{"id":"https://openalex.org/W4416257511","doi":"https://doi.org/10.48550/arxiv.2509.18154","title":"MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe","display_name":"MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe","publication_year":2025,"publication_date":"2025-09-16","ids":{"openalex":"https://openalex.org/W4416257511","doi":"https://doi.org/10.48550/arxiv.2509.18154"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2509.18154","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.18154","pdf_url":"https://arxiv.org/pdf/2509.18154","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2509.18154","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101790974","display_name":"Tianyu Yu","orcid":"https://orcid.org/0000-0003-0609-0994","observed_orcids":["https://orcid.org/0000-0003-0609-0994"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Tianyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Wang, Zefan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zefan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109704013","display_name":"Chongyi Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Chongyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Huang, Fuwei","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Fuwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016986549","display_name":"Wing\u2010Kin Ma","orcid":"https://orcid.org/0000-0001-7314-3537","observed_orcids":["https://orcid.org/0000-0001-7314-3537"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Wenshuo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"He, Zhihui","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Zhihui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Cai, Tianchi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Tianchi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Chen, Weize","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Weize","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032028380","display_name":"Yuxiang Huang","orcid":"https://orcid.org/0000-0003-1329-8814","observed_orcids":["https://orcid.org/0000-0003-1329-8814"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Yuxiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111700047","display_name":"Y. Xu H. Zhang L.M. Zhao","orcid":"https://orcid.org/0009-0007-6446-7119","observed_orcids":["https://orcid.org/0009-0007-6446-7119"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Yuanqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047505566","display_name":"Bokai Xu","orcid":"https://orcid.org/0000-0002-0224-416X","observed_orcids":["https://orcid.org/0000-0002-0224-416X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Bokai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Cui, Junbo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Junbo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100297804","display_name":"Yingjing Xu","orcid":"https://orcid.org/0009-0005-5856-1972","observed_orcids":["https://orcid.org/0009-0005-5856-1972"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Yingjing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120499911","display_name":"Liqing Ruan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ruan, Liqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Luoyuan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Luoyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100608411","display_name":"Hanyu Liu","orcid":"https://orcid.org/0000-0003-2394-5421","observed_orcids":["https://orcid.org/0000-0003-2394-5421","https://orcid.org/0009-0004-1002-0344","https://orcid.org/0009-0002-4150-0784","https://orcid.org/0009-0008-9850-8983","https://orcid.org/0009-0009-8533-8636"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Hanyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Tang, Jingkun","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Jingkun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Liu, Hongyuan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Hongyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Guo, Qining","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Qining","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101758824","display_name":"Wang Hu","orcid":"https://orcid.org/0000-0003-1373-3951","observed_orcids":["https://orcid.org/0000-0003-1373-3951"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Wenhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"He, Bingxiang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Bingxiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066557235","display_name":"Jie Zhou","orcid":"https://orcid.org/0000-0002-3836-4163","observed_orcids":["https://orcid.org/0000-0002-3836-4163","https://orcid.org/0009-0001-7276-8487"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Jie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102882093","display_name":"Jie Cai","orcid":"https://orcid.org/0000-0003-4935-649X","observed_orcids":["https://orcid.org/0000-0003-4935-649X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Jie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Qi, Ji","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qi, Ji","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Guo, Zonghao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Zonghao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Chen, Chi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Chi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087934040","display_name":"Guoyang Zeng","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeng, Guoyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104130542","display_name":"Yuxuan Li","orcid":"https://orcid.org/0009-0000-8613-148X","observed_orcids":["https://orcid.org/0009-0000-8613-148X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yuxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102542015","display_name":"Ganqu Cui","orcid":"https://orcid.org/0000-0001-6385-8547","observed_orcids":["https://orcid.org/0000-0001-6385-8547"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Ganqu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037110401","display_name":"Ning Ding","orcid":"https://orcid.org/0000-0002-3689-0696","observed_orcids":["https://orcid.org/0000-0002-3689-0696"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Ning","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110711399","display_name":"Han Xu","orcid":"https://orcid.org/0009-0007-0468-6336","observed_orcids":["https://orcid.org/0009-0007-0468-6336"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Xu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5075674109","display_name":"Yuan Yao","orcid":"https://orcid.org/0000-0002-8811-8200","observed_orcids":["https://orcid.org/0000-0002-8811-8200","https://orcid.org/0009-0000-5855-0011"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yao, Yuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100320722","display_name":"Zhiyuan Liu","orcid":"https://orcid.org/0000-0002-6331-0810","observed_orcids":["https://orcid.org/0000-0002-6331-0810"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zhiyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5046448314","display_name":"Maosong Sun","orcid":"https://orcid.org/0000-0002-6011-6115","observed_orcids":["https://orcid.org/0000-0002-6011-6115"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Maosong","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":1,"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":1.0,"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":1.0,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0,"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":0.0,"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/multimodal-large-language-models","display_name":"multimodal large language models","score":0.949999988079071},{"id":"https://openalex.org/keywords/text-recognition","display_name":"text recognition","score":0.9200000166893005},{"id":"https://openalex.org/keywords/hybrid-reinforcement-learning","display_name":"hybrid reinforcement learning","score":0.9760000109672546},{"id":"https://openalex.org/keywords/training-efficiency","display_name":"training efficiency","score":0.5820000171661377},{"id":"https://openalex.org/keywords/inference-efficiency","display_name":"inference efficiency","score":0.7570000290870667}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7989000082015991},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.6969000101089478},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6715999841690063},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.628600001335144},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.5498999953269958},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5415999889373779},{"id":"https://openalex.org/C2778671685","wikidata":"https://www.wikidata.org/wiki/Q219239","display_name":"Recipe","level":2,"score":0.5281000137329102},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.46970000863075256},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.43130001425743103},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4271000027656555},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4203000068664551},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.4027000069618225},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.37950000166893005},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.3560999929904938},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.33480000495910645},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3237000107765198},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3165999948978424},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.2939999997615814},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.28929999470710754},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.26809999346733093},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.25369998812675476},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2509.18154","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.18154","pdf_url":"https://arxiv.org/pdf/2509.18154","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2509.18154","doi":"https://doi.org/10.48550/arxiv.2509.18154","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.18154","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2509.18154","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.18154","pdf_url":"https://arxiv.org/pdf/2509.18154","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4406489046","https://openalex.org/W4417136227","https://openalex.org/W4416119086","https://openalex.org/W4405248620","https://openalex.org/W4390437716","https://openalex.org/W4403808773","https://openalex.org/W4405089583","https://openalex.org/W4394709684","https://openalex.org/W7127589246","https://openalex.org/W4415020574"],"abstract_inverted_index":{"Multimodal":[0],"Large":[1],"Language":[2],"Models":[3],"(MLLMs)":[4],"are":[5],"undergoing":[6],"rapid":[7],"progress":[8],"and":[9,19,33,51,64,78,87,94,105,127,169],"represent":[10],"the":[11,37,137,148],"frontier":[12],"of":[13,173],"AI":[14],"development.":[15],"However,":[16],"their":[17],"training":[18,65],"inference":[20,171],"efficiency":[21,50],"have":[22],"emerged":[23],"as":[24,125,133],"a":[25,67,80,95],"core":[26,57],"bottleneck":[27],"in":[28,59,102,112],"making":[29],"MLLMs":[30],"more":[31],"accessible":[32],"scalable.":[34],"To":[35],"address":[36],"challenges,":[38],"we":[39],"present":[40],"MiniCPM-V":[41,117,153],"4.5,":[42],"an":[43],"8B":[44],"parameter":[45],"model":[46,60,70],"designed":[47],"for":[48,72,84,100],"high":[49],"strong":[52,138],"performance.":[53],"We":[54],"introduce":[55],"three":[56],"improvements":[58],"architecture,":[61],"data":[62,92],"strategy":[63,99],"method:":[66],"unified":[68,81],"3D-Resampler":[69],"architecture":[71],"highly":[73],"compact":[74],"encoding":[75],"over":[76],"images":[77],"videos,":[79],"learning":[82,98],"paradigm":[83],"document":[85],"knowledge":[86],"text":[88],"recognition":[89],"without":[90],"heavy":[91],"engineering,":[93],"hybrid":[96],"reinforcement":[97],"proficiency":[101],"both":[103],"short":[104],"long":[106],"reasoning":[107],"modes.":[108],"Comprehensive":[109],"experimental":[110],"results":[111],"OpenCompass":[113],"evaluation":[114],"show":[115],"that":[116],"4.5":[118,154],"surpasses":[119],"widely":[120,149],"used":[121],"proprietary":[122],"models":[123,131,159],"such":[124,132],"GPT-4o-latest,":[126],"significantly":[128],"larger":[129],"open-source":[130],"Qwen2.5-VL":[134,174],"72B.":[135],"Notably,":[136],"performance":[139,157],"is":[140],"achieved":[141],"with":[142],"remarkable":[143],"efficiency.":[144],"For":[145],"example,":[146],"on":[147],"adopted":[150],"VideoMME":[151],"benchmark,":[152],"achieves":[155],"state-of-the-art":[156],"among":[158],"under":[160],"30B":[161],"size,":[162],"using":[163],"just":[164],"46.7\\%":[165],"GPU":[166],"memory":[167],"cost":[168],"8.7\\%":[170],"time":[172],"7B.":[175]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-10-18T09:31:52.927876","created_date":"2025-10-10T00:00:00"}