{"id":"https://openalex.org/W4412888203","doi":"https://doi.org/10.18653/v1/2025.findings-acl.620","title":"Forgotten Polygons: Multimodal Large Language Models are Shape-Blind","display_name":"Forgotten Polygons: Multimodal Large Language Models are Shape-Blind","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888203","doi":"https://doi.org/10.18653/v1/2025.findings-acl.620"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.620","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.620","pdf_url":"https://aclanthology.org/2025.findings-acl.620.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: ACL 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-acl.620.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5081709342","display_name":"William J. Rudman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"William Rudman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019199746","display_name":"Michal Golovanevsky","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Michal Golovanevsky","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040676689","display_name":"Amir Bar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Amir Bar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5091920997","display_name":"Vedant Palit","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Vedant Palit","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001226970","display_name":"Yann LeCun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yann LeCun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014921416","display_name":"Carsten Eickhoff","orcid":"https://orcid.org/0000-0001-9895-4061"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Carsten Eickhoff","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5070578596","display_name":"Ritambhara Singh","orcid":"https://orcid.org/0000-0002-7523-160X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ritambhara Singh","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":true,"cited_by_count":2,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"11983","last_page":"11998"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9771000146865845,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9771000146865845,"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/T10028","display_name":"Topic Modeling","score":0.9523000121116638,"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/computer-science","display_name":"Computer science","score":0.7010325193405151},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4486366808414459},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.38737547397613525},{"id":"https://openalex.org/keywords/computer-graphics","display_name":"Computer graphics (images)","score":0.33087605237960815}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7010325193405151},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4486366808414459},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.38737547397613525},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.33087605237960815}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.620","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.620","pdf_url":"https://aclanthology.org/2025.findings-acl.620.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: ACL 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.620","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.620","pdf_url":"https://aclanthology.org/2025.findings-acl.620.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: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Quality Education","score":0.4699999988079071,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888203.pdf","grobid_xml":"https://content.openalex.org/works/W4412888203.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Despite":[0],"strong":[1],"performance":[2,25],"on":[3,26,90,150],"vision-language":[4],"tasks,":[5],"Multimodal":[6],"Large":[7],"Language":[8],"Models":[9],"(MLLMs)":[10],"struggle":[11],"with":[12,15,65],"mathematical":[13,138],"problemsolving,":[14],"both":[16,109],"open-source":[17],"and":[18,47,85,111,172],"state-ofthe-art":[19],"models":[20,67],"falling":[21],"short":[22],"of":[23,40,82,108,121],"human":[24],"visual-math":[27],"benchmarks.To":[28],"systematically":[29],"examine":[30],"visual-mathematical":[31],"reasoning":[32,56,139,165],"in":[33,62,72,145,166],"MLLMs,":[34],"we":[35,128],"(1)":[36],"evaluate":[37],"their":[38],"understanding":[39],"geometric":[41],"primitives,":[42],"(2)":[43],"test":[44],"multi-step":[45,137],"reasoning,":[46],"(3)":[48],"explore":[49],"a":[50],"potential":[51],"solution":[52],"to":[53,104,158],"improve":[54],"visual":[55,126,143,180],"capabilities.Our":[57],"findings":[58,160],"reveal":[59],"fundamental":[60],"shortcomings":[61],"shape":[63],"recognition,":[64],"top":[66],"achieving":[68],"under":[69],"50%":[70],"accuracy":[71,149],"identifying":[73],"regular":[74],"polygons.We":[75],"analyze":[76],"these":[77],"failures":[78],"through":[79],"the":[80,106,119],"lens":[81],"dual-process":[83],"theory":[84],"show":[86],"that":[87,162],"MLLMs":[88,102,167],"rely":[89],"System":[91,98,163],"1":[92],"(intuitive,":[93],"memorized":[94],"associations)":[95],"rather":[96],"than":[97],"2":[99,164],"(deliberate":[100],"reasoning).Consequently,":[101],"fail":[103],"count":[105],"sides":[107],"familiar":[110],"novel":[112],"shapes,":[113],"suggesting":[114],"they":[115],"have":[116],"neither":[117],"learned":[118],"concept":[120],"\"sides\"":[122],"nor":[123],"effectively":[124],"process":[125],"inputs.Finally,":[127],"propose":[129],"Visually":[130],"Cued":[131],"Chain-of-Thought":[132],"(VC-CoT)":[133],"prompting,":[134],"which":[135],"enhances":[136],"by":[140],"explicitly":[141],"referencing":[142],"annotations":[144],"diagrams,":[146],"boosting":[147],"GPT-4o's":[148],"an":[151,169],"irregular":[152],"polygon":[153],"sidecounting":[154],"task":[155],"from":[156],"7%":[157],"93%.Our":[159],"suggest":[161],"remains":[168],"open":[170],"problem,":[171],"visually-guided":[173],"prompting":[174],"is":[175],"essential":[176],"for":[177],"successfully":[178],"engaging":[179],"reasoning.":[181]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}