{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UR6NTD7I7OCLX235PODHH7N37W","short_pith_number":"pith:UR6NTD7I","schema_version":"1.0","canonical_sha256":"a47cd98fe8fb84bbeb7d7b8673fdbbfd9708e029b3bf2fa9f59519c48e0725ec","source":{"kind":"arxiv","id":"2608.08212","version":1},"attestation_state":"computed","paper":{"title":"Harmful Content Is Not Enough: Continuation Framing Moderates In-Context Emergent Misalignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Di Liang, Peiyang Liu, Wei Ye, Xi Wang, Ziqiang Cui","submitted_at":"2026-08-08T16:13:38Z","abstract_excerpt":"In-context learning (ICL) can induce emergent misalignment (EM), where narrow misaligned examples alter answers to unrelated questions. Existing prompts, however, conflate harmful-text exposure with an invitation to continue assistant behavior. We hold harmful answers fixed while varying their delivery as demonstrations, evidence, assistant history, or tool output. Across ten independently sampled contexts, demonstration framing raises broad EM by $30$--$32$ percentage points on a susceptible Gemini model; the gap survives domain exclusion, semantic clustering, unseen questions, and four promp"},"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":"2608.08212","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-08T16:13:38Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1cffe068b97f98c853bc3af94fe1ae15d7e86e34cfa6cd6f5e76290b91f0d6bf","abstract_canon_sha256":"916079133b35ea185007216058daad687fc22a721bdc1b413526845272954e24"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:21:59.893283Z","signature_b64":"KkbdfvW5oyXIV+uyPzdpNVD+d5/v4RtZCmDPxqRfPYuQSjA4NF79lpwRxHG7iPnaqHX4xMQ7s8RgAUz8/ca+Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a47cd98fe8fb84bbeb7d7b8673fdbbfd9708e029b3bf2fa9f59519c48e0725ec","last_reissued_at":"2026-08-11T01:21:59.890912Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:21:59.890912Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Harmful Content Is Not Enough: Continuation Framing Moderates In-Context Emergent Misalignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Di Liang, Peiyang Liu, Wei Ye, Xi Wang, Ziqiang Cui","submitted_at":"2026-08-08T16:13:38Z","abstract_excerpt":"In-context learning (ICL) can induce emergent misalignment (EM), where narrow misaligned examples alter answers to unrelated questions. Existing prompts, however, conflate harmful-text exposure with an invitation to continue assistant behavior. We hold harmful answers fixed while varying their delivery as demonstrations, evidence, assistant history, or tool output. Across ten independently sampled contexts, demonstration framing raises broad EM by $30$--$32$ percentage points on a susceptible Gemini model; the gap survives domain exclusion, semantic clustering, unseen questions, and four promp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08212","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/2608.08212/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":"2608.08212","created_at":"2026-08-11T01:21:59.891630+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.08212v1","created_at":"2026-08-11T01:21:59.891630+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08212","created_at":"2026-08-11T01:21:59.891630+00:00"},{"alias_kind":"pith_short_12","alias_value":"UR6NTD7I7OCL","created_at":"2026-08-11T01:21:59.891630+00:00"},{"alias_kind":"pith_short_16","alias_value":"UR6NTD7I7OCLX235","created_at":"2026-08-11T01:21:59.891630+00:00"},{"alias_kind":"pith_short_8","alias_value":"UR6NTD7I","created_at":"2026-08-11T01:21:59.891630+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W","json":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W.json","graph_json":"https://pith.science/api/pith-number/UR6NTD7I7OCLX235PODHH7N37W/graph.json","events_json":"https://pith.science/api/pith-number/UR6NTD7I7OCLX235PODHH7N37W/events.json","paper":"https://pith.science/paper/UR6NTD7I"},"agent_actions":{"view_html":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W","download_json":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W.json","view_paper":"https://pith.science/paper/UR6NTD7I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.08212&json=true","fetch_graph":"https://pith.science/api/pith-number/UR6NTD7I7OCLX235PODHH7N37W/graph.json","fetch_events":"https://pith.science/api/pith-number/UR6NTD7I7OCLX235PODHH7N37W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W/action/storage_attestation","attest_author":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W/action/author_attestation","sign_citation":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W/action/citation_signature","submit_replication":"https://pith.science/pith/UR6NTD7I7OCLX235PODHH7N37W/action/replication_record"}},"created_at":"2026-08-11T01:21:59.891630+00:00","updated_at":"2026-08-11T01:21:59.891630+00:00"}