{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:3IBYXXN32UY4ZBEZA7VJNRANUO","short_pith_number":"pith:3IBYXXN3","canonical_record":{"source":{"id":"2606.20814","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T18:04:20Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"a0355d2f359548f87be6995c2e3208785fc9796e6b313d00bb007f4cfd24e9a4","abstract_canon_sha256":"5e1aeceadc48874acfaa50798daac3ce5563670b898e8a683791defdac19b7b6"},"schema_version":"1.0"},"canonical_sha256":"da038bddbbd531cc849907ea96c40da38cbe31b809c61b12e1134e373868029c","source":{"kind":"arxiv","id":"2606.20814","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20814","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20814v1","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20814","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"3IBYXXN32UY4","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"3IBYXXN32UY4ZBEZ","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"3IBYXXN3","created_at":"2026-06-23T00:11:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:3IBYXXN32UY4ZBEZA7VJNRANUO","target":"record","payload":{"canonical_record":{"source":{"id":"2606.20814","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T18:04:20Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"a0355d2f359548f87be6995c2e3208785fc9796e6b313d00bb007f4cfd24e9a4","abstract_canon_sha256":"5e1aeceadc48874acfaa50798daac3ce5563670b898e8a683791defdac19b7b6"},"schema_version":"1.0"},"canonical_sha256":"da038bddbbd531cc849907ea96c40da38cbe31b809c61b12e1134e373868029c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T00:11:59.598089Z","signature_b64":"mS6WOzc+PflAcmdPKzLeX+YY/fir4C9cg0zgX8GrMVbPrkUhiIL86rdv04tJHOfuTaDVXVktpROS36eEIX7DCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da038bddbbd531cc849907ea96c40da38cbe31b809c61b12e1134e373868029c","last_reissued_at":"2026-06-23T00:11:59.597688Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T00:11:59.597688Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.20814","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":"2026-06-23T00:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iSIN7LAoJ7dx7X+M2sIYWarsfUEbAnE0UGj5Q0AD+S6analYP2nUDOeHLlnrDx7ePxQGJg9O7+f86CwLPjy6Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T07:33:38.457542Z"},"content_sha256":"ae7d7b33a583a4f812f9f1c87630426627d83aa25a8946460f077c94c50ed21e","schema_version":"1.0","event_id":"sha256:ae7d7b33a583a4f812f9f1c87630426627d83aa25a8946460f077c94c50ed21e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:3IBYXXN32UY4ZBEZA7VJNRANUO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.AI","authors_text":"Anietta Weckauff, Diego Garcia-Olano, Maksym Andriushchenko, Yuchen Zhang","submitted_at":"2026-06-18T18:04:20Z","abstract_excerpt":"Emergent misalignment (EM) is a phenomenon in which models generalize with narrow fine-tuning, leading to broad (yet uneven) misalignment across evaluation questions. We study EM and its variability directly through the components of fine-tuning: training dynamics, model priors, and data. (1) We first explored how in-domain training loss relates to out-of-domain alignment scores across datasets and model families. Then, we tried to induce potential alternative local minima through different learning schedules for one narrow fine-tuning, but did not find strong runs with better broad alignment "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20814","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/2606.20814/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":"2026-06-23T00:11:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"knHciS3F9cc8L87YygRlbdDCA2ZJ3W3xx7MpD6hz1WFkMObPRqhdXcWfK4yQEhkEz+zB1UidTuDT23AKw2Y0CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T07:33:38.457913Z"},"content_sha256":"da25eef209c5e4f600e57b377bcc1d2ef158ac6768a5c8dcd9b33075843b423d","schema_version":"1.0","event_id":"sha256:da25eef209c5e4f600e57b377bcc1d2ef158ac6768a5c8dcd9b33075843b423d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/bundle.json","state_url":"https://pith.science/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/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":"2026-06-28T07:33:38Z","links":{"resolver":"https://pith.science/pith/3IBYXXN32UY4ZBEZA7VJNRANUO","bundle":"https://pith.science/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/bundle.json","state":"https://pith.science/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3IBYXXN32UY4ZBEZA7VJNRANUO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:3IBYXXN32UY4ZBEZA7VJNRANUO","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":"5e1aeceadc48874acfaa50798daac3ce5563670b898e8a683791defdac19b7b6","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T18:04:20Z","title_canon_sha256":"a0355d2f359548f87be6995c2e3208785fc9796e6b313d00bb007f4cfd24e9a4"},"schema_version":"1.0","source":{"id":"2606.20814","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20814","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20814v1","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20814","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_12","alias_value":"3IBYXXN32UY4","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_16","alias_value":"3IBYXXN32UY4ZBEZ","created_at":"2026-06-23T00:11:59Z"},{"alias_kind":"pith_short_8","alias_value":"3IBYXXN3","created_at":"2026-06-23T00:11:59Z"}],"graph_snapshots":[{"event_id":"sha256:da25eef209c5e4f600e57b377bcc1d2ef158ac6768a5c8dcd9b33075843b423d","target":"graph","created_at":"2026-06-23T00:11:59Z","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/2606.20814/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Emergent misalignment (EM) is a phenomenon in which models generalize with narrow fine-tuning, leading to broad (yet uneven) misalignment across evaluation questions. We study EM and its variability directly through the components of fine-tuning: training dynamics, model priors, and data. (1) We first explored how in-domain training loss relates to out-of-domain alignment scores across datasets and model families. Then, we tried to induce potential alternative local minima through different learning schedules for one narrow fine-tuning, but did not find strong runs with better broad alignment ","authors_text":"Anietta Weckauff, Diego Garcia-Olano, Maksym Andriushchenko, Yuchen Zhang","cross_cats":["cs.LG","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T18:04:20Z","title":"What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20814","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:ae7d7b33a583a4f812f9f1c87630426627d83aa25a8946460f077c94c50ed21e","target":"record","created_at":"2026-06-23T00:11:59Z","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":"5e1aeceadc48874acfaa50798daac3ce5563670b898e8a683791defdac19b7b6","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T18:04:20Z","title_canon_sha256":"a0355d2f359548f87be6995c2e3208785fc9796e6b313d00bb007f4cfd24e9a4"},"schema_version":"1.0","source":{"id":"2606.20814","kind":"arxiv","version":1}},"canonical_sha256":"da038bddbbd531cc849907ea96c40da38cbe31b809c61b12e1134e373868029c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"da038bddbbd531cc849907ea96c40da38cbe31b809c61b12e1134e373868029c","first_computed_at":"2026-06-23T00:11:59.597688Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-23T00:11:59.597688Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mS6WOzc+PflAcmdPKzLeX+YY/fir4C9cg0zgX8GrMVbPrkUhiIL86rdv04tJHOfuTaDVXVktpROS36eEIX7DCA==","signature_status":"signed_v1","signed_at":"2026-06-23T00:11:59.598089Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.20814","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ae7d7b33a583a4f812f9f1c87630426627d83aa25a8946460f077c94c50ed21e","sha256:da25eef209c5e4f600e57b377bcc1d2ef158ac6768a5c8dcd9b33075843b423d"],"state_sha256":"50ce441045d4035928229e6c997b5ec132bc17d000eb360a7b918060aef555fe"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AroK+2smWytIRM9v2ZHcX9ZqnJfPkLsRfl6b/jfPTlMysopkt2FzQvaAhX1MC4KcQi9lVqCP/Yr9C5D9svC5DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-28T07:33:38.459903Z","bundle_sha256":"f050194f262c7421dc3cec98276b8b4b10b42578746aca8c12787b4f17a134f9"}}