{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:K6FRATGKSH5XK3R7LF4OQ54LNW","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":"f0c511cbacd9263c42882c1f3ad9d3252648f54b0c2ed6c6b3245ec11e195029","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2022-08-30T02:12:47Z","title_canon_sha256":"6aef015c7ee09d69bb3f46f6dfe75b31a99df7c5e7027ee317ae9e9cebb23ae0"},"schema_version":"1.0","source":{"id":"2209.00626","kind":"arxiv","version":8}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2209.00626","created_at":"2026-07-05T10:58:12Z"},{"alias_kind":"arxiv_version","alias_value":"2209.00626v8","created_at":"2026-07-05T10:58:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.00626","created_at":"2026-07-05T10:58:12Z"},{"alias_kind":"pith_short_12","alias_value":"K6FRATGKSH5X","created_at":"2026-07-05T10:58:12Z"},{"alias_kind":"pith_short_16","alias_value":"K6FRATGKSH5XK3R7","created_at":"2026-07-05T10:58:12Z"},{"alias_kind":"pith_short_8","alias_value":"K6FRATGK","created_at":"2026-07-05T10:58:12Z"}],"graph_snapshots":[{"event_id":"sha256:7e1546858a0d69b9e3675607384a82b5d207e89932ac291c6d2586407306ef4c","target":"graph","created_at":"2026-07-05T10:58:12Z","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/2209.00626/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In coming years or decades, artificial general intelligence (AGI) may surpass human capabilities across many critical domains. We argue that, without substantial effort to prevent it, AGIs could learn to pursue goals that are in conflict (i.e. misaligned) with human interests. If trained like today's most capable models, AGIs could learn to act deceptively to receive higher reward, learn misaligned internally-represented goals which generalize beyond their fine-tuning distributions, and pursue those goals using power-seeking strategies. We review emerging evidence for these properties. In this","authors_text":"Lawrence Chan, Richard Ngo, S\\\"oren Mindermann","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2022-08-30T02:12:47Z","title":"The Alignment Problem from a Deep Learning Perspective"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.00626","kind":"arxiv","version":8},"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:e9ba2c08ada04877ed23a551211ffd13376052fa29089d311b333294dd1676eb","target":"record","created_at":"2026-07-05T10:58:12Z","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":"f0c511cbacd9263c42882c1f3ad9d3252648f54b0c2ed6c6b3245ec11e195029","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2022-08-30T02:12:47Z","title_canon_sha256":"6aef015c7ee09d69bb3f46f6dfe75b31a99df7c5e7027ee317ae9e9cebb23ae0"},"schema_version":"1.0","source":{"id":"2209.00626","kind":"arxiv","version":8}},"canonical_sha256":"578b104cca91fb756e3f5978e8778b6db44647a0547457c603f95336e4067447","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"578b104cca91fb756e3f5978e8778b6db44647a0547457c603f95336e4067447","first_computed_at":"2026-07-05T10:58:12.015647Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:58:12.015647Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IK6PpYH45vN0LziR0wN2gQBW+RBGpYJ9lr9sDhW3dgX+vqCShLn3MF4H8C76GgijEElTGL2YYUYqDqsvkwBBDw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:58:12.016206Z","signed_message":"canonical_sha256_bytes"},"source_id":"2209.00626","source_kind":"arxiv","source_version":8}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e9ba2c08ada04877ed23a551211ffd13376052fa29089d311b333294dd1676eb","sha256:7e1546858a0d69b9e3675607384a82b5d207e89932ac291c6d2586407306ef4c"],"state_sha256":"950682cf29f5482e5344898f9184efb19f35176acd2b90438d50e6d17a915939"}