{"id":"https://openalex.org/W4416034987","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.217","title":"Sparkle: Mastering Basic Spatial Capabilities in Vision Language Models Elicits Generalization to Spatial Reasoning","display_name":"Sparkle: Mastering Basic Spatial Capabilities in Vision Language Models Elicits Generalization to Spatial Reasoning","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416034987","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.217"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.217","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.217","pdf_url":"https://aclanthology.org/2025.findings-emnlp.217.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.217.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102021157","display_name":"Yihong Tang","orcid":"https://orcid.org/0000-0002-3684-3072"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yihong Tang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114534567","display_name":"Ao Qu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ao Qu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060872432","display_name":"Zhaokai Wang","orcid":"https://orcid.org/0009-0007-4357-7802"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhaokai Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027933499","display_name":"Dingyi Zhuang","orcid":"https://orcid.org/0000-0003-3208-6016"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dingyi Zhuang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036524439","display_name":"Zhaofeng Wu","orcid":"https://orcid.org/0000-0002-1470-5648"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhaofeng Wu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115596456","display_name":"Wei Ma","orcid":"https://orcid.org/0000-0002-7239-2010"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Ma","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101998459","display_name":"Shenhao Wang","orcid":"https://orcid.org/0000-0003-4374-8193"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shenhao Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027813095","display_name":"Yunhan Zheng","orcid":"https://orcid.org/0000-0001-5114-7561"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yunhan Zheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071764531","display_name":"Zhan Zhao","orcid":"https://orcid.org/0000-0002-3139-7579"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhan Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5023905102","display_name":"Jinhua Zhao","orcid":"https://orcid.org/0000-0002-1929-7583"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jinhua Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":10,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.34297289,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"4083","last_page":"4103"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11596","display_name":"Constraint Satisfaction and Optimization","score":0.3617999851703644,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T11596","display_name":"Constraint Satisfaction and Optimization","score":0.3617999851703644,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T12694","display_name":"Categorization, perception, and language","score":0.12250000238418579,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.12120000272989273,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.5976999998092651},{"id":"https://openalex.org/keywords/spatial-intelligence","display_name":"Spatial intelligence","score":0.5042999982833862},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.3515999913215637},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.34290000796318054},{"id":"https://openalex.org/keywords/spatial-relation","display_name":"Spatial relation","score":0.31150001287460327}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6499000191688538},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.5976999998092651},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5924999713897705},{"id":"https://openalex.org/C155911833","wikidata":"https://www.wikidata.org/wiki/Q3817354","display_name":"Spatial intelligence","level":2,"score":0.5042999982833862},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.41440001130104065},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3515999913215637},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.34290000796318054},{"id":"https://openalex.org/C27511587","wikidata":"https://www.wikidata.org/wiki/Q2178623","display_name":"Spatial relation","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C2780878386","wikidata":"https://www.wikidata.org/wiki/Q1659648","display_name":"Visual language","level":2,"score":0.30480000376701355},{"id":"https://openalex.org/C159620131","wikidata":"https://www.wikidata.org/wiki/Q1938983","display_name":"Spatial analysis","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.2646999955177307},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.26460000872612}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.217","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.217","pdf_url":"https://aclanthology.org/2025.findings-emnlp.217.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.217","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.217","pdf_url":"https://aclanthology.org/2025.findings-emnlp.217.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416034987.pdf","grobid_xml":"https://content.openalex.org/works/W4416034987.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision":[0],"language":[1],"models":[2],"(VLMs)":[3],"perform":[4],"well":[5],"on":[6,29,71,98,151,156],"many":[7],"tasks":[8,27,53,101,153],"but":[9,154],"often":[10],"fail":[11],"at":[12],"spatial":[13,25,48,63,73,78,100,161,169,177,187],"reasoning,":[14],"which":[15],"is":[16],"essential":[17],"for":[18,140,184],"navigation":[19],"and":[20,88,106,135,158,179],"interaction":[21],"with":[22,146],"physical":[23],"environments.Many":[24],"reasoning":[26,64,79,105,162,178],"depend":[28],"fundamental":[30],"two-dimensional":[31],"(2D)":[32],"skills,":[33],"yet":[34],"our":[35],"evaluation":[36],"shows":[37],"that":[38,54,91,102,123,143,166],"state-of-the-art":[39],"VLMs":[40,66,144],"give":[41],"implausible":[42],"or":[43],"incorrect":[44],"answers":[45],"to":[46,113,127],"composite":[47,157],"problems,":[49],"including":[50],"simple":[51],"pathfinding":[52],"humans":[55],"solve":[56],"effortlessly.To":[57],"address":[58],"this,":[59,117],"we":[60,118],"enhance":[61],"2D":[62,77],"in":[65],"by":[67],"training":[68],"them":[69],"only":[70,150],"basic":[72,152,168],"capabilities.We":[74],"first":[75],"disentangle":[76],"into":[80],"three":[81,133],"core":[82],"components:":[83],"direction":[84],"comprehension,":[85],"distance":[86],"estimation,":[87],"localization.We":[89],"hypothesize":[90],"mastering":[92],"these":[93,132],"skills":[94,170],"substantially":[95],"improves":[96],"performance":[97],"complex":[99,176],"require":[103],"advanced":[104],"combinatorial":[107],"problem":[108],"solving,":[109],"while":[110],"also":[111,155],"generalizing":[112],"real-world":[114,160],"scenarios.To":[115],"test":[116],"introduce":[119],"Sparkle,":[120],"a":[121,181],"framework":[122],"generates":[124],"synthetic":[125,172],"data":[126],"provide":[128],"targeted":[129],"supervision":[130],"across":[131],"capabilities":[134],"yields":[136],"an":[137],"instruction":[138],"dataset":[139],"each.Experiments":[141],"show":[142],"fine-tuned":[145],"Sparkle":[147],"improve":[148],"not":[149],"out-of-distribution":[159],"tasks.These":[163],"results":[164],"indicate":[165],"enhancing":[167],"through":[171],"generalization":[173],"effectively":[174],"advances":[175],"offers":[180],"systematic":[182],"strategy":[183],"boosting":[185],"the":[186],"understanding":[188],"of":[189],"VLMs.":[190]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-11-08T00:00:00"}