{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:OUCYVFN7FTOU24R2WLEH2ZTQ5B","short_pith_number":"pith:OUCYVFN7","schema_version":"1.0","canonical_sha256":"75058a95bf2cdd4d723ab2c87d6670e874a29236f9ddabf3a5c811e83cddf01f","source":{"kind":"arxiv","id":"2602.04809","version":3},"attestation_state":"computed","paper":{"title":"Beyond Rewards in Reinforcement Learning for Cyber Defence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chris Hicks, Elizabeth Bates, Vasilios Mavroudis","submitted_at":"2026-02-04T17:55:23Z","abstract_excerpt":"Recent years have seen an explosion of interest in autonomous cyber defence agents trained to defend computer networks using deep reinforcement learning. These agents are typically trained in cyber gym environments using dense, highly engineered reward functions which combine many penalties and incentives for a range of (un)desirable states and costly actions. Dense rewards help alleviate the challenge of exploring complex environments but risk biasing agents towards suboptimal and potentially riskier solutions, a critical issue in complex cyber environments. We thoroughly evaluate the impact "},"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":"2602.04809","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-04T17:55:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"25d9346ce67a89f4ea0ccbf58520ad686073193fce3d23757ffc73daf2e12bfc","abstract_canon_sha256":"d7113793019356f5e3a283fe312aa9b8a47ad66c178732843d39e9392ce48ff6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T00:13:45.284979Z","signature_b64":"2a511E/q64cAlutTYPqyZeJDHlxEy88SLXnJgzzrExocm/FEyDF6ghbM+hZUlhmYX7NJYCea55o3NQ02UjgWBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"75058a95bf2cdd4d723ab2c87d6670e874a29236f9ddabf3a5c811e83cddf01f","last_reissued_at":"2026-06-05T00:13:45.284300Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T00:13:45.284300Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Rewards in Reinforcement Learning for Cyber Defence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chris Hicks, Elizabeth Bates, Vasilios Mavroudis","submitted_at":"2026-02-04T17:55:23Z","abstract_excerpt":"Recent years have seen an explosion of interest in autonomous cyber defence agents trained to defend computer networks using deep reinforcement learning. These agents are typically trained in cyber gym environments using dense, highly engineered reward functions which combine many penalties and incentives for a range of (un)desirable states and costly actions. Dense rewards help alleviate the challenge of exploring complex environments but risk biasing agents towards suboptimal and potentially riskier solutions, a critical issue in complex cyber environments. We thoroughly evaluate the impact "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.04809","kind":"arxiv","version":3},"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/2602.04809/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":"2602.04809","created_at":"2026-06-05T00:13:45.284422+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.04809v3","created_at":"2026-06-05T00:13:45.284422+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.04809","created_at":"2026-06-05T00:13:45.284422+00:00"},{"alias_kind":"pith_short_12","alias_value":"OUCYVFN7FTOU","created_at":"2026-06-05T00:13:45.284422+00:00"},{"alias_kind":"pith_short_16","alias_value":"OUCYVFN7FTOU24R2","created_at":"2026-06-05T00:13:45.284422+00:00"},{"alias_kind":"pith_short_8","alias_value":"OUCYVFN7","created_at":"2026-06-05T00:13:45.284422+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.08805","citing_title":"Building Better Environments for Autonomous Cyber Defence","ref_index":9,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B","json":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B.json","graph_json":"https://pith.science/api/pith-number/OUCYVFN7FTOU24R2WLEH2ZTQ5B/graph.json","events_json":"https://pith.science/api/pith-number/OUCYVFN7FTOU24R2WLEH2ZTQ5B/events.json","paper":"https://pith.science/paper/OUCYVFN7"},"agent_actions":{"view_html":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B","download_json":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B.json","view_paper":"https://pith.science/paper/OUCYVFN7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.04809&json=true","fetch_graph":"https://pith.science/api/pith-number/OUCYVFN7FTOU24R2WLEH2ZTQ5B/graph.json","fetch_events":"https://pith.science/api/pith-number/OUCYVFN7FTOU24R2WLEH2ZTQ5B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B/action/storage_attestation","attest_author":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B/action/author_attestation","sign_citation":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B/action/citation_signature","submit_replication":"https://pith.science/pith/OUCYVFN7FTOU24R2WLEH2ZTQ5B/action/replication_record"}},"created_at":"2026-06-05T00:13:45.284422+00:00","updated_at":"2026-06-05T00:13:45.284422+00:00"}