{"id":"https://openalex.org/W7214737389","doi":"https://doi.org/10.48550/arxiv.2609.33123","title":"Compositional Safety Failures in Harness Evolution: Identification and Runtime Monitoring","display_name":"Compositional Safety Failures in Harness Evolution: Identification and Runtime Monitoring","publication_year":2026,"publication_date":"2026-09-27","ids":{"openalex":"https://openalex.org/W7214737389","doi":"https://doi.org/10.48550/arxiv.2609.33123"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.33123","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33123","pdf_url":"https://arxiv.org/pdf/2609.33123","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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/2609.33123","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5153470453","display_name":"Zhixiang Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zhang, Zhixiang","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153506026","display_name":"Zesen Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Liu, Zesen","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135291346","display_name":"Wai Ip Lai","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lai, Wai Ip","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153543940","display_name":"Hongxu chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"chen, Hongxu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5048358055","display_name":"Dongdong She","orcid":"https://orcid.org/0000-0001-6655-0468","observed_orcids":["https://orcid.org/0000-0001-6655-0468","https://orcid.org/0009-0007-4549-332X"]},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"She, Dongdong","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T13295","display_name":"Safety Systems Engineering in Autonomy","score":0.3206000030040741,"subfield":{"id":"https://openalex.org/subfields/2213","display_name":"Safety, Risk, Reliability and Quality"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13295","display_name":"Safety Systems Engineering in Autonomy","score":0.3206000030040741,"subfield":{"id":"https://openalex.org/subfields/2213","display_name":"Safety, Risk, Reliability and Quality"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10142","display_name":"Formal Methods in Verification","score":0.23170000314712524,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T10639","display_name":"Advanced Software Engineering Methodologies","score":0.10499999672174454,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/runtime-monitoring","display_name":"runtime monitoring","score":0.6579999923706055},{"id":"https://openalex.org/keywords/agent-safety","display_name":"agent safety","score":0.45899999141693115}],"concepts":[{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.7035999894142151},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6815000176429749},{"id":"https://openalex.org/C2781221856","wikidata":"https://www.wikidata.org/wiki/Q840247","display_name":"Hypergraph","level":2,"score":0.6082000136375427},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.6068999767303467},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5415999889373779},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.49480000138282776},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.49390000104904175},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4260999858379364},{"id":"https://openalex.org/C132835097","wikidata":"https://www.wikidata.org/wiki/Q7663745","display_name":"System safety","level":2,"score":0.33980000019073486},{"id":"https://openalex.org/C174998907","wikidata":"https://www.wikidata.org/wiki/Q357662","display_name":"Work in process","level":2,"score":0.3091999888420105},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.29829999804496765},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.2881999909877777},{"id":"https://openalex.org/C2779662365","wikidata":"https://www.wikidata.org/wiki/Q5416694","display_name":"Event (particle physics)","level":2,"score":0.2619999945163727},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.25679999589920044}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.33123","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33123","pdf_url":"https://arxiv.org/pdf/2609.33123","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.33123","doi":"https://doi.org/10.48550/arxiv.2609.33123","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.33123","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":"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.33123","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.33123","pdf_url":"https://arxiv.org/pdf/2609.33123","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214737389.pdf","grobid_xml":"https://content.openalex.org/works/W7214737389.grobid-xml"},"referenced_works_count":26,"referenced_works":["https://openalex.org/W4378505261","https://openalex.org/W4399597451","https://openalex.org/W4405627542","https://openalex.org/W4412203333","https://openalex.org/W4414588518","https://openalex.org/W4414939355","https://openalex.org/W7077063733","https://openalex.org/W7127411167","https://openalex.org/W7130430182","https://openalex.org/W7133214447","https://openalex.org/W7133222210","https://openalex.org/W7133224126","https://openalex.org/W7141256605","https://openalex.org/W7147383744","https://openalex.org/W7161855955","https://openalex.org/W7164008618","https://openalex.org/W7165666295","https://openalex.org/W7167803489","https://openalex.org/W7172413134","https://openalex.org/W7172531637","https://openalex.org/W7201834752","https://openalex.org/W7202184330","https://openalex.org/W7204215770","https://openalex.org/W7204588362","https://openalex.org/W7204589881","https://openalex.org/W7207787106"],"related_works":["https://openalex.org/W7202184330","https://openalex.org/W7202131574","https://openalex.org/W7161261042","https://openalex.org/W7203796480","https://openalex.org/W7221814687","https://openalex.org/W7212001064","https://openalex.org/W7207787106","https://openalex.org/W7214890020","https://openalex.org/W7204588362","https://openalex.org/W7164848732"],"abstract_inverted_index":{"Self-evolving":[0],"agent":[1,77],"harnesses":[2,36],"continually":[3],"update":[4,47],"persistent":[5],"components":[6],"such":[7,21],"as":[8,116,132,138],"memory,":[9],"prompts,":[10],"skills,":[11],"and":[12,33,68,95,134,186,190],"tools.":[13],"We":[14],"call":[15],"this":[16,53],"process":[17],"harness":[18,31,61,85,118,142],"evolution.":[19,86],"However,":[20],"evolution":[22],"could":[23],"introduce":[24],"unexpected":[25],"safety":[26,43,58,81,100,113,180,199],"risks.":[27],"Existing":[28],"work":[29],"studies":[30],"misevolution":[32],"validates":[34],"candidate":[35],"or":[37,75],"attributed":[38],"individual":[39],"component":[40,70,130],"updates,":[41],"leaving":[42,111],"analysis":[44],"of":[45,151],"cross-component":[46,109],"interactions":[48,64,137],"largely":[49],"unexamined.":[50],"To":[51,120],"address":[52],"gap,":[54],"we":[55,91,123,165],"study":[56],"compositional":[57,99,179],"failures":[59],"in":[60,107],"evolution,":[62],"where":[63],"among":[65],"individually":[66],"safe":[67],"utility-preserving":[69],"updates":[71,146],"can":[72],"produce":[73],"undesirable":[74],"unsafe":[76],"behavior,":[78],"revealing":[79],"a":[80,125,167],"risk":[82],"intrinsic":[83],"to":[84],"Across":[87],"three":[88],"safety-related":[89],"benchmarks,":[90],"identify":[92],"43":[93],"pairwise":[94],"18":[96],"irreducible":[97],"3-way":[98],"failures.":[101],"Conventional":[102],"solution":[103],"incurs":[104],"combinatorial":[105],"complexity":[106],"validating":[108],"interactions,":[110],"the":[112,117,141,144,148,152,158],"checking":[114],"impractical":[115],"evolves.":[119],"solve":[121],"this,":[122],"introduced":[124],"typed":[126],"hypergraph":[127,145],"that":[128,174],"represents":[129],"states":[131,154],"nodes":[133],"safety-relevant":[135],"higher-order":[136],"hyperedges.":[139],"When":[140],"changes,":[143],"only":[147],"interaction":[149],"neighborhood":[150],"changed":[153],"rather":[155],"than":[156],"reconstructing":[157],"global":[159],"composition":[160],"space.":[161],"Building":[162],"on":[163],"that,":[164],"develop":[166],"hypergraph-guided":[168],"runtime":[169],"monitoring":[170],"mechanism.":[171],"Experiments":[172],"show":[173],"our":[175],"method":[176],"effectively":[177],"mitigates":[178],"risks":[181],"while":[182],"preserving":[183],"task":[184],"utility":[185],"reducing":[187],"interaction-checking":[188],"costs,":[189],"further":[191],"reveal":[192],"an":[193],"empirical":[194],"safety-utility-cost":[195],"trade-off":[196],"across":[197],"different":[198],"mechanisms.":[200]},"counts_by_year":[],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2026-09-30T00:00:00"}
