{"id":"https://openalex.org/W7153085568","doi":"https://doi.org/10.48550/arxiv.2604.08468","title":"TTVS: Boosting Self-Exploring Reinforcement Learning via Test-time Variational Synthesis","display_name":"TTVS: Boosting Self-Exploring Reinforcement Learning via Test-time Variational Synthesis","publication_year":2026,"publication_date":"2026-04-09","ids":{"openalex":"https://openalex.org/W7153085568","doi":"https://doi.org/10.48550/arxiv.2604.08468"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.08468","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08468","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.08468","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011529251","display_name":"Sikai Bai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bai, Sikai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133328915","display_name":"Haoxi Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Haoxi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100436819","display_name":"Jie Zhang","orcid":"https://orcid.org/0000-0002-8925-1011"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133333341","display_name":"Yongjiang Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yongjiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133371521","display_name":"Song Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Song","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"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/T10028","display_name":"Topic Modeling","score":0.550000011920929,"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/T10028","display_name":"Topic Modeling","score":0.550000011920929,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.11219999939203262,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.08020000159740448,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7290999889373779},{"id":"https://openalex.org/keywords/overfitting","display_name":"Overfitting","score":0.7111999988555908},{"id":"https://openalex.org/keywords/boosting","display_name":"Boosting (machine learning)","score":0.5436999797821045},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4575999975204468},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.37459999322891235},{"id":"https://openalex.org/keywords/verifiable-secret-sharing","display_name":"Verifiable secret sharing","score":0.33399999141693115},{"id":"https://openalex.org/keywords/temporal-difference-learning","display_name":"Temporal difference learning","score":0.3244999945163727}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7669000029563904},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7290999889373779},{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.7111999988555908},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6639999747276306},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.5436999797821045},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4575999975204468},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4438999891281128},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.37459999322891235},{"id":"https://openalex.org/C85847156","wikidata":"https://www.wikidata.org/wiki/Q59015987","display_name":"Verifiable secret sharing","level":3,"score":0.33399999141693115},{"id":"https://openalex.org/C196340769","wikidata":"https://www.wikidata.org/wiki/Q7698910","display_name":"Temporal difference learning","level":3,"score":0.3244999945163727},{"id":"https://openalex.org/C176248197","wikidata":"https://www.wikidata.org/wiki/Q458526","display_name":"Probably approximately correct learning","level":4,"score":0.3133000135421753},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3018999993801117},{"id":"https://openalex.org/C116222747","wikidata":"https://www.wikidata.org/wiki/Q220888","display_name":"Falsifiability","level":2,"score":0.29670000076293945},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.29100000858306885},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.2867000102996826},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.2637999951839447},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25850000977516174},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.25220000743865967}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.08468","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08468","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.08468","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08468","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":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Despite":[0],"significant":[1],"advances":[2],"in":[3,21],"Large":[4],"Reasoning":[5],"Models":[6],"(LRMs)":[7],"driven":[8],"by":[9,52,81],"reinforcement":[10],"learning":[11,53],"with":[12,133],"verifiable":[13],"rewards":[14],"(RLVR),":[15],"this":[16,65],"paradigm":[17],"is":[18,29],"fundamentally":[19],"limited":[20],"specialized":[22],"or":[23,32],"novel":[24,74],"domains":[25],"where":[26],"such":[27],"supervision":[28],"prohibitively":[30],"expensive":[31],"unavailable,":[33],"posing":[34],"a":[35,46,73,106],"key":[36],"challenge":[37],"for":[38],"test-time":[39,43,154,161],"adaptation.":[40],"While":[41],"existing":[42],"methods":[44,163],"offer":[45],"potential":[47],"solution,":[48],"they":[49],"are":[50],"constrained":[51],"from":[54,87],"static":[55,102],"query":[56],"sets,":[57],"risking":[58],"overfitting":[59],"to":[60,79,116],"textual":[61],"patterns.":[62],"To":[63],"address":[64],"gap,":[66],"we":[67],"introduce":[68],"Test-Time":[69],"Variational":[70,98],"Synthesis":[71],"(TTVS),":[72],"framework":[75],"that":[76],"enables":[77],"LRMs":[78],"self-evolve":[80],"dynamically":[82],"augmenting":[83],"the":[84,114],"training":[85],"stream":[86,108],"unlabeled":[88,153],"test":[89,103],"queries.":[90],"TTVS":[91,142,156],"comprises":[92],"two":[93],"synergistic":[94],"modules:":[95],"(1)":[96],"Online":[97],"Synthesis,":[99],"which":[100,129],"transforms":[101],"queries":[104],"into":[105],"dynamic":[107],"of":[109],"diverse,":[110],"semantically-equivalent":[111],"variations,":[112],"enforcing":[113],"model":[115,148],"learn":[117],"underlying":[118],"problem":[119],"logic":[120],"rather":[121],"than":[122],"superficial":[123],"patterns;":[124],"(2)":[125],"Test-time":[126],"Hybrid":[127],"Exploration,":[128],"balances":[130],"accuracy-driven":[131],"exploitation":[132],"consistency-driven":[134],"exploration":[135],"across":[136,146],"synthetic":[137],"variants.":[138],"Extensive":[139],"experiments":[140],"show":[141],"yields":[143],"superior":[144],"performance":[145],"eight":[147],"architectures.":[149],"Notably,":[150],"using":[151],"only":[152,158],"data,":[155],"not":[157],"surpasses":[159],"other":[160],"adaptation":[162],"but":[164],"also":[165],"outperforms":[166],"state-of-the-art":[167],"supervised":[168],"RL-based":[169],"techniques":[170],"trained":[171],"on":[172],"vast,":[173],"high-quality":[174],"labeled":[175],"data.":[176]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-11T00:00:00"}