{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5EN3ABVUVTFHHTVZLIIX5QD4GU","short_pith_number":"pith:5EN3ABVU","schema_version":"1.0","canonical_sha256":"e91bb006b4acca73ceb95a117ec07c350e369cb0547630cd77723b39418d87f4","source":{"kind":"arxiv","id":"2603.13356","version":2},"attestation_state":"computed","paper":{"title":"Learning When to Trust in Contextual Social Bandits","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Majid Ghasemi, Mark Crowley","submitted_at":"2026-03-09T01:35:37Z","abstract_excerpt":"Robust reinforcement learning typically assumes that feedback sources are either globally trustworthy or corrupted within a fixed global budget. We identify a more subtle failure mode that escapes this dichotomy, which we call \\emph{Contextual Sycophancy}. In this failure, evaluators are truthful in benign contexts but systematically biased in critical ones, so that no single evaluator is reliable everywhere and the corrupt evaluators may form a \\emph{majority} in the contexts that matter. Our first result is an information-theoretic lower bound. We exhibit two problem instances that induce \\e"},"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":"2603.13356","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-03-09T01:35:37Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"196a11c90621ffcb5b39fe6ecb5fdc5abb05feb4c10155f013506391f6010fcd","abstract_canon_sha256":"fc8d750e903b8d9f62596b7b329cfc39f19cad16b9ab0a691979dc4c8b53f092"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-15T01:21:51.343325Z","signature_b64":"+SYp9RWUwsZbERWaAlyiBgEoVn+CE3y+FXrNIacd4vFDwYtfuPDf3Zm0dlslqRQ/nQYWs9qW/MLDWEqRwVNIBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e91bb006b4acca73ceb95a117ec07c350e369cb0547630cd77723b39418d87f4","last_reissued_at":"2026-07-15T01:21:51.342369Z","signature_status":"signed_v1","first_computed_at":"2026-07-15T01:21:51.342369Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning When to Trust in Contextual Social Bandits","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Majid Ghasemi, Mark Crowley","submitted_at":"2026-03-09T01:35:37Z","abstract_excerpt":"Robust reinforcement learning typically assumes that feedback sources are either globally trustworthy or corrupted within a fixed global budget. We identify a more subtle failure mode that escapes this dichotomy, which we call \\emph{Contextual Sycophancy}. In this failure, evaluators are truthful in benign contexts but systematically biased in critical ones, so that no single evaluator is reliable everywhere and the corrupt evaluators may form a \\emph{majority} in the contexts that matter. Our first result is an information-theoretic lower bound. We exhibit two problem instances that induce \\e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.13356","kind":"arxiv","version":2},"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/2603.13356/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":"2603.13356","created_at":"2026-07-15T01:21:51.342833+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.13356v2","created_at":"2026-07-15T01:21:51.342833+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.13356","created_at":"2026-07-15T01:21:51.342833+00:00"},{"alias_kind":"pith_short_12","alias_value":"5EN3ABVUVTFH","created_at":"2026-07-15T01:21:51.342833+00:00"},{"alias_kind":"pith_short_16","alias_value":"5EN3ABVUVTFHHTVZ","created_at":"2026-07-15T01:21:51.342833+00:00"},{"alias_kind":"pith_short_8","alias_value":"5EN3ABVU","created_at":"2026-07-15T01:21:51.342833+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/5EN3ABVUVTFHHTVZLIIX5QD4GU","json":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU.json","graph_json":"https://pith.science/api/pith-number/5EN3ABVUVTFHHTVZLIIX5QD4GU/graph.json","events_json":"https://pith.science/api/pith-number/5EN3ABVUVTFHHTVZLIIX5QD4GU/events.json","paper":"https://pith.science/paper/5EN3ABVU"},"agent_actions":{"view_html":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU","download_json":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU.json","view_paper":"https://pith.science/paper/5EN3ABVU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.13356&json=true","fetch_graph":"https://pith.science/api/pith-number/5EN3ABVUVTFHHTVZLIIX5QD4GU/graph.json","fetch_events":"https://pith.science/api/pith-number/5EN3ABVUVTFHHTVZLIIX5QD4GU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU/action/storage_attestation","attest_author":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU/action/author_attestation","sign_citation":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU/action/citation_signature","submit_replication":"https://pith.science/pith/5EN3ABVUVTFHHTVZLIIX5QD4GU/action/replication_record"}},"created_at":"2026-07-15T01:21:51.342833+00:00","updated_at":"2026-07-15T01:21:51.342833+00:00"}