{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RML4BJ5HC55QY6EXCBNFCY7FC2","short_pith_number":"pith:RML4BJ5H","canonical_record":{"source":{"id":"2608.05624","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T05:42:30Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"5082ba19ef935b2c22ed53093b107612fbe0874247fdfbc6efdafbdf84f17743","abstract_canon_sha256":"4a30d717fd0e1989d7f7dddabb84103839ac92edf6e3df754b72f19267dfca87"},"schema_version":"1.0"},"canonical_sha256":"8b17c0a7a7177b0c7897105a5163e516884defed345945561741be953513db6f","source":{"kind":"arxiv","id":"2608.05624","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.05624","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"arxiv_version","alias_value":"2608.05624v1","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.05624","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_12","alias_value":"RML4BJ5HC55Q","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_16","alias_value":"RML4BJ5HC55QY6EX","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_8","alias_value":"RML4BJ5H","created_at":"2026-08-07T00:50:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RML4BJ5HC55QY6EXCBNFCY7FC2","target":"record","payload":{"canonical_record":{"source":{"id":"2608.05624","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T05:42:30Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"5082ba19ef935b2c22ed53093b107612fbe0874247fdfbc6efdafbdf84f17743","abstract_canon_sha256":"4a30d717fd0e1989d7f7dddabb84103839ac92edf6e3df754b72f19267dfca87"},"schema_version":"1.0"},"canonical_sha256":"8b17c0a7a7177b0c7897105a5163e516884defed345945561741be953513db6f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-07T00:50:19.670907Z","signature_b64":"9jY0ENvLm+U9XS11zXAMVpgDgaiBOubkNAGfsYGgrlMbnm55tPYIAFNT8jf3+Zy69BElngIwth359FwYr/t+Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8b17c0a7a7177b0c7897105a5163e516884defed345945561741be953513db6f","last_reissued_at":"2026-08-07T00:50:19.669252Z","signature_status":"signed_v1","first_computed_at":"2026-08-07T00:50:19.669252Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.05624","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":"2026-08-07T00:50:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ggfLFSRyeUFmnxTUW9fu9ETlq9IN9CFSUDT9qNnuRCTglSc9MgIjuZ+CClAqIzTh1aRKC3DqTn4051SZHHpGCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T19:34:27.787717Z"},"content_sha256":"eec705b7d5c33f06e1115cecaf5f13c41e65c92a00bfa124884666897e26922c","schema_version":"1.0","event_id":"sha256:eec705b7d5c33f06e1115cecaf5f13c41e65c92a00bfa124884666897e26922c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RML4BJ5HC55QY6EXCBNFCY7FC2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Measuring and Detecting Harmful AI Sycophancy","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bohan Jiang, Dawei Li, Huan Liu, Yasin Silva","submitted_at":"2026-08-06T05:42:30Z","abstract_excerpt":"Sycophantic responses are becoming pervasive in large language models (LLMs), and prior work has pointed out that some of them could be harmful. This paper focuses on one harmful sycophancy: preference-induced stance reversal sycophancy (PSRS), where a model reverses an initial stance merely to align with a user's stated preference. While existing research mainly measures how sycophantic a model is, we go further and ask whether PSRS can also be detected automatically from a single response. To investigate this at scale, we introduce CAP (Contrastive Anchor Probing), a framework for collecting"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.05624","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/2608.05624/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":"2026-08-07T00:50:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"S4/+2hb65rJL95wCFdth/kzSzWom9b3AlMvF+tG96wIATLquYPv3+BxP7wufU31RylOgEVkYan77xoP6v3PwBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T19:34:27.788214Z"},"content_sha256":"0170af803a56833f37f3e6153ea71a9ad851ecdcf2eca8963fdc68f30d80592b","schema_version":"1.0","event_id":"sha256:0170af803a56833f37f3e6153ea71a9ad851ecdcf2eca8963fdc68f30d80592b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/bundle.json","state_url":"https://pith.science/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/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":"2026-08-08T19:34:27Z","links":{"resolver":"https://pith.science/pith/RML4BJ5HC55QY6EXCBNFCY7FC2","bundle":"https://pith.science/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/bundle.json","state":"https://pith.science/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RML4BJ5HC55QY6EXCBNFCY7FC2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RML4BJ5HC55QY6EXCBNFCY7FC2","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":"4a30d717fd0e1989d7f7dddabb84103839ac92edf6e3df754b72f19267dfca87","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T05:42:30Z","title_canon_sha256":"5082ba19ef935b2c22ed53093b107612fbe0874247fdfbc6efdafbdf84f17743"},"schema_version":"1.0","source":{"id":"2608.05624","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.05624","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"arxiv_version","alias_value":"2608.05624v1","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.05624","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_12","alias_value":"RML4BJ5HC55Q","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_16","alias_value":"RML4BJ5HC55QY6EX","created_at":"2026-08-07T00:50:19Z"},{"alias_kind":"pith_short_8","alias_value":"RML4BJ5H","created_at":"2026-08-07T00:50:19Z"}],"graph_snapshots":[{"event_id":"sha256:0170af803a56833f37f3e6153ea71a9ad851ecdcf2eca8963fdc68f30d80592b","target":"graph","created_at":"2026-08-07T00:50:19Z","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/2608.05624/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Sycophantic responses are becoming pervasive in large language models (LLMs), and prior work has pointed out that some of them could be harmful. This paper focuses on one harmful sycophancy: preference-induced stance reversal sycophancy (PSRS), where a model reverses an initial stance merely to align with a user's stated preference. While existing research mainly measures how sycophantic a model is, we go further and ask whether PSRS can also be detected automatically from a single response. To investigate this at scale, we introduce CAP (Contrastive Anchor Probing), a framework for collecting","authors_text":"Bohan Jiang, Dawei Li, Huan Liu, Yasin Silva","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T05:42:30Z","title":"Measuring and Detecting Harmful AI Sycophancy"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.05624","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:eec705b7d5c33f06e1115cecaf5f13c41e65c92a00bfa124884666897e26922c","target":"record","created_at":"2026-08-07T00:50:19Z","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":"4a30d717fd0e1989d7f7dddabb84103839ac92edf6e3df754b72f19267dfca87","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T05:42:30Z","title_canon_sha256":"5082ba19ef935b2c22ed53093b107612fbe0874247fdfbc6efdafbdf84f17743"},"schema_version":"1.0","source":{"id":"2608.05624","kind":"arxiv","version":1}},"canonical_sha256":"8b17c0a7a7177b0c7897105a5163e516884defed345945561741be953513db6f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8b17c0a7a7177b0c7897105a5163e516884defed345945561741be953513db6f","first_computed_at":"2026-08-07T00:50:19.669252Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-07T00:50:19.669252Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"9jY0ENvLm+U9XS11zXAMVpgDgaiBOubkNAGfsYGgrlMbnm55tPYIAFNT8jf3+Zy69BElngIwth359FwYr/t+Bw==","signature_status":"signed_v1","signed_at":"2026-08-07T00:50:19.670907Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.05624","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:eec705b7d5c33f06e1115cecaf5f13c41e65c92a00bfa124884666897e26922c","sha256:0170af803a56833f37f3e6153ea71a9ad851ecdcf2eca8963fdc68f30d80592b"],"state_sha256":"a2bc904687aaa3c04f25e8100e0780e07422a8afe63cfa4f8b7b5ed6224485e1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ixu2pJPCw7RtCYVglzOQyUB1vFYsItjXTiS0Ii3fSctRAOeImSoyDHWbmOeQaQnVVcujI4s1Q/dUHUYNNtYlCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T19:34:27.792148Z","bundle_sha256":"215e5096985143e8d9d4cbd58e6f3d1cd003b47e2aee40379bc35f47b20a7db7"}}