{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WWKZ6SAP4S2UGYHDRLZLIBP42U","short_pith_number":"pith:WWKZ6SAP","schema_version":"1.0","canonical_sha256":"b5959f480fe4b54360e38af2b405fcd52db41c2abf800f269f5036e3a0f599b9","source":{"kind":"arxiv","id":"2408.08959","version":3},"attestation_state":"computed","paper":{"title":"Trust-Oriented Adaptive Guardrails for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Jinwei Hu, Xiaowei Huang, Yi Dong","submitted_at":"2024-08-16T18:07:48Z","abstract_excerpt":"Guardrail, an emerging mechanism designed to ensure that large language models (LLMs) align with human values by moderating harmful or toxic responses, requires a sociotechnical approach in their design. This paper addresses a critical issue: existing guardrails lack a well-founded methodology to accommodate the diverse needs of different user groups, particularly concerning access rights. Supported by trust modeling (primarily on `social' aspect) and enhanced with online in-context learning via retrieval-augmented generation (on `technical' aspect), we introduce an adaptive guardrail mechanis"},"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":"2408.08959","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-08-16T18:07:48Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1703164ae1edb32a4bc718b0b5622e4e739316240f1e63d1b8a1edd71296ba73","abstract_canon_sha256":"bd0b3259885078be708bd75be6192667969ce92427116e06c75e02a82f3f632e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:15:17.443196Z","signature_b64":"RcXVkVyz5QxaJIvNxGXkeUgFgkQ0lrCx4LX6RUEWDP9q3D8tHBZqJ6eosxZglCzNya8BZBn+4kQd4dAke48+AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b5959f480fe4b54360e38af2b405fcd52db41c2abf800f269f5036e3a0f599b9","last_reissued_at":"2026-07-05T11:15:17.442704Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:15:17.442704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Trust-Oriented Adaptive Guardrails for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Jinwei Hu, Xiaowei Huang, Yi Dong","submitted_at":"2024-08-16T18:07:48Z","abstract_excerpt":"Guardrail, an emerging mechanism designed to ensure that large language models (LLMs) align with human values by moderating harmful or toxic responses, requires a sociotechnical approach in their design. This paper addresses a critical issue: existing guardrails lack a well-founded methodology to accommodate the diverse needs of different user groups, particularly concerning access rights. Supported by trust modeling (primarily on `social' aspect) and enhanced with online in-context learning via retrieval-augmented generation (on `technical' aspect), we introduce an adaptive guardrail mechanis"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.08959","kind":"arxiv","version":3},"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/2408.08959/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":"2408.08959","created_at":"2026-07-05T11:15:17.442763+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.08959v3","created_at":"2026-07-05T11:15:17.442763+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.08959","created_at":"2026-07-05T11:15:17.442763+00:00"},{"alias_kind":"pith_short_12","alias_value":"WWKZ6SAP4S2U","created_at":"2026-07-05T11:15:17.442763+00:00"},{"alias_kind":"pith_short_16","alias_value":"WWKZ6SAP4S2UGYHD","created_at":"2026-07-05T11:15:17.442763+00:00"},{"alias_kind":"pith_short_8","alias_value":"WWKZ6SAP","created_at":"2026-07-05T11:15:17.442763+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.28908","citing_title":"Who Does Your AI Work For? Designing Conversational Agents as Digital Fiduciaries","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U","json":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U.json","graph_json":"https://pith.science/api/pith-number/WWKZ6SAP4S2UGYHDRLZLIBP42U/graph.json","events_json":"https://pith.science/api/pith-number/WWKZ6SAP4S2UGYHDRLZLIBP42U/events.json","paper":"https://pith.science/paper/WWKZ6SAP"},"agent_actions":{"view_html":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U","download_json":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U.json","view_paper":"https://pith.science/paper/WWKZ6SAP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.08959&json=true","fetch_graph":"https://pith.science/api/pith-number/WWKZ6SAP4S2UGYHDRLZLIBP42U/graph.json","fetch_events":"https://pith.science/api/pith-number/WWKZ6SAP4S2UGYHDRLZLIBP42U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U/action/storage_attestation","attest_author":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U/action/author_attestation","sign_citation":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U/action/citation_signature","submit_replication":"https://pith.science/pith/WWKZ6SAP4S2UGYHDRLZLIBP42U/action/replication_record"}},"created_at":"2026-07-05T11:15:17.442763+00:00","updated_at":"2026-07-05T11:15:17.442763+00:00"}