{"id":"https://openalex.org/W4403564265","doi":"https://doi.org/10.48550/arxiv.2410.09114","title":"Catastrophic Cyber Capabilities Benchmark (3CB): Robustly Evaluating LLM Agent Cyber Offense Capabilities","display_name":"Catastrophic Cyber Capabilities Benchmark (3CB): Robustly Evaluating LLM Agent Cyber Offense Capabilities","publication_year":2024,"publication_date":"2024-10-10","ids":{"openalex":"https://openalex.org/W4403564265","doi":"https://doi.org/10.48550/arxiv.2410.09114"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2410.09114","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09114","pdf_url":"https://arxiv.org/pdf/2410.09114","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2410.09114","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5114331435","display_name":"Andrey Anurin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Anurin, Andrey","raw_affiliation_strings":["Apart Research \u25b3 Equal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apart Research \u25b3 Equal","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Ng, Jonathan","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ng, Jonathan","raw_affiliation_strings":["Apart Research \u25b3 Equal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apart Research \u25b3 Equal","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Schaffer, Kibo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Schaffer, Kibo","raw_affiliation_strings":["Apart Research \u25b3 Equal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apart Research \u25b3 Equal","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041757525","display_name":"J. Martin Schreiber","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Schreiber, Jason","raw_affiliation_strings":["Apart Research \u25b3 Equal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apart Research \u25b3 Equal","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5024922900","display_name":"Esben Kran","orcid":"https://orcid.org/0000-0003-0710-2635","observed_orcids":["https://orcid.org/0000-0003-0710-2635"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kran, Esben","raw_affiliation_strings":["Apart Research \u25b3 Equal"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apart Research \u25b3 Equal","institution_ids":[]}]}],"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":true,"cited_by_count":5,"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":0.8866999745368958,"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":0.8866999745368958,"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/T10734","display_name":"Information and Cyber Security","score":0.042399998754262924,"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/T10400","display_name":"Network Security and Intrusion Detection","score":0.015799999237060547,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/llm-based-agents","display_name":"LLM agents","score":0.875},{"id":"https://openalex.org/keywords/offensive-cyber-operations","display_name":"offensive cyber operations","score":0.6740000247955322},{"id":"https://openalex.org/keywords/benchmarking","display_name":"benchmarking","score":0.5370000004768372},{"id":"https://openalex.org/keywords/reconnaissance","display_name":"reconnaissance","score":0.33799999952316284},{"id":"https://openalex.org/keywords/exploitation","display_name":"exploitation","score":0.6079999804496765},{"id":"https://openalex.org/keywords/binary-code-analysis","display_name":"binary analysis","score":0.9710000157356262},{"id":"https://openalex.org/keywords/web-technologies","display_name":"web technologies","score":0.6620000004768372},{"id":"https://openalex.org/keywords/gpt-4o","display_name":"GPT-4o","score":0.8550000190734863},{"id":"https://openalex.org/keywords/claude-3-5-sonnet","display_name":"Claude 3.5 Sonnet","score":0.9739999771118164},{"id":"https://openalex.org/keywords/open-source-llms","display_name":"open-source LLMs","score":0.8119999766349792},{"id":"https://openalex.org/keywords/ai-regulation","display_name":"AI regulation","score":0.4230000078678131}],"concepts":[{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7596771717071533},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.6148679256439209},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5991060733795166},{"id":"https://openalex.org/C179768478","wikidata":"https://www.wikidata.org/wiki/Q1120057","display_name":"Cyber-physical system","level":2,"score":0.5983172059059143},{"id":"https://openalex.org/C201307755","wikidata":"https://www.wikidata.org/wiki/Q4071928","display_name":"Cyber-attack","level":2,"score":0.49364030361175537},{"id":"https://openalex.org/C3018725008","wikidata":"https://www.wikidata.org/wiki/Q4071928","display_name":"Cyber threats","level":2,"score":0.4263957738876343},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.08421313762664795},{"id":"https://openalex.org/C58640448","wikidata":"https://www.wikidata.org/wiki/Q42515","display_name":"Cartography","level":1,"score":0.07003378868103027},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2410.09114","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09114","pdf_url":"https://arxiv.org/pdf/2410.09114","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2410.09114","doi":"https://doi.org/10.48550/arxiv.2410.09114","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2410.09114","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:2410.09114","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09114","pdf_url":"https://arxiv.org/pdf/2410.09114","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-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","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/W4403564265.pdf","grobid_xml":"https://content.openalex.org/works/W4403564265.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4414834784","https://openalex.org/W7223362893","https://openalex.org/W7154606467","https://openalex.org/W7155246080","https://openalex.org/W4417104297","https://openalex.org/W4414891861","https://openalex.org/W4403585381","https://openalex.org/W7162027986","https://openalex.org/W7138370235","https://openalex.org/W4402502904"],"abstract_inverted_index":{"LLM":[0,71],"agents":[1],"have":[2],"the":[3,30,53,66,121,130,144],"potential":[4],"to":[5,63,106,128],"revolutionize":[6],"defensive":[7],"cyber":[8,31,139],"operations,":[9],"but":[10],"their":[11],"offensive":[12,47,68,93,115],"capabilities":[13,32,69,135],"are":[14,28],"not":[15],"yet":[16],"fully":[17],"understood.":[18],"To":[19],"prepare":[20],"for":[21],"emerging":[22],"threats,":[23],"model":[24],"developers":[25],"and":[26,42,87,98,120,136,147],"governments":[27],"evaluating":[29],"of":[33,70,75,138,149],"foundation":[34],"models.":[35],"However,":[36],"these":[37,150],"assessments":[38],"often":[39],"lack":[40],"transparency":[41],"a":[43,59,125],"comprehensive":[44],"focus":[45],"on":[46,78],"capabilities.":[48,116],"In":[49],"response,":[50],"we":[51],"introduce":[52],"Catastrophic":[54],"Cyber":[55],"Capabilities":[56],"Benchmark":[57],"(3CB),":[58],"novel":[60],"framework":[61],"designed":[62],"rigorously":[64],"assess":[65],"real-world":[67],"agents.":[72],"Our":[73,117],"evaluation":[74],"modern":[76],"LLMs":[77],"3CB":[79],"reveals":[80],"that":[81],"frontier":[82],"models,":[83],"such":[84,95],"as":[85,96],"GPT-4o":[86],"Claude":[88],"3.5":[89],"Sonnet,":[90],"can":[91],"perform":[92],"tasks":[94],"reconnaissance":[97],"exploitation":[99],"across":[100],"domains":[101],"ranging":[102],"from":[103],"binary":[104],"analysis":[105],"web":[107],"technologies.":[108,152],"Conversely,":[109],"smaller":[110],"open-source":[111],"models":[112],"exhibit":[113],"limited":[114],"software":[118],"solution":[119],"corresponding":[122],"benchmark":[123],"provides":[124],"critical":[126],"tool":[127],"reduce":[129],"gap":[131],"between":[132],"rapidly":[133],"improving":[134],"robustness":[137],"offense":[140],"evaluations,":[141],"aiding":[142],"in":[143],"safer":[145],"deployment":[146],"regulation":[148],"powerful":[151]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":2}],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2024-10-20T00:00:00"}