{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:WUT4QINO4N7AHTNN2XRPR6MSZT","short_pith_number":"pith:WUT4QINO","schema_version":"1.0","canonical_sha256":"b527c821aee37e03cdadd5e2f8f992ccd5e6dfd870b1f13fcda15bb959082cc1","source":{"kind":"arxiv","id":"2605.23974","version":1},"attestation_state":"computed","paper":{"title":"AERIC: Anticipatory Hidden-State Monitoring for Implicit Harmful Dialogue","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jihyung Park, Junfeng Jiao, Saleh Afroogh","submitted_at":"2026-05-13T14:18:29Z","abstract_excerpt":"Current language models create two safety challenges: risk must be detected early enough to avoid exposing harmful continuation, and the harmfulness itself may be implicit rather than signaled by overtly toxic text. Existing response-level guards are strong at judging completed text, and native streaming guards move closer to token time, but both settings leave open whether a lightweight monitor can anticipate implicit harmful drift from the generator's own internal trajectory. We study anticipatory same-pass monitoring, where a safety monitor may read hidden states produced during ordinary de"},"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":"2605.23974","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T14:18:29Z","cross_cats_sorted":[],"title_canon_sha256":"1939924d3df121e886f427e9b2de56bec870c91df46132b005e895ae8d7ce66d","abstract_canon_sha256":"9c936d24f6f0a16ceca284f8305247ac77af934bbd56a9dde29e248b0464d345"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T00:02:24.708220Z","signature_b64":"EoCZW1eA22sBurcjNZCDRHdawtfXaH5ukV+IStT4LC5e4fVyl60nJqmJYtlzq1GnzXZyClLCVV9nlE2KsBi7Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b527c821aee37e03cdadd5e2f8f992ccd5e6dfd870b1f13fcda15bb959082cc1","last_reissued_at":"2026-05-26T00:02:24.707455Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T00:02:24.707455Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AERIC: Anticipatory Hidden-State Monitoring for Implicit Harmful Dialogue","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jihyung Park, Junfeng Jiao, Saleh Afroogh","submitted_at":"2026-05-13T14:18:29Z","abstract_excerpt":"Current language models create two safety challenges: risk must be detected early enough to avoid exposing harmful continuation, and the harmfulness itself may be implicit rather than signaled by overtly toxic text. Existing response-level guards are strong at judging completed text, and native streaming guards move closer to token time, but both settings leave open whether a lightweight monitor can anticipate implicit harmful drift from the generator's own internal trajectory. We study anticipatory same-pass monitoring, where a safety monitor may read hidden states produced during ordinary de"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23974","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/2605.23974/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":"2605.23974","created_at":"2026-05-26T00:02:24.707565+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23974v1","created_at":"2026-05-26T00:02:24.707565+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23974","created_at":"2026-05-26T00:02:24.707565+00:00"},{"alias_kind":"pith_short_12","alias_value":"WUT4QINO4N7A","created_at":"2026-05-26T00:02:24.707565+00:00"},{"alias_kind":"pith_short_16","alias_value":"WUT4QINO4N7AHTNN","created_at":"2026-05-26T00:02:24.707565+00:00"},{"alias_kind":"pith_short_8","alias_value":"WUT4QINO","created_at":"2026-05-26T00:02:24.707565+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/WUT4QINO4N7AHTNN2XRPR6MSZT","json":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT.json","graph_json":"https://pith.science/api/pith-number/WUT4QINO4N7AHTNN2XRPR6MSZT/graph.json","events_json":"https://pith.science/api/pith-number/WUT4QINO4N7AHTNN2XRPR6MSZT/events.json","paper":"https://pith.science/paper/WUT4QINO"},"agent_actions":{"view_html":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT","download_json":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT.json","view_paper":"https://pith.science/paper/WUT4QINO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23974&json=true","fetch_graph":"https://pith.science/api/pith-number/WUT4QINO4N7AHTNN2XRPR6MSZT/graph.json","fetch_events":"https://pith.science/api/pith-number/WUT4QINO4N7AHTNN2XRPR6MSZT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT/action/storage_attestation","attest_author":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT/action/author_attestation","sign_citation":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT/action/citation_signature","submit_replication":"https://pith.science/pith/WUT4QINO4N7AHTNN2XRPR6MSZT/action/replication_record"}},"created_at":"2026-05-26T00:02:24.707565+00:00","updated_at":"2026-05-26T00:02:24.707565+00:00"}