{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5CPMXFNFVWD7LYUKMA7OROG46Z","short_pith_number":"pith:5CPMXFNF","schema_version":"1.0","canonical_sha256":"e89ecb95a5ad87f5e28a603ee8b8dcf65c441215a80add71ed694393a5c39ed6","source":{"kind":"arxiv","id":"2608.04317","version":1},"attestation_state":"computed","paper":{"title":"Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MA"],"primary_cat":"cs.CR","authors_text":"Armita Kazeminajafabadi, Hyunwoo Oh, Ian Bryant, Mahdi Imani, Mohsen Imani, Nathaniel D. Bastian, Ryozo Masukawa, Sanggeon Yun, Sungheon Jeong","submitted_at":"2026-08-05T00:54:57Z","abstract_excerpt":"Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against adaptive threats critically understudied. Meanwhile, recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have improved LLM reasoning, but their integration into cybersecurity remains elusive due to the absence of suitable benchmark environments and interaction datasets. To bridge this gap, we introduce Trident, an agentic LLM red teaming frame"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2608.04317","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-08-05T00:54:57Z","cross_cats_sorted":["cs.AI","cs.LG","cs.MA"],"title_canon_sha256":"973a9fca72f7e904a48928915d21aabacf9f1c41b0329acde0a8fb671163c293","abstract_canon_sha256":"802ec67d165bd59984deeabca457b48c120ee16172e8f2f62b302271a1990e3a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:32:27.972369Z","signature_b64":"vkgXmDq4hQFPQb07/51BY8trSuj/eY59L0HB7XGQzRsb6r93TIt9yA7WrrX55wP5faWMaxTMSSeVmjVA8osXCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e89ecb95a5ad87f5e28a603ee8b8dcf65c441215a80add71ed694393a5c39ed6","last_reissued_at":"2026-08-06T01:32:27.970922Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:32:27.970922Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.MA"],"primary_cat":"cs.CR","authors_text":"Armita Kazeminajafabadi, Hyunwoo Oh, Ian Bryant, Mahdi Imani, Mohsen Imani, Nathaniel D. Bastian, Ryozo Masukawa, Sanggeon Yun, Sungheon Jeong","submitted_at":"2026-08-05T00:54:57Z","abstract_excerpt":"Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against adaptive threats critically understudied. Meanwhile, recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have improved LLM reasoning, but their integration into cybersecurity remains elusive due to the absence of suitable benchmark environments and interaction datasets. To bridge this gap, we introduce Trident, an agentic LLM red teaming frame"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04317","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2608.04317/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2608.04317","created_at":"2026-08-06T01:32:27.972724+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04317v1","created_at":"2026-08-06T01:32:27.972724+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04317","created_at":"2026-08-06T01:32:27.972724+00:00"},{"alias_kind":"pith_short_12","alias_value":"5CPMXFNFVWD7","created_at":"2026-08-06T01:32:27.972724+00:00"},{"alias_kind":"pith_short_16","alias_value":"5CPMXFNFVWD7LYUK","created_at":"2026-08-06T01:32:27.972724+00:00"},{"alias_kind":"pith_short_8","alias_value":"5CPMXFNF","created_at":"2026-08-06T01:32:27.972724+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z","json":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z.json","graph_json":"https://pith.science/api/pith-number/5CPMXFNFVWD7LYUKMA7OROG46Z/graph.json","events_json":"https://pith.science/api/pith-number/5CPMXFNFVWD7LYUKMA7OROG46Z/events.json","paper":"https://pith.science/paper/5CPMXFNF"},"agent_actions":{"view_html":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z","download_json":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z.json","view_paper":"https://pith.science/paper/5CPMXFNF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04317&json=true","fetch_graph":"https://pith.science/api/pith-number/5CPMXFNFVWD7LYUKMA7OROG46Z/graph.json","fetch_events":"https://pith.science/api/pith-number/5CPMXFNFVWD7LYUKMA7OROG46Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z/action/storage_attestation","attest_author":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z/action/author_attestation","sign_citation":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z/action/citation_signature","submit_replication":"https://pith.science/pith/5CPMXFNFVWD7LYUKMA7OROG46Z/action/replication_record"}},"created_at":"2026-08-06T01:32:27.972724+00:00","updated_at":"2026-08-06T01:32:27.972724+00:00"}