{"id":"https://openalex.org/W7204137760","doi":"https://doi.org/10.48550/arxiv.2608.21247","title":"Just Noticeable Difference Modeling for Token Compression in Vision-Language-Action Models","display_name":"Just Noticeable Difference Modeling for Token Compression in Vision-Language-Action Models","publication_year":2026,"publication_date":"2026-08-21","ids":{"openalex":"https://openalex.org/W7204137760","doi":"https://doi.org/10.48550/arxiv.2608.21247"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.21247","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21247","pdf_url":"https://arxiv.org/pdf/2608.21247","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.21247","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5132857178","display_name":"Zhuoyuan Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Li, Zhuoyuan","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134108332","display_name":"Rui Zhao","orcid":"https://orcid.org/0000-0002-9577-7714","observed_orcids":["https://orcid.org/0000-0002-9577-7714"]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zhao, Rui","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5143736222","display_name":"Jin Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Wang, Jin","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5121675028","display_name":"Hanwei Zhu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zhu, Hanwei","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138725213","display_name":"Cong Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zhang, Cong","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034069364","display_name":"Giuseppe Valenzise","orcid":"https://orcid.org/0000-0002-5840-5743","observed_orcids":["https://orcid.org/0000-0002-5840-5743"]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Valenzise, Giuseppe","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124455472","display_name":"Weisi Lin","orcid":"https://orcid.org/0000-0001-9866-1947","observed_orcids":["https://orcid.org/0000-0001-9866-1947"]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Lin, Weisi","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5019678322","display_name":"Kin\u2010Man Lam","orcid":"https://orcid.org/0000-0002-0422-8454","observed_orcids":["https://orcid.org/0000-0002-0422-8454"]},"institutions":[{"id":"https://openalex.org/I14243506","display_name":"Hong Kong Polytechnic University","ror":"https://ror.org/0030zas98","country_code":"HK","type":"education","lineage":["https://openalex.org/I14243506"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Lam, Kin-Man","raw_affiliation_strings":["Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical and Electronic Engineering , The Hong Kong Polytechnic University , Hong Kong","institution_ids":["https://openalex.org/I14243506"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I14243506"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.38600000739097595,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.38600000739097595,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.25290000438690186,"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/T10653","display_name":"Robot Manipulation and Learning","score":0.21529999375343323,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/token-compression","display_name":"token compression","score":0.9549999833106995},{"id":"https://openalex.org/keywords/just-noticeable-difference","display_name":"just noticeable difference","score":0.9670000076293945},{"id":"https://openalex.org/keywords/vision-language-action-models","display_name":"vision-language-action models","score":0.9409999847412109},{"id":"https://openalex.org/keywords/embodied-agents","display_name":"embodied agents","score":0.8199999928474426},{"id":"https://openalex.org/keywords/token-pruning","display_name":"token pruning","score":0.7850000262260437},{"id":"https://openalex.org/keywords/kv-cache-reuse","display_name":"KV cache reuse","score":0.8820000290870667},{"id":"https://openalex.org/keywords/inference-cost-reduction","display_name":"inference cost reduction","score":0.5210000276565552},{"id":"https://openalex.org/keywords/openvla","display_name":"OpenVLA","score":0.531000018119812},{"id":"https://openalex.org/keywords/libero-benchmark","display_name":"LIBERO benchmark","score":0.753000020980835}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.677299976348877},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5996999740600586},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.47999998927116394},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4507000148296356},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.43320000171661377},{"id":"https://openalex.org/C57654395","wikidata":"https://www.wikidata.org/wiki/Q1097775","display_name":"Compression artifact","level":5,"score":0.4117000102996826},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4083000123500824},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.3806000053882599},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.35910001397132874},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.3578000068664551},{"id":"https://openalex.org/C94835093","wikidata":"https://www.wikidata.org/wiki/Q3113333","display_name":"Data compression ratio","level":5,"score":0.34779998660087585},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3093999922275543},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3001999855041504},{"id":"https://openalex.org/C100609095","wikidata":"https://www.wikidata.org/wiki/Q1335050","display_name":"Embodied cognition","level":2,"score":0.2881999909877777},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.26510000228881836},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2632000148296356},{"id":"https://openalex.org/C25797200","wikidata":"https://www.wikidata.org/wiki/Q828137","display_name":"Compression ratio","level":3,"score":0.2621999979019165},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.26019999384880066},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.25999999046325684},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.2554999887943268},{"id":"https://openalex.org/C2779110517","wikidata":"https://www.wikidata.org/wiki/Q1240788","display_name":"Supervisor","level":2,"score":0.25440001487731934}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.21247","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21247","pdf_url":"https://arxiv.org/pdf/2608.21247","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.2608.21247","doi":"https://doi.org/10.48550/arxiv.2608.21247","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.21247","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:2608.21247","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21247","pdf_url":"https://arxiv.org/pdf/2608.21247","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204137760.pdf","grobid_xml":"https://content.openalex.org/works/W7204137760.grobid-xml"},"referenced_works_count":16,"referenced_works":["https://openalex.org/W4306886919","https://openalex.org/W4398192846","https://openalex.org/W4399414358","https://openalex.org/W4399695759","https://openalex.org/W4400034355","https://openalex.org/W4403346092","https://openalex.org/W4414079054","https://openalex.org/W4415059540","https://openalex.org/W4415250071","https://openalex.org/W4416661803","https://openalex.org/W4417174884","https://openalex.org/W4417536563","https://openalex.org/W6948104436","https://openalex.org/W7081567423","https://openalex.org/W7083693084","https://openalex.org/W7115560929"],"related_works":["https://openalex.org/W7222513017","https://openalex.org/W7106685829","https://openalex.org/W7203873914","https://openalex.org/W7114933106","https://openalex.org/W7161030779","https://openalex.org/W4417151041","https://openalex.org/W7151091321","https://openalex.org/W7165618836","https://openalex.org/W4415250071","https://openalex.org/W4415037022"],"abstract_inverted_index":{"Token":[0],"compression":[1,60,217,244,249],"has":[2],"become":[3],"a":[4,88,157,178,187,212],"key":[5,81],"technique":[6],"for":[7,33,83,215,228],"reducing":[8],"the":[9,80,107,120,152,172,198,232],"inference":[10],"cost":[11],"of":[12,109,156],"large":[13],"foundation":[14],"models,":[15],"with":[16,235],"approaches":[17],"such":[18,66],"as":[19,67,211],"token":[20,89,165,223],"pruning":[21],"and":[22,30,44,72,222,237],"KV-cache":[23],"reuse":[24,221],"widely":[25],"adopted":[26],"in":[27,97,119,161,192],"vision-language":[28],"models":[29],"recently":[31],"explored":[32],"embodied":[34,37,146],"agents.":[35],"In":[36],"agents,":[38],"tokens":[39,227],"not":[40],"only":[41,77,170],"support":[42],"perception":[43,147],"semantic":[45],"understanding":[46],"but":[47],"also":[48],"directly":[49],"affect":[50],"latency-sensitive":[51],"closed-loop":[52,162],"robot":[53],"action":[54,154,174],"prediction.":[55],"Existing":[56],"schemes":[57],"typically":[58],"guide":[59],"using":[61],"redundancy":[62],"or":[63],"importance":[64],"cues,":[65],"visual":[68,122],"similarity,":[69],"attention":[70],"scores,":[71],"saliency.":[73],"However,":[74],"these":[75],"cues":[76],"indirectly":[78],"measure":[79],"factor":[82],"safe":[84],"compression:":[85],"how":[86],"much":[87],"can":[90],"change":[91,166],"before":[92],"causing":[93],"an":[94],"unacceptable":[95],"deviation":[96,175],"downstream":[98,131],"actions.":[99],"This":[100],"receiver-dependent":[101],"tolerance":[102,118],"is":[103,167],"closely":[104],"related":[105],"to":[106,130,145,196],"principle":[108],"just":[110],"noticeable":[111],"difference":[112],"(JND).":[113],"Classical":[114],"JND":[115,126,143,190],"characterizes":[116],"signal":[117],"human":[121],"system,":[123],"while":[124,202],"machine-oriented":[125],"extends":[127,142],"this":[128,136,183],"concept":[129],"machine":[132],"responses.":[133,205],"Building":[134],"on":[135,231],"progression,":[137],"we":[138,185],"introduce":[139],"Action-JND,":[140],"which":[141],"modeling":[144],"by":[148],"defining":[149],"noticeability":[150],"through":[151],"language-conditioned":[153],"response":[155],"vision-language-action":[158],"(VLA)":[159],"policy":[160,204],"control.":[163],"A":[164],"considered":[168],"admissible":[169],"when":[171],"induced":[173],"remains":[176],"within":[177],"tolerated":[179],"margin.":[180],"To":[181],"realize":[182],"concept,":[184],"develop":[186],"lightweight":[188],"token-wise":[189],"estimator":[191],"deep":[193],"visual-feature":[194],"space":[195],"predict":[197],"maximum":[199],"tolerable":[200],"perturbation":[201],"preserving":[203],"The":[206],"resulting":[207],"action-tolerance":[208],"score":[209],"serves":[210],"plug-and-play":[213],"criterion":[214],"VLA":[216],"paradigms,":[218],"including":[219],"stale-KV":[220],"pruning,":[224],"prioritizing":[225],"action-tolerant":[226],"compression.":[229],"Experiments":[230],"LIBERO":[233],"benchmark":[234],"OpenVLA":[236],"OpenVLA-OFT":[238],"demonstrate":[239],"that":[240],"Action-JND":[241],"consistently":[242],"improves":[243],"reliability,":[245],"especially":[246],"under":[247],"aggressive":[248],"ratios.":[250]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-25T00:00:00"}
