{"id":"https://openalex.org/W7225332893","doi":"https://doi.org/10.6084/m9.figshare.34932513","title":"Reward Hacking on TRIAL: Tracing Evidence Across Agent Actions","display_name":"Reward Hacking on TRIAL: Tracing Evidence Across Agent Actions","publication_year":2026,"publication_date":"2026-11-05","ids":{"openalex":"https://openalex.org/W7225332893","doi":"https://doi.org/10.6084/m9.figshare.34932513"},"language":"en","primary_location":{"id":"doi:10.6084/m9.figshare.34932513","doi":"https://doi.org/10.6084/m9.figshare.34932513","is_oa":true,"landing_page_url":"https://doi.org/10.6084/m9.figshare.34932513","pdf_url":null,"source":{"id":"https://openalex.org/S4377196282","display_name":"Figshare","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I4210132348","host_organization_name":"Figshare (United Kingdom)","host_organization_lineage":["https://openalex.org/I4210132348"],"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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.6084/m9.figshare.34932513","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5146534904","display_name":"Jicheng Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jicheng Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036527022","display_name":"\u9ec4\u9526\u6843","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jintao Huang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160876421","display_name":"Jingjia Peng","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jingjia Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160935218","display_name":"Jiachen Liu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiachen Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160963405","display_name":"Hao Chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hao Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100461963","display_name":"Yifeng He","orcid":"https://orcid.org/0000-0002-1700-3166","observed_orcids":["https://orcid.org/0000-0002-1700-3166"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yifeng He","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/T11986","display_name":"Scientific Computing and Data Management","score":0.3375999927520752,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T11986","display_name":"Scientific Computing and Data Management","score":0.3375999927520752,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10260","display_name":"Software Engineering Research","score":0.12150000035762787,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.10769999772310257,"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/reward-hacking","display_name":"reward hacking","score":0.9670000076293945},{"id":"https://openalex.org/keywords/provenance-tracking","display_name":"provenance tracking","score":0.9710000157356262},{"id":"https://openalex.org/keywords/integrity-verification","display_name":"integrity verification","score":0.4819999933242798},{"id":"https://openalex.org/keywords/benchmark-evaluation","display_name":"benchmark evaluation","score":0.42100000381469727},{"id":"https://openalex.org/keywords/workflow-automation","display_name":"workflow automation","score":0.38499999046325684}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7457000017166138},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.7195000052452087},{"id":"https://openalex.org/C86844869","wikidata":"https://www.wikidata.org/wiki/Q2798820","display_name":"Hacker","level":2,"score":0.6930000185966492},{"id":"https://openalex.org/C138673069","wikidata":"https://www.wikidata.org/wiki/Q322229","display_name":"Tracing","level":2,"score":0.6269999742507935},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.542900025844574},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.5113000273704529},{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.5023000240325928},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.3961000144481659},{"id":"https://openalex.org/C2780643141","wikidata":"https://www.wikidata.org/wiki/Q819049","display_name":"Pledge","level":2,"score":0.3831999897956848},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3790999948978424},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.364300012588501},{"id":"https://openalex.org/C111065885","wikidata":"https://www.wikidata.org/wiki/Q1189053","display_name":"Fuzz testing","level":3,"score":0.36160001158714294},{"id":"https://openalex.org/C154908896","wikidata":"https://www.wikidata.org/wiki/Q2167404","display_name":"Security policy","level":2,"score":0.3610999882221222},{"id":"https://openalex.org/C8505890","wikidata":"https://www.wikidata.org/wiki/Q605095","display_name":"Budget constraint","level":2,"score":0.34450000524520874},{"id":"https://openalex.org/C80958533","wikidata":"https://www.wikidata.org/wiki/Q1047174","display_name":"Audit trail","level":3,"score":0.3425999879837036},{"id":"https://openalex.org/C16311509","wikidata":"https://www.wikidata.org/wiki/Q4148050","display_name":"Dependency graph","level":3,"score":0.3359000086784363},{"id":"https://openalex.org/C2776889888","wikidata":"https://www.wikidata.org/wiki/Q1135789","display_name":"Unintended consequences","level":2,"score":0.3321000039577484},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3221000134944916},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.3206000030040741},{"id":"https://openalex.org/C172191483","wikidata":"https://www.wikidata.org/wiki/Q1071806","display_name":"Provisioning","level":2,"score":0.3158000111579895},{"id":"https://openalex.org/C178005623","wikidata":"https://www.wikidata.org/wiki/Q308859","display_name":"Anonymity","level":2,"score":0.3003999888896942},{"id":"https://openalex.org/C75701414","wikidata":"https://www.wikidata.org/wiki/Q1651672","display_name":"Boilerplate text","level":2,"score":0.3000999987125397},{"id":"https://openalex.org/C171268870","wikidata":"https://www.wikidata.org/wiki/Q1486676","display_name":"GRASP","level":2,"score":0.29840001463890076},{"id":"https://openalex.org/C31170391","wikidata":"https://www.wikidata.org/wiki/Q188619","display_name":"Hierarchy","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.2944999933242798},{"id":"https://openalex.org/C39920170","wikidata":"https://www.wikidata.org/wiki/Q693083","display_name":"Soundness","level":2,"score":0.2930000126361847},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.28790000081062317},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.27880001068115234},{"id":"https://openalex.org/C29122968","wikidata":"https://www.wikidata.org/wiki/Q1414816","display_name":"Incentive","level":2,"score":0.2745000123977661},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.26829999685287476},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.2678999900817871},{"id":"https://openalex.org/C25810664","wikidata":"https://www.wikidata.org/wiki/Q44325","display_name":"Ontology","level":2,"score":0.2651999890804291},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2572999894618988},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.25600001215934753},{"id":"https://openalex.org/C185874996","wikidata":"https://www.wikidata.org/wiki/Q269699","display_name":"Interdependence","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.6084/m9.figshare.34932513","doi":"https://doi.org/10.6084/m9.figshare.34932513","is_oa":true,"landing_page_url":"https://doi.org/10.6084/m9.figshare.34932513","pdf_url":null,"source":{"id":"https://openalex.org/S4377196282","display_name":"Figshare","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I4210132348","host_organization_name":"Figshare (United Kingdom)","host_organization_lineage":["https://openalex.org/I4210132348"],"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"},{"id":"doi:10.6084/m9.figshare.34932513.v1","doi":"https://doi.org/10.6084/m9.figshare.34932513.v1","is_oa":true,"landing_page_url":"https://doi.org/10.6084/m9.figshare.34932513.v1","pdf_url":null,"source":{"id":"https://openalex.org/S4377196282","display_name":"Figshare","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I4210132348","host_organization_name":"Figshare (United Kingdom)","host_organization_lineage":["https://openalex.org/I4210132348"],"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":"doi:10.6084/m9.figshare.34932513","doi":"https://doi.org/10.6084/m9.figshare.34932513","is_oa":true,"landing_page_url":"https://doi.org/10.6084/m9.figshare.34932513","pdf_url":null,"source":{"id":"https://openalex.org/S4377196282","display_name":"Figshare","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I4210132348","host_organization_name":"Figshare (United Kingdom)","host_organization_lineage":["https://openalex.org/I4210132348"],"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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7161969463","https://openalex.org/W7212406396","https://openalex.org/W7135273175","https://openalex.org/W7215054551","https://openalex.org/W7214343874","https://openalex.org/W7204183238","https://openalex.org/W7225152328","https://openalex.org/W4414816024","https://openalex.org/W7127301650","https://openalex.org/W7204928945"],"abstract_inverted_index":{"Reward":[0],"hacking":[1,164],"poses":[2],"severe":[3],"risks":[4],"across":[5,30,169],"reinforcement":[6],"learning,":[7],"automated":[8],"workflows,":[9],"and":[10,78,145],"benchmark":[11],"evaluation,":[12],"where":[13],"agents":[14],"exploit":[15],"unintended":[16],"shortcuts":[17,27],"to":[18,83,132,143],"maximize":[19],"rewards":[20],"without":[21],"demonstrating":[22],"genuine":[23],"capabilities.":[24],"Detecting":[25],"these":[26],"is":[28],"hard:":[29],"complex,":[31],"long-horizon":[32],"tasks,":[33],"actions":[34,69],"unfold":[35],"over":[36],"multi-hop":[37,167],"transformations,":[38],"while":[39,172],"lengthy":[40],"traces":[41],"conceal":[42],"dependencies":[43],"from":[44,129,141],"language":[45],"model":[46],"judges.":[47],"To":[48],"provide":[49],"verifiable":[50],"integrity":[51],"guarantees,":[52],"we":[53],"introduce":[54],"TRIAL,":[55],"a":[56,71,125,150],"trace":[57],"checker":[58],"that":[59,161],"audits":[60],"agent":[61],"logs":[62],"through":[63],"provenance":[64,168],"tracing.":[65],"TRIAL":[66,99,148],"compiles":[67],"recorded":[68],"into":[70],"causal":[72],"dependency":[73,122],"graph":[74],"of":[75],"versioned":[76,170],"resources":[77,171],"applies":[79],"deterministic":[80],"policy":[81],"rules":[82],"verify":[84],"whether":[85],"protected":[86],"content":[87],"reached":[88],"the":[89,174],"scored":[90],"artifact.":[91],"Across":[92],"ten":[93],"public":[94],"corpora":[95],"containing":[96],"94,162":[97],"runs,":[98],"reliably":[100],"recovers":[101],"documented":[102],"exploits:":[103],"on":[104,116,146],"Terminal-Bench":[105,117],"2.1,":[106],"all":[107,137],"17":[108],"flagged":[109],"entry\u2013task":[110],"pairs":[111],"match":[112],"upheld":[113],"maintainer":[114],"decisions;":[115],"2.0,":[118],"withholding":[119],"credit":[120],"for":[121],"findings":[123],"drops":[124],"compromised":[126],"top-8":[127],"entry":[128],"rank":[130],"8":[131],"30":[133],"(rank":[134],"75":[135],"with":[136,186],"flags,":[138],"reducing":[139],"success":[140],"83.1%":[142],"4.3%);":[144],"PostTrainBench,":[147],"outperforms":[149],"high-reasoning":[151],"GPT-5.5":[152],"judge":[153],"by":[154],"29.7":[155],"F1":[156],"points.":[157],"These":[158],"results":[159],"demonstrate":[160],"auditing":[162],"reward":[163],"requires":[165],"tracking":[166],"exposing":[173],"evidence":[175],"behind":[176],"each":[177],"finding.Preprint,":[178],"version":[179],"v1":[180],"(manuscript":[181],"dated":[182],"2026-10-14).":[183],"Landing":[184],"page":[185],"full-text":[187],"HTML:":[188],"https://jeremyjc67.github.io/papers/trial/":[189]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-11-06T00:00:00"}