264999985694885},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.29019999504089355},{"id":"https://openalex.org/C206352148","wikidata":"https://www.wikidata.org/wiki/Q209306","display_name":"Hyperbolic geometry","level":3,"score":0.2881999909877777},{"id":"https://openalex.org/C1667742","wikidata":"https://www.wikidata.org/wiki/Q10927554","display_name":"Image retrieval","level":3,"score":0.2867000102996826},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2849999964237213},{"id":"https://openalex.org/C183322885","wikidata":"https://www.wikidata.org/wiki/Q17007702","display_name":"Context model","level":3,"score":0.2797999978065491},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.27379998564720154},{"id":"https://openalex.org/C149364088","wikidata":"https://www.wikidata.org/wiki/Q185917","display_name":"Translation (biology)","level":4,"score":0.27070000767707825},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.27059999108314514},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.2644999921321869},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.2599000036716461},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.25209999084472656}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.00428","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.00428","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,"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.2607.00428","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.00428","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,"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"CLIP":[0,46,111],"(Contrastive":[1],"Language-Image":[2],"Pre-training)":[3],"has":[4],"become":[5],"a":[6,98],"de":[7],"facto":[8],"paradigm":[9],"for":[10,80],"image-text":[11,56],"alignment,":[12],"but":[13],"it":[14,185,217],"struggles":[15],"with":[16,47,152,166],"long-context":[17,138,161,181],"descriptions":[18],"(>77":[19],"tokens)":[20],"due":[21],"to":[22,66,122,188,218],"absolute":[23],"positional":[24,49],"encoding":[25],"and":[26,76,87,125,144,171],"pretraining":[27],"on":[28,157],"short":[29,141],"captions.":[30],"In":[31,183],"long":[32],"contexts,":[33],"sentences":[34],"are":[35],"often":[36,52],"reordered,":[37],"summarized,":[38],"or":[39],"partially":[40],"omitted.":[41],"Although":[42],"prior":[43,201],"works":[44],"extend":[45],"longer":[48],"encodings,":[50],"they":[51],"suffer":[53],"from":[54],"degraded":[55],"alignment":[57,107],"under":[58,195],"such":[59],"text":[60],"perturbations.":[61],"We":[62,203],"attribute":[63],"this":[64,93],"limitation":[65],"the":[67,104,199],"Euclidean":[68,110],"contrastive":[69],"objective,":[70],"which":[71],"enforces":[72],"strict":[73],"one-to-one":[74],"matching":[75],"lacks":[77],"explicit":[78],"mechanisms":[79],"modeling":[81],"hierarchical":[82,124,131],"relationships":[83,148],"between":[84],"global":[85],"context":[86],"its":[88,120],"constituent":[89,140],"elements.":[90],"To":[91],"address":[92],"issue,":[94],"we":[95],"propose":[96],"HyFL-CLIP,":[97],"hyperbolic":[99,113,150],"fine-tuning":[100],"framework":[101],"that":[102,176],"distills":[103],"well-established":[105],"text-image":[106],"learned":[108],"in":[109,191],"into":[112,211],"space":[114],"via":[115,149],"cross-manifold":[116],"similarity":[117],"distillation,":[118],"leveraging":[119],"geometry":[121],"capture":[123],"entailment":[126,151],"relations.":[127],"Our":[128],"method":[129],"models":[130],"semantics":[132],"by":[133,215],"linking":[134],"summarized":[135],"token-wise":[136],"features,":[137],"descriptions,":[139],"textual":[142,196],"components,":[143],"images,":[145],"capturing":[146],"part-whole":[147],"Einstein":[153],"midpoint":[154],"aggregation.":[155],"Experiments":[156],"diverse":[158],"benchmarks,":[159],"including":[160],"cross-modal":[162,164,173,193],"retrieval,":[163,170,174],"retrieval":[165,194],"caption":[167],"perturbations,":[168],"intra-modality":[169],"short-text":[172],"show":[175,205],"HyFL-CLIP":[177,206],"achieves":[178],"more":[179],"robust":[180],"understanding.":[182],"particular,":[184],"yields":[186],"up":[187],"19.5%":[189],"improvement":[190],"long-text":[192],"perturbations":[197],"over":[198],"best":[200],"method.":[202],"also":[204],"can":[207],"be":[208],"seamlessly":[209],"integrated":[210],"other":[212],"model":[213],"frameworks":[214],"applying":[216],"Stable":[219],"Diffusion":[220],"XL":[221],"(SDXL).":[222]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-03T00:00:00"}