{"id":"https://openalex.org/W7166886611","doi":"https://doi.org/10.48550/arxiv.2606.31599","title":"Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning","display_name":"Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning","publication_year":2026,"publication_date":"2026-06-30","ids":{"openalex":"https://openalex.org/W7166886611","doi":"https://doi.org/10.48550/arxiv.2606.31599"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.31599","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31599","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":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.2606.31599","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5104264401","display_name":"Kaitao Chen","orcid":"https://orcid.org/0000-0001-8422-7920"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Kaitao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016584164","display_name":"Weidong Zhao","orcid":"https://orcid.org/0000-0003-4144-7303"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Weiqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139757487","display_name":"Jiamin Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Jiamin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139672763","display_name":"Qihao Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Qihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139756549","display_name":"Shangquan Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Shangquan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139721945","display_name":"Chunfeng Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Chunfeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139743506","display_name":"Xiaosong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xiaosong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139802224","display_name":"Mu Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Mu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5058203667","display_name":"Mianxin Liu","orcid":"https://orcid.org/0000-0001-5171-778X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Mianxin","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":0.9588000178337097,"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":0.9588000178337097,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.015599999576807022,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.00419999985024333,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7656000256538391},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6541000008583069},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6283000111579895},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.5288000106811523},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4821000099182129},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.3716000020503998},{"id":"https://openalex.org/keywords/policy-learning","display_name":"Policy learning","score":0.3474000096321106},{"id":"https://openalex.org/keywords/task-analysis","display_name":"Task analysis","score":0.32760000228881836}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7656000256538391},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7353000044822693},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6541000008583069},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6283000111579895},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6021999716758728},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5288000106811523},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4821000099182129},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.44780001044273376},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.3716000020503998},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.3474000096321106},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.32760000228881836},{"id":"https://openalex.org/C40506919","wikidata":"https://www.wikidata.org/wiki/Q7452469","display_name":"Sequence learning","level":2,"score":0.3206000030040741},{"id":"https://openalex.org/C2780929884","wikidata":"https://www.wikidata.org/wiki/Q737460","display_name":"Blindness","level":2,"score":0.3068999946117401},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.29019999504089355},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.28690001368522644},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.27239999175071716},{"id":"https://openalex.org/C115086926","wikidata":"https://www.wikidata.org/wiki/Q17004651","display_name":"Causal reasoning","level":3,"score":0.26910001039505005},{"id":"https://openalex.org/C2776650193","wikidata":"https://www.wikidata.org/wiki/Q264661","display_name":"Obstacle","level":2,"score":0.26600000262260437},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2606000006198883},{"id":"https://openalex.org/C2776654903","wikidata":"https://www.wikidata.org/wiki/Q2601463","display_name":"SAFER","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2533999979496002}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.31599","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31599","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":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.2606.31599","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31599","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.7882817983627319,"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision-language":[0],"models":[1],"(VLMs)":[2],"combining":[3],"reinforcement":[4],"learning":[5,104],"(RL)":[6],"ignite":[7],"remarkable":[8],"progress":[9],"in":[10],"multimodal":[11,57,168],"reasoning,":[12],"yet":[13],"still":[14],"struggle":[15],"with":[16,81],"medical":[17,42,56,126,167],"images,":[18],"which":[19],"typically":[20],"exhibit":[21],"extremely":[22],"sparse":[23],"visual":[24,34,51,131],"evidence":[25],"to":[26,69,133],"inform":[27],"clinical":[28],"decision-making.":[29],"We":[30],"recognize":[31],"that":[32],"pruning":[33,53,72],"tokens":[35,132],"outside":[36],"the":[37,91,101,108,119,136],"grounding":[38,89],"region":[39],"greatly":[40],"enhances":[41],"reasoning.":[43,169],"However,":[44],"a":[45,64,142],"united":[46],"RL":[47,66],"framework":[48],"for":[49,166],"active":[50],"token":[52,71],"(VTP)":[54],"and":[55,73,115,148,159],"reasoning":[58,95],"remains":[59],"unestablished.":[60],"Here,":[61],"we":[62,99],"propose":[63],"dual-stream":[65],"framework,":[67],"ViToS,":[68],"fulfill":[70],"question":[74],"answering.":[75],"ViToS":[76,155],"trains":[77],"one":[78,86],"policy":[79,103,121],"model":[80],"two":[82],"task":[83],"branches,":[84],"where":[85],"focuses":[87],"on":[88,124,146,152],"while":[90,140],"other":[92],"conducts":[93],"token-sparse":[94],"after":[96],"VTP.":[97],"Furthermore,":[98],"solve":[100],"coupled":[102],"problem":[105],"by":[106],"introducing":[107],"cross-feedback":[109],"sequential":[110],"optimization,":[111],"avoiding":[112],"gradient":[113],"conflict":[114],"facilitating":[116],"convergence":[117],"of":[118,135],"shared":[120],"model.":[122],"Evaluated":[123],"seven":[125],"benchmarks,":[127],"our":[128],"method":[129],"reduces":[130],"77%":[134],"original":[137],"sequence":[138],"length":[139],"achieving":[141],"108.27%":[143],"relative":[144,150],"performance":[145,151,158],"Lingshu-7B":[147],"104.16%":[149],"HuatuoGPT-Vision-7B.":[153],"Overall,":[154],"delivers":[156],"superior":[157],"inference":[160],"speedup,":[161],"establishing":[162],"an":[163],"efficient":[164],"paradigm":[165]},"counts_by_year":[],"updated_date":"2026-07-02T06:18:51.028212","created_date":"2026-07-02T00:00:00"}