{"id":"https://openalex.org/W7160981632","doi":"https://doi.org/10.48550/arxiv.2605.10987","title":"AESOP: Adversarial Execution-path Selection to Overload Deep Learning Pipelines","display_name":"AESOP: Adversarial Execution-path Selection to Overload Deep Learning Pipelines","publication_year":2026,"publication_date":"2026-05-09","ids":{"openalex":"https://openalex.org/W7160981632","doi":"https://doi.org/10.48550/arxiv.2605.10987"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.10987","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10987","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.2605.10987","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5057784628","display_name":"Tingxi Li","orcid":"https://orcid.org/0000-0002-1726-6097"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Tingxi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125514739","display_name":"Mingfang Ji","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Mingfang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136012297","display_name":"Ravishka Shemal Rathnasuriya","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rathnasuriya, Ravishka Shemal","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136073154","display_name":"Simin Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Simin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136058398","display_name":"Yitao Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Yitao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136004515","display_name":"Wei Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Wei","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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9857000112533569,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9857000112533569,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.002899999963119626,"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/T11424","display_name":"Security and Verification in Computing","score":0.0010000000474974513,"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/latency","display_name":"Latency (audio)","score":0.5206000208854675},{"id":"https://openalex.org/keywords/pipeline-transport","display_name":"Pipeline transport","score":0.5026999711990356},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.49799999594688416},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.486299991607666},{"id":"https://openalex.org/keywords/adversarial-system","display_name":"Adversarial system","score":0.47769999504089355},{"id":"https://openalex.org/keywords/workload","display_name":"Workload","score":0.4593000113964081},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4205999970436096},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.3903000056743622}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6506999731063843},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.5206000208854675},{"id":"https://openalex.org/C175309249","wikidata":"https://www.wikidata.org/wiki/Q725864","display_name":"Pipeline transport","level":2,"score":0.5026999711990356},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.49799999594688416},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.486299991607666},{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.47769999504089355},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.4593000113964081},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4205999970436096},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.3903000056743622},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.37560001015663147},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.36070001125335693},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.35370001196861267},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.33869999647140503},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.337799996137619},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.3377000093460083},{"id":"https://openalex.org/C191172861","wikidata":"https://www.wikidata.org/wiki/Q7899321","display_name":"Upstream (networking)","level":2,"score":0.3359000086784363},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.3334999978542328},{"id":"https://openalex.org/C115874739","wikidata":"https://www.wikidata.org/wiki/Q825377","display_name":"Critical path method","level":2,"score":0.33149999380111694},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3278000056743622},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3197000026702881},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3118000030517578},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.3041999936103821},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.301800012588501},{"id":"https://openalex.org/C2781215313","wikidata":"https://www.wikidata.org/wiki/Q3493345","display_name":"SPARK (programming language)","level":2,"score":0.2883000075817108},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28130000829696655},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.27790001034736633},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2524000108242035}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.10987","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10987","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.2605.10987","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10987","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":{"Modern":[0],"machine":[1],"learning":[2],"deployments":[3],"increasingly":[4],"compose":[5],"specialized":[6],"models":[7,87],"into":[8],"dynamic":[9],"inference":[10],"pipelines,":[11],"where":[12,115],"upstream":[13],"components":[14],"produce":[15],"intermediate":[16],"predictions":[17],"that":[18,75,82],"determine":[19],"the":[20,46,102,116,124,184],"workload":[21,55],"and":[22,53,93,128,156,165,172,203],"inputs":[23,92],"of":[24,29,49],"downstream":[25],"components.":[26],"The":[27],"cost":[28,48],"processing":[30],"an":[31,79],"input":[32],"is":[33,66,118,186],"therefore":[34],"not":[35,187],"determined":[36],"by":[37,42,99],"any":[38],"single":[39,86],"model,":[40],"but":[41,189],"two":[43],"coupled":[44],"factors:":[45],"per-inference":[47],"each":[50],"invoked":[51],"component":[52],"its":[54],"volume.":[56],"Because":[57],"these":[58],"pipelines":[59,146,191],"run":[60],"under":[61],"hard":[62],"real-time":[63],"constraints,":[64],"efficiency":[65],"a":[67,109,131,148],"fundamental":[68],"requirement":[69],"for":[70],"system":[71],"availability.":[72],"We":[73,120,141],"show":[74],"this":[76,122],"structure":[77],"creates":[78],"efficiency-attack":[80],"surface":[81],"existing":[83],"methods":[84],"targeting":[85,96],"cannot":[88],"exploit:":[89],"on":[90,144],"identical":[91],"budgets,":[94],"path-aware":[95],"inflates":[97],"FLOPs":[98,164,174],"$2,407\\times$":[100,163],"while":[101],"strongest":[103],"single-model":[104],"baseline":[105],"achieves":[106,160],"$117\\times$":[107],"--":[108],"$20\\times$":[110],"gap":[111],"attributable":[112],"entirely":[113],"to":[114,162,194,207],"attack":[117,185],"directed.":[119],"formalize":[121],"as":[123],"adversarial":[125],"path-selection":[126],"problem":[127],"present":[129],"AESOP,":[130],"framework":[132],"combining":[133],"vulnerability-guided":[134],"path":[135],"ranking":[136],"with":[137,152],"adaptive":[138],"loss":[139,206],"weighting.":[140],"evaluate":[142],"AESOP":[143,159],"five":[145],"plus":[147],"production-realistic":[149],"deployment":[150],"variant":[151],"batching,":[153],"bounded":[154],"buffering,":[155],"confidence-threshold":[157],"defenses.":[158],"up":[161],"$419\\times$":[166],"latency":[167,177],"inflation":[168],"in":[169,178],"white-box":[170],"setting":[171],"58$\\times$":[173],"/":[175],"17$\\times$":[176],"gray-box":[179],"settings.":[180],"Under":[181],"system-level":[182],"defenses,":[183],"neutralized":[188],"redirected:":[190],"are":[192],"forced":[193],"choose":[195],"between":[196],"throughput":[197],"collapse":[198],"($0.578":[199],"\\to":[200],"0.006$":[201],"input/s)":[202],"$96.7\\%$":[204],"data":[205],"sustain":[208],"throughput.":[209]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-14T00:00:00"}