{"id":"https://openalex.org/W7139939368","doi":"https://doi.org/10.48550/arxiv.2603.18118","title":"Insight-V++: Towards Advanced Long-Chain Visual Reasoning with Multimodal Large Language Models","display_name":"Insight-V++: Towards Advanced Long-Chain Visual Reasoning with Multimodal Large Language Models","publication_year":2026,"publication_date":"2026-03-18","ids":{"openalex":"https://openalex.org/W7139939368","doi":"https://doi.org/10.48550/arxiv.2603.18118"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.18118","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18118","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.2603.18118","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130223047","display_name":"Yuhao Dong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, Yuhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130218391","display_name":"Zuyan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zuyan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130214784","display_name":"Shulin Tian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Shulin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082694832","display_name":"Yongming Rao","orcid":"https://orcid.org/0000-0003-3952-8753"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rao, Yongming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5130225906","display_name":"Ziwei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Ziwei","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":0.9729999899864197,"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":0.9729999899864197,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.0034000000450760126,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.003100000089034438,"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/visual-reasoning","display_name":"Visual reasoning","score":0.5026000142097473},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.4700999855995178},{"id":"https://openalex.org/keywords/spatial-intelligence","display_name":"Spatial intelligence","score":0.39800000190734863},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.3930000066757202},{"id":"https://openalex.org/keywords/qualitative-reasoning","display_name":"Qualitative reasoning","score":0.38499999046325684},{"id":"https://openalex.org/keywords/automated-reasoning","display_name":"Automated reasoning","score":0.34929999709129333},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.3422999978065491},{"id":"https://openalex.org/keywords/opportunistic-reasoning","display_name":"Opportunistic reasoning","score":0.3391999900341034},{"id":"https://openalex.org/keywords/reasoning-system","display_name":"Reasoning system","score":0.33660000562667847}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7892000079154968},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5968000292778015},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.5026000142097473},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.4700999855995178},{"id":"https://openalex.org/C155911833","wikidata":"https://www.wikidata.org/wiki/Q3817354","display_name":"Spatial intelligence","level":2,"score":0.39800000190734863},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.3930000066757202},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.38499999046325684},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3808000087738037},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3499000072479248},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.34929999709129333},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.3422999978065491},{"id":"https://openalex.org/C86827895","wikidata":"https://www.wikidata.org/wiki/Q7098582","display_name":"Opportunistic reasoning","level":4,"score":0.3391999900341034},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.33660000562667847},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.3301999866962433},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.3271999955177307},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3228999972343445},{"id":"https://openalex.org/C157657479","wikidata":"https://www.wikidata.org/wiki/Q2367247","display_name":"Closed captioning","level":3,"score":0.32030001282691956},{"id":"https://openalex.org/C2780660688","wikidata":"https://www.wikidata.org/wiki/Q25052564","display_name":"Multimodal learning","level":2,"score":0.31299999356269836},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.31209999322891235},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.29420000314712524},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.2922999858856201},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.25999999046325684},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.2526000142097473},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.25}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.18118","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18118","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.2603.18118","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18118","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":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0,21],"Language":[1,22],"Models":[2,23],"(LLMs)":[3],"have":[4],"achieved":[5],"remarkable":[6],"reliability":[7],"and":[8,39,91,123,130,166,172,210,219],"advanced":[9],"capabilities":[10,18,226],"through":[11],"extended":[12],"test-time":[13],"reasoning.":[14],"However,":[15],"extending":[16],"these":[17,153],"to":[19,30,118,127],"Multi-modal":[20],"(MLLMs)":[24],"remains":[25],"a":[26,31,49,65,73,111,115,124,199],"significant":[27,213],"challenge":[28],"due":[29],"critical":[32],"scarcity":[33],"of":[34],"high-quality,":[35],"long-chain":[36],"reasoning":[37,53,87,116,171,189,221],"data":[38,75,105],"optimized":[40],"training":[41],"pipelines.":[42],"To":[43,151],"bridge":[44],"this":[45],"gap,":[46],"we":[47,109,185],"present":[48],"unified":[50],"multi-agent":[51,196],"visual":[52],"framework":[54,137],"that":[55,82,98],"systematically":[56],"evolves":[57],"from":[58,181],"our":[59,135],"foundational":[60],"image-centric":[61],"model,":[62],"Insight-V,":[63],"into":[64],"generalized":[66],"spatial-temporal":[67,170],"architecture,":[68],"Insight-V++.":[69],"We":[70],"first":[71],"propose":[72],"scalable":[74],"generation":[76,191],"pipeline":[77],"equipped":[78],"with":[79,102],"multi-granularity":[80],"assessment":[81],"autonomously":[83],"synthesizes":[84],"structured,":[85],"complex":[86],"trajectories":[88],"across":[89,216],"image":[90,218],"video":[92,158,220],"domains":[93],"without":[94],"human":[95],"intervention.":[96],"Recognizing":[97],"directly":[99],"supervising":[100],"MLLMs":[101],"such":[103],"intricate":[104],"yields":[106],"sub-optimal":[107],"results,":[108],"design":[110],"dual-agent":[112],"architecture":[113],"comprising":[114],"agent":[117,126],"execute":[119],"extensive":[120],"analytical":[121],"chains,":[122],"summary":[125,183],"critically":[128],"evaluate":[129],"distill":[131],"final":[132],"outcomes.":[133],"While":[134],"initial":[136],"utilized":[138],"Direct":[139],"Preference":[140],"Optimization":[141],"(DPO),":[142],"its":[143],"off-policy":[144],"nature":[145],"fundamentally":[146],"constrained":[147],"reinforcement":[148],"learning":[149],"potential.":[150],"overcome":[152],"limitations,":[154],"particularly":[155],"for":[156],"long-horizon":[157],"understanding,":[159],"Insight-V++":[160],"introduces":[161],"two":[162],"novel":[163],"algorithms,":[164],"ST-GRPO":[165],"J-GRPO,":[167],"which":[168],"enhance":[169],"improve":[173],"evaluative":[174],"robustness.":[175],"Crucially,":[176],"by":[177],"leveraging":[178],"reliable":[179],"feedback":[180],"the":[182,194],"agent,":[184],"guide":[186],"an":[187],"iterative":[188],"path":[190],"process,":[192],"retraining":[193],"entire":[195],"system":[197],"in":[198],"continuous,":[200],"self-improving":[201],"loop.":[202],"Extensive":[203],"experiments":[204],"on":[205,227],"base":[206],"models":[207],"like":[208],"LLaVA-NeXT":[209],"Qwen2.5-VL":[211],"demonstrate":[212],"performance":[214],"gains":[215],"challenging":[217],"benchmarks":[222],"while":[223],"preserving":[224],"strong":[225],"traditional":[228],"perception-focused":[229],"tasks.":[230]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-21T00:00:00"}