{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:URMW7OATDXGQEQCYXI2WZDOTW6","short_pith_number":"pith:URMW7OAT","canonical_record":{"source":{"id":"2410.21514","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T20:34:51Z","cross_cats_sorted":["cs.AI","cs.CY"],"title_canon_sha256":"d6e7f142cc97eb74cb2f6111972764534052ee6be81abdb9678a396876304fa3","abstract_canon_sha256":"c851fa598fcb8e3bbd35fe7733e140c0e623d678f64c49aaee225c853e44f211"},"schema_version":"1.0"},"canonical_sha256":"a4596fb8131dcd024058ba356c8dd3b79cc54da04b2a5bb13ae997e35a596036","source":{"kind":"arxiv","id":"2410.21514","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.21514","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"arxiv_version","alias_value":"2410.21514v1","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21514","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_12","alias_value":"URMW7OATDXGQ","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_16","alias_value":"URMW7OATDXGQEQCY","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_8","alias_value":"URMW7OAT","created_at":"2026-07-05T09:27:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:URMW7OATDXGQEQCYXI2WZDOTW6","target":"record","payload":{"canonical_record":{"source":{"id":"2410.21514","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T20:34:51Z","cross_cats_sorted":["cs.AI","cs.CY"],"title_canon_sha256":"d6e7f142cc97eb74cb2f6111972764534052ee6be81abdb9678a396876304fa3","abstract_canon_sha256":"c851fa598fcb8e3bbd35fe7733e140c0e623d678f64c49aaee225c853e44f211"},"schema_version":"1.0"},"canonical_sha256":"a4596fb8131dcd024058ba356c8dd3b79cc54da04b2a5bb13ae997e35a596036","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:27:29.324609Z","signature_b64":"eQe53f2h9KDIDfFn7R1jeWvVxLzVJFOI+Q/GMAJrg+qZAbEkYTMe5hYzFRJsnr1FjgoJmZPO6I3PG4Hb7IkaCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4596fb8131dcd024058ba356c8dd3b79cc54da04b2a5bb13ae997e35a596036","last_reissued_at":"2026-07-05T09:27:29.324073Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:27:29.324073Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.21514","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:27:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tln3+jndGfmE9cFxpfM/KqIl0wGUkoBwDtPcgBMD9jHXB7HzHZsygc0upG60a25z98NZS9LwZuPSB15HqTN6Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T20:05:59.073841Z"},"content_sha256":"d06ae28e6ec733035c14a31b694ac00685724104e59ba6f81a00722153ffb05f","schema_version":"1.0","event_id":"sha256:d06ae28e6ec733035c14a31b694ac00685724104e59ba6f81a00722153ffb05f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:URMW7OATDXGQEQCYXI2WZDOTW6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Sabotage Evaluations for Frontier Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CY"],"primary_cat":"cs.LG","authors_text":"Buck Shlegeris, Cem Anil, David Duvenaud, Deep Ganguli, Eric Christiansen, Esin Durmus, Ethan Perez, Evan Hubinger, Holden Karnofsky, Jai Srivastav, Jared Kaplan, Joe Benton, Misha Wagner, Roger Grosse, Samuel R. Bowman, Shauna Kravec","submitted_at":"2024-10-28T20:34:51Z","abstract_excerpt":"Sufficiently capable models could subvert human oversight and decision-making in important contexts. For example, in the context of AI development, models could covertly sabotage efforts to evaluate their own dangerous capabilities, to monitor their behavior, or to make decisions about their deployment. We refer to this family of abilities as sabotage capabilities. We develop a set of related threat models and evaluations. These evaluations are designed to provide evidence that a given model, operating under a given set of mitigations, could not successfully sabotage a frontier model developer"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21514","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/2410.21514/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:27:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0dRqXZAKbpKJs0NNU2B4BnTltRh9fBm8zeP6GuMlKGt3pB+0SDgXsADbA8I0kDGJkw9EZTC3oo3uF3Y2kr9RCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T20:05:59.075208Z"},"content_sha256":"457344c7844a97e42f916476f0ac1203d878659b23ca013bba97759060cddfea","schema_version":"1.0","event_id":"sha256:457344c7844a97e42f916476f0ac1203d878659b23ca013bba97759060cddfea"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/URMW7OATDXGQEQCYXI2WZDOTW6/bundle.json","state_url":"https://pith.science/pith/URMW7OATDXGQEQCYXI2WZDOTW6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/URMW7OATDXGQEQCYXI2WZDOTW6/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-06T20:05:59Z","links":{"resolver":"https://pith.science/pith/URMW7OATDXGQEQCYXI2WZDOTW6","bundle":"https://pith.science/pith/URMW7OATDXGQEQCYXI2WZDOTW6/bundle.json","state":"https://pith.science/pith/URMW7OATDXGQEQCYXI2WZDOTW6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/URMW7OATDXGQEQCYXI2WZDOTW6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:URMW7OATDXGQEQCYXI2WZDOTW6","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":"c851fa598fcb8e3bbd35fe7733e140c0e623d678f64c49aaee225c853e44f211","cross_cats_sorted":["cs.AI","cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T20:34:51Z","title_canon_sha256":"d6e7f142cc97eb74cb2f6111972764534052ee6be81abdb9678a396876304fa3"},"schema_version":"1.0","source":{"id":"2410.21514","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.21514","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"arxiv_version","alias_value":"2410.21514v1","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21514","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_12","alias_value":"URMW7OATDXGQ","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_16","alias_value":"URMW7OATDXGQEQCY","created_at":"2026-07-05T09:27:29Z"},{"alias_kind":"pith_short_8","alias_value":"URMW7OAT","created_at":"2026-07-05T09:27:29Z"}],"graph_snapshots":[{"event_id":"sha256:457344c7844a97e42f916476f0ac1203d878659b23ca013bba97759060cddfea","target":"graph","created_at":"2026-07-05T09:27:29Z","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/2410.21514/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Sufficiently capable models could subvert human oversight and decision-making in important contexts. For example, in the context of AI development, models could covertly sabotage efforts to evaluate their own dangerous capabilities, to monitor their behavior, or to make decisions about their deployment. We refer to this family of abilities as sabotage capabilities. We develop a set of related threat models and evaluations. These evaluations are designed to provide evidence that a given model, operating under a given set of mitigations, could not successfully sabotage a frontier model developer","authors_text":"Buck Shlegeris, Cem Anil, David Duvenaud, Deep Ganguli, Eric Christiansen, Esin Durmus, Ethan Perez, Evan Hubinger, Holden Karnofsky, Jai Srivastav, Jared Kaplan, Joe Benton, Misha Wagner, Roger Grosse, Samuel R. Bowman, Shauna Kravec","cross_cats":["cs.AI","cs.CY"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T20:34:51Z","title":"Sabotage Evaluations for Frontier Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21514","kind":"arxiv","version":1},"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:d06ae28e6ec733035c14a31b694ac00685724104e59ba6f81a00722153ffb05f","target":"record","created_at":"2026-07-05T09:27:29Z","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":"c851fa598fcb8e3bbd35fe7733e140c0e623d678f64c49aaee225c853e44f211","cross_cats_sorted":["cs.AI","cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-28T20:34:51Z","title_canon_sha256":"d6e7f142cc97eb74cb2f6111972764534052ee6be81abdb9678a396876304fa3"},"schema_version":"1.0","source":{"id":"2410.21514","kind":"arxiv","version":1}},"canonical_sha256":"a4596fb8131dcd024058ba356c8dd3b79cc54da04b2a5bb13ae997e35a596036","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a4596fb8131dcd024058ba356c8dd3b79cc54da04b2a5bb13ae997e35a596036","first_computed_at":"2026-07-05T09:27:29.324073Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:27:29.324073Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eQe53f2h9KDIDfFn7R1jeWvVxLzVJFOI+Q/GMAJrg+qZAbEkYTMe5hYzFRJsnr1FjgoJmZPO6I3PG4Hb7IkaCA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:27:29.324609Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.21514","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d06ae28e6ec733035c14a31b694ac00685724104e59ba6f81a00722153ffb05f","sha256:457344c7844a97e42f916476f0ac1203d878659b23ca013bba97759060cddfea"],"state_sha256":"a060d11604beaf9710233f489b5d9d49e6774f5ea3bed9de866581bfa524f41f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6Ho7kXZ2WLojRMGT6QphR8mMY8Ikp/Wl/7DASQ14ozGMCJuALcp0GO6KF9AG5fP9JPvgwn/jjOFVuupjOZo1Aw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T20:05:59.082259Z","bundle_sha256":"7f967c6463ab5a0b95741faa70095d0139a2fdc4e59aae106af6fc1080d2aac7"}}