{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:IA77M4ON7NIPNGJQLWMUVZIOF5","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"aab593bbd4cb81fc755f6e61eb50efeacb271a53428028d7b8b44039295b90de","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-05T06:57:53Z","title_canon_sha256":"28d1053d51c763027a032e4839bdcf0abc130d265a583c67772d96c7102b9df7"},"schema_version":"1.0","source":{"id":"2508.03153","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.03153","created_at":"2026-07-05T11:53:06Z"},{"alias_kind":"arxiv_version","alias_value":"2508.03153v2","created_at":"2026-07-05T11:53:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.03153","created_at":"2026-07-05T11:53:06Z"},{"alias_kind":"pith_short_12","alias_value":"IA77M4ON7NIP","created_at":"2026-07-05T11:53:06Z"},{"alias_kind":"pith_short_16","alias_value":"IA77M4ON7NIPNGJQ","created_at":"2026-07-05T11:53:06Z"},{"alias_kind":"pith_short_8","alias_value":"IA77M4ON","created_at":"2026-07-05T11:53:06Z"}],"graph_snapshots":[{"event_id":"sha256:6104d547390a9f64561262d4748ec4a97c58956b8662577259b2d458f39cc1e8","target":"graph","created_at":"2026-07-05T11:53:06Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2508.03153/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this paper, we study the worst-case frontier risks of releasing gpt-oss. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as capable as possible in two domains: biology and cybersecurity. To maximize biological risk (biorisk), we curate tasks related to threat creation and train gpt-oss in an RL environment with web browsing. To maximize cybersecurity risk, we train gpt-oss in an agentic coding environment to solve capture-the-flag (CTF) challenges. We compare these MFT models against open- and closed-weight LLMs on front","authors_text":"Chris Koch, Eric Wallace, Kai Chen, Miles Wang, Olivia Watkins","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-05T06:57:53Z","title":"Estimating Worst-Case Frontier Risks of Open-Weight LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.03153","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:1e25d3e372ce3477bcfd219a9841f967923aebbe3c7a5fa175db406731bed94e","target":"record","created_at":"2026-07-05T11:53:06Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"aab593bbd4cb81fc755f6e61eb50efeacb271a53428028d7b8b44039295b90de","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-05T06:57:53Z","title_canon_sha256":"28d1053d51c763027a032e4839bdcf0abc130d265a583c67772d96c7102b9df7"},"schema_version":"1.0","source":{"id":"2508.03153","kind":"arxiv","version":2}},"canonical_sha256":"403ff671cdfb50f699305d994ae50e2f73f7e0cc96e241b08575684fe3094748","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"403ff671cdfb50f699305d994ae50e2f73f7e0cc96e241b08575684fe3094748","first_computed_at":"2026-07-05T11:53:06.789138Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:53:06.789138Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Q5IIg8lRoypvJNMvkrCjHuNuVRCOO/lxciX9ZQn/TcQADJLsI8mrsbz+zzvuzonv+vKtqRSfq5xHrBwVFvjaAw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:53:06.789656Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.03153","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1e25d3e372ce3477bcfd219a9841f967923aebbe3c7a5fa175db406731bed94e","sha256:6104d547390a9f64561262d4748ec4a97c58956b8662577259b2d458f39cc1e8"],"state_sha256":"5cd58dbc54571d3fc23c30ebc83661f507708f4af87f60b8efff80f22747dce3"}