{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:L6I2QLL4YDBS5643BVF2DEN2TI","short_pith_number":"pith:L6I2QLL4","schema_version":"1.0","canonical_sha256":"5f91a82d7cc0c32efb9b0d4ba191ba9a31337df34d019d045a3afbdb3a4988b7","source":{"kind":"arxiv","id":"2402.01822","version":2},"attestation_state":"computed","paper":{"title":"Building Guardrails for Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Gaojie Jin, Jie Meng, Jinwei Hu, Ronghui Mu, Wenjie Ruan, Xiaowei Huang, Xingyu Zhao, Yi Dong, Yi Qi","submitted_at":"2024-02-02T16:35:00Z","abstract_excerpt":"As Large Language Models (LLMs) become more integrated into our daily lives, it is crucial to identify and mitigate their risks, especially when the risks can have profound impacts on human users and societies. Guardrails, which filter the inputs or outputs of LLMs, have emerged as a core safeguarding technology. This position paper takes a deep look at current open-source solutions (Llama Guard, Nvidia NeMo, Guardrails AI), and discusses the challenges and the road towards building more complete solutions. Drawing on robust evidence from previous research, we advocate for a systematic approac"},"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":"2402.01822","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-02T16:35:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"22c96054c0186ac49deb1d34ceeb8e01945dde2084f82abb5b879e2f51723e98","abstract_canon_sha256":"725d128c905b6d288102c4c812e98f921a9c22d8bbb4737deb8adbbc7f29c202"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:40.629853Z","signature_b64":"izTT03nw/dsRaHjoh71alXV2dAjCq4dwC3gHfnbTqWwkD9sYotYFXKtc6aahnOrV1ZmlKlsXKPbbwmEm4EQDAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5f91a82d7cc0c32efb9b0d4ba191ba9a31337df34d019d045a3afbdb3a4988b7","last_reissued_at":"2026-07-05T08:24:40.629417Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:40.629417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Building Guardrails for Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Gaojie Jin, Jie Meng, Jinwei Hu, Ronghui Mu, Wenjie Ruan, Xiaowei Huang, Xingyu Zhao, Yi Dong, Yi Qi","submitted_at":"2024-02-02T16:35:00Z","abstract_excerpt":"As Large Language Models (LLMs) become more integrated into our daily lives, it is crucial to identify and mitigate their risks, especially when the risks can have profound impacts on human users and societies. Guardrails, which filter the inputs or outputs of LLMs, have emerged as a core safeguarding technology. This position paper takes a deep look at current open-source solutions (Llama Guard, Nvidia NeMo, Guardrails AI), and discusses the challenges and the road towards building more complete solutions. Drawing on robust evidence from previous research, we advocate for a systematic approac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01822","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/2402.01822/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":"2402.01822","created_at":"2026-07-05T08:24:40.629472+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.01822v2","created_at":"2026-07-05T08:24:40.629472+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01822","created_at":"2026-07-05T08:24:40.629472+00:00"},{"alias_kind":"pith_short_12","alias_value":"L6I2QLL4YDBS","created_at":"2026-07-05T08:24:40.629472+00:00"},{"alias_kind":"pith_short_16","alias_value":"L6I2QLL4YDBS5643","created_at":"2026-07-05T08:24:40.629472+00:00"},{"alias_kind":"pith_short_8","alias_value":"L6I2QLL4","created_at":"2026-07-05T08:24:40.629472+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08028","citing_title":"From Prompts to Contracts: Harness Engineering for Auditable Enterprise LLM Agents","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2605.26942","citing_title":"Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint)","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2505.16120","citing_title":"LLM-Powered AI Agent Systems and Their Applications in Industry","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19035","citing_title":"Trustworthy Agent Network: Trust in Agent Networks Must Be Baked In, Not Bolted On","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19940","citing_title":"Robotics-Inspired Guardrails for Foundation Models in Socially Sensitive Domains","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2603.22868","citing_title":"Agent-Sentry: Bounding LLM Agents via Execution Provenance","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10990","citing_title":"Skill Drift Is Contract Violation: Proactive Maintenance for LLM Agent Skill Libraries","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2402.19173","citing_title":"StarCoder 2 and The Stack v2: The Next Generation","ref_index":196,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05329","citing_title":"Understanding Annotator Safety Policy with Interpretability","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01462","citing_title":"LocalAlign: Enabling Generalizable Prompt Injection Defense via Generation of Near-Target Adversarial Examples for Alignment Training","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11666","citing_title":"Playing Along: Learning a Double-Agent Defender for Belief Steering via Theory of Mind","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04446","citing_title":"Misrouter: Exploiting Routing Mechanisms for Input-Only Attacks on Mixture-of-Experts LLMs","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI","json":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI.json","graph_json":"https://pith.science/api/pith-number/L6I2QLL4YDBS5643BVF2DEN2TI/graph.json","events_json":"https://pith.science/api/pith-number/L6I2QLL4YDBS5643BVF2DEN2TI/events.json","paper":"https://pith.science/paper/L6I2QLL4"},"agent_actions":{"view_html":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI","download_json":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI.json","view_paper":"https://pith.science/paper/L6I2QLL4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.01822&json=true","fetch_graph":"https://pith.science/api/pith-number/L6I2QLL4YDBS5643BVF2DEN2TI/graph.json","fetch_events":"https://pith.science/api/pith-number/L6I2QLL4YDBS5643BVF2DEN2TI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI/action/storage_attestation","attest_author":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI/action/author_attestation","sign_citation":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI/action/citation_signature","submit_replication":"https://pith.science/pith/L6I2QLL4YDBS5643BVF2DEN2TI/action/replication_record"}},"created_at":"2026-07-05T08:24:40.629472+00:00","updated_at":"2026-07-05T08:24:40.629472+00:00"}