{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:XWXF2MK2L2V7BVXMVBVCG43UWB","short_pith_number":"pith:XWXF2MK2","canonical_record":{"source":{"id":"2605.21006","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T10:43:17Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"2b02046cce0d121366f204a85355fcb59644ad42a1bb66a115ddae54af837dcb","abstract_canon_sha256":"627a611ae08ac9d28eb54424d01db7967b2fe83cb55e4ece0863da270c0e0d1c"},"schema_version":"1.0"},"canonical_sha256":"bdae5d315a5eabf0d6eca86a237374b048a0af19989895c34e4e94fda65484c0","source":{"kind":"arxiv","id":"2605.21006","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21006","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21006v1","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21006","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_12","alias_value":"XWXF2MK2L2V7","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_16","alias_value":"XWXF2MK2L2V7BVXM","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_8","alias_value":"XWXF2MK2","created_at":"2026-05-21T01:05:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:XWXF2MK2L2V7BVXMVBVCG43UWB","target":"record","payload":{"canonical_record":{"source":{"id":"2605.21006","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T10:43:17Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"2b02046cce0d121366f204a85355fcb59644ad42a1bb66a115ddae54af837dcb","abstract_canon_sha256":"627a611ae08ac9d28eb54424d01db7967b2fe83cb55e4ece0863da270c0e0d1c"},"schema_version":"1.0"},"canonical_sha256":"bdae5d315a5eabf0d6eca86a237374b048a0af19989895c34e4e94fda65484c0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:05:31.997715Z","signature_b64":"mBita2WL4y5q7rEg+O0u92ZxD6rwT9M4bJQrXzH4iWfylZAcmBulQsGcMaQ88/JP3+IyJULXDwgXdOBIGunqDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bdae5d315a5eabf0d6eca86a237374b048a0af19989895c34e4e94fda65484c0","last_reissued_at":"2026-05-21T01:05:31.997228Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:05:31.997228Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.21006","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":"2026-05-21T01:05:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XlRaAznt7RoPrYxPo9PiX2B/DxRCxvhqeCm+IWO51asRJUKTsXOWBNRC2GxE1nZ8LBbl6cLMceuIv6LmYW86DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T23:54:39.686442Z"},"content_sha256":"1d1c212be6794fa3d1aaaa5988efe433e9e8f199c6510b2a934197d4d3fec815","schema_version":"1.0","event_id":"sha256:1d1c212be6794fa3d1aaaa5988efe433e9e8f199c6510b2a934197d4d3fec815"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:XWXF2MK2L2V7BVXMVBVCG43UWB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Playing Devil's Advocate: Off-the-Shelf Persona Vectors Rival Targeted Steering for Sycophancy","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ishaan Kelkar, Madhur Panwar, Maheep Chaudhary, Nebras Alam, Vasu Sharma, Vikram Kakaria","submitted_at":"2026-05-20T10:43:17Z","abstract_excerpt":"We study the effect of different persona on \\textbf{sycophancy}: model's agreement with users even when the user is incorrect. The standard mitigation, Contrastive Activation Addition (CAA), derives a steering direction from labelled pairs of sycophantic and honest responses. This study evaluates whether off-the-shelf persona steering vectors, originally developed for general role-playing and not trained on sycophancy data, can serve as an alternative. In two instruction-tuned models, steering toward personas characterised by doubt or scrutiny reduces sycophancy to approximately $68\\%$ and $98"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21006","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/2605.21006/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":"2026-05-21T01:05:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ML/NrwpcaBpB7puwI8Y9MCliJjbdVy/qWTWBTuoz1VKToDNhfk2X9n670nHkFW1KaxZs58RDESEaJtwl2L7QAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T23:54:39.687157Z"},"content_sha256":"04329a66910a134e0599b51b6f17011527226ed58ae092a1c98aa1503de4c815","schema_version":"1.0","event_id":"sha256:04329a66910a134e0599b51b6f17011527226ed58ae092a1c98aa1503de4c815"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/bundle.json","state_url":"https://pith.science/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/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":"2026-05-22T23:54:39Z","links":{"resolver":"https://pith.science/pith/XWXF2MK2L2V7BVXMVBVCG43UWB","bundle":"https://pith.science/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/bundle.json","state":"https://pith.science/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XWXF2MK2L2V7BVXMVBVCG43UWB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:XWXF2MK2L2V7BVXMVBVCG43UWB","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":"627a611ae08ac9d28eb54424d01db7967b2fe83cb55e4ece0863da270c0e0d1c","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T10:43:17Z","title_canon_sha256":"2b02046cce0d121366f204a85355fcb59644ad42a1bb66a115ddae54af837dcb"},"schema_version":"1.0","source":{"id":"2605.21006","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21006","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21006v1","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21006","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_12","alias_value":"XWXF2MK2L2V7","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_16","alias_value":"XWXF2MK2L2V7BVXM","created_at":"2026-05-21T01:05:31Z"},{"alias_kind":"pith_short_8","alias_value":"XWXF2MK2","created_at":"2026-05-21T01:05:31Z"}],"graph_snapshots":[{"event_id":"sha256:04329a66910a134e0599b51b6f17011527226ed58ae092a1c98aa1503de4c815","target":"graph","created_at":"2026-05-21T01:05:31Z","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/2605.21006/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We study the effect of different persona on \\textbf{sycophancy}: model's agreement with users even when the user is incorrect. The standard mitigation, Contrastive Activation Addition (CAA), derives a steering direction from labelled pairs of sycophantic and honest responses. This study evaluates whether off-the-shelf persona steering vectors, originally developed for general role-playing and not trained on sycophancy data, can serve as an alternative. In two instruction-tuned models, steering toward personas characterised by doubt or scrutiny reduces sycophancy to approximately $68\\%$ and $98","authors_text":"Ishaan Kelkar, Madhur Panwar, Maheep Chaudhary, Nebras Alam, Vasu Sharma, Vikram Kakaria","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T10:43:17Z","title":"Playing Devil's Advocate: Off-the-Shelf Persona Vectors Rival Targeted Steering for Sycophancy"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21006","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":"sha256:1d1c212be6794fa3d1aaaa5988efe433e9e8f199c6510b2a934197d4d3fec815","target":"record","created_at":"2026-05-21T01:05:31Z","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":"627a611ae08ac9d28eb54424d01db7967b2fe83cb55e4ece0863da270c0e0d1c","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-20T10:43:17Z","title_canon_sha256":"2b02046cce0d121366f204a85355fcb59644ad42a1bb66a115ddae54af837dcb"},"schema_version":"1.0","source":{"id":"2605.21006","kind":"arxiv","version":1}},"canonical_sha256":"bdae5d315a5eabf0d6eca86a237374b048a0af19989895c34e4e94fda65484c0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bdae5d315a5eabf0d6eca86a237374b048a0af19989895c34e4e94fda65484c0","first_computed_at":"2026-05-21T01:05:31.997228Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T01:05:31.997228Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mBita2WL4y5q7rEg+O0u92ZxD6rwT9M4bJQrXzH4iWfylZAcmBulQsGcMaQ88/JP3+IyJULXDwgXdOBIGunqDg==","signature_status":"signed_v1","signed_at":"2026-05-21T01:05:31.997715Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.21006","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1d1c212be6794fa3d1aaaa5988efe433e9e8f199c6510b2a934197d4d3fec815","sha256:04329a66910a134e0599b51b6f17011527226ed58ae092a1c98aa1503de4c815"],"state_sha256":"be6672a6ae210643016c0d6b10f8959b43613656975d94cef79fb69dcec6ce73"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ya8ogLC2tEP2jZ/Lttte4WVRQq4Ejt3Yn90/ZPlmtSTH2bCpXRJdRIYx/KKWvKLB9wZr57BTqGbXUmeQAp/UAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T23:54:39.690860Z","bundle_sha256":"a857ca7e596443ced1632a9c6c344e9f3a1a952670907551977d9076ae2852fa"}}