{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:FYYFKCUPWPLKRYWPP3EJZPG2MB","short_pith_number":"pith:FYYFKCUP","schema_version":"1.0","canonical_sha256":"2e30550a8fb3d6a8e2cf7ec89cbcda6066262cc2b0faefbe96e6ca6e8524e298","source":{"kind":"arxiv","id":"2502.08142","version":1},"attestation_state":"computed","paper":{"title":"Bridging the Safety Gap: A Guardrail Pipeline for Trustworthy LLM Inferences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chaoyang He, Salman Avestimehr, Shanshan Han","submitted_at":"2025-02-12T05:48:57Z","abstract_excerpt":"We present Wildflare GuardRail, a guardrail pipeline designed to enhance the safety and reliability of Large Language Model (LLM) inferences by systematically addressing risks across the entire processing workflow. Wildflare GuardRail integrates several core functional modules, including Safety Detector that identifies unsafe inputs and detects hallucinations in model outputs while generating root-cause explanations, Grounding that contextualizes user queries with information retrieved from vector databases, Customizer that adjusts outputs in real time using lightweight, rule-based wrappers, a"},"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":"2502.08142","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T05:48:57Z","cross_cats_sorted":[],"title_canon_sha256":"a953b6080514f9e6b5854519f735fee0cba1770a29bfc73eb977220af3456594","abstract_canon_sha256":"4c693c248fad864377b35a35511b90bdbe32ffcf3ce1c06b79f29add02a20072"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:13:16.069957Z","signature_b64":"vclRypW8RX8ZHSEKH3kumTW31ITnOXsBL/Uv+vDvzU1eeDbLiUUvVqheF/PwfUT+sjMJLu+dyfbdZ7WbWRk/Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e30550a8fb3d6a8e2cf7ec89cbcda6066262cc2b0faefbe96e6ca6e8524e298","last_reissued_at":"2026-07-05T10:13:16.069433Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:13:16.069433Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bridging the Safety Gap: A Guardrail Pipeline for Trustworthy LLM Inferences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chaoyang He, Salman Avestimehr, Shanshan Han","submitted_at":"2025-02-12T05:48:57Z","abstract_excerpt":"We present Wildflare GuardRail, a guardrail pipeline designed to enhance the safety and reliability of Large Language Model (LLM) inferences by systematically addressing risks across the entire processing workflow. Wildflare GuardRail integrates several core functional modules, including Safety Detector that identifies unsafe inputs and detects hallucinations in model outputs while generating root-cause explanations, Grounding that contextualizes user queries with information retrieved from vector databases, Customizer that adjusts outputs in real time using lightweight, rule-based wrappers, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08142","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/2502.08142/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":"2502.08142","created_at":"2026-07-05T10:13:16.069498+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.08142v1","created_at":"2026-07-05T10:13:16.069498+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08142","created_at":"2026-07-05T10:13:16.069498+00:00"},{"alias_kind":"pith_short_12","alias_value":"FYYFKCUPWPLK","created_at":"2026-07-05T10:13:16.069498+00:00"},{"alias_kind":"pith_short_16","alias_value":"FYYFKCUPWPLKRYWP","created_at":"2026-07-05T10:13:16.069498+00:00"},{"alias_kind":"pith_short_8","alias_value":"FYYFKCUP","created_at":"2026-07-05T10:13:16.069498+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.23196","citing_title":"Prompt Overflow: What the Guardrail Inspects Is Not What the Model Infers","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17497","citing_title":"Generative AI Technologies, Techniques & Tensions: A Primer","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB","json":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB.json","graph_json":"https://pith.science/api/pith-number/FYYFKCUPWPLKRYWPP3EJZPG2MB/graph.json","events_json":"https://pith.science/api/pith-number/FYYFKCUPWPLKRYWPP3EJZPG2MB/events.json","paper":"https://pith.science/paper/FYYFKCUP"},"agent_actions":{"view_html":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB","download_json":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB.json","view_paper":"https://pith.science/paper/FYYFKCUP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.08142&json=true","fetch_graph":"https://pith.science/api/pith-number/FYYFKCUPWPLKRYWPP3EJZPG2MB/graph.json","fetch_events":"https://pith.science/api/pith-number/FYYFKCUPWPLKRYWPP3EJZPG2MB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB/action/storage_attestation","attest_author":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB/action/author_attestation","sign_citation":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB/action/citation_signature","submit_replication":"https://pith.science/pith/FYYFKCUPWPLKRYWPP3EJZPG2MB/action/replication_record"}},"created_at":"2026-07-05T10:13:16.069498+00:00","updated_at":"2026-07-05T10:13:16.069498+00:00"}