Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-01T04:12:58.815158Z
Paper Citation Record · LEDGER
As of 17 August 2026, this Paper Citation Record lists 71 of 71 outbound references and 0 inbound Pith citation observations for arXiv:2607.22926.
A citation records a reference. It does not transfer a finding from one paper to another.
Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-01T04:12:58.815158Z
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-08-17T06:30:58.91139+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links
A source-named dated measurement, never combined with another source.
Source: cited_works
71 of 71 outbound references displayed
External citation measurements
No source-named external measurement is stored.
Observation e386d600-11db-4447-bbb0-3bdb46b0c2e6 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI and Bates, Stephen and Fisch, Adam and Lei, Lihua and Schuster, Tal , title =
Reference 1
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 21d0965f-d4e5-49b1-8cf8-68b917901aa1 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 2
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2240337a-c29d-4f24-abe9-8073c986602f · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 225 , year =
Reference 3
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6fbce0bd-8b3f-4142-83d2-04e419ad532a · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Proceedings of the 42nd International Conference on Machine Learning , year =
Reference 4
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d91a120a-d822-4b24-8f99-537d6bc3322b · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 5
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6fc1e2b1-9943-474c-9b93-735521a4edc9 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Official Journal of the European Union , url =
Reference 6
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ac01214c-91e6-44fb-82ec-d5947046d5ff · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Advances in Neural Information Processing Systems , volume =
Reference 7
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7f192b9d-e95c-4bdf-bb3b-21641cfa5905 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Computer Aided Verification , series =
Reference 8
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation e7227c6f-5b58-41c3-abae-49fc78e7dac6 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Journal of Logic and Algebraic Programming , volume =
Reference 9
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 4b1c7390-57a4-43a6-8556-803e189d5c89 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Proceedings of the 41st International Conference on Machine Learning , series =
Reference 10
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2f581593-88e8-4a40-a6b1-6ec21efb9731 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2023 , doi =
Reference 11
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 75006494-161e-473a-8a78-eb8d15910679 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2024 , doi =
Reference 12
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation e2d3c197-53ee-48ab-af77-2bdf6bd41595 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 13
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2267c734-e9cf-4c28-8228-74fa2dfb08be · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , pages =
Reference 14
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 621d31ba-bd70-4d5e-9aa8-ded0b18d444b · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI and Schroeder, Michael D
Reference 15
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 45209cb0-d433-4458-be9a-e791f14313d8 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 16
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation c076d59f-544b-491e-97d1-10a78912dd96 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Advances in Neural Information Processing Systems , volume =
Reference 17
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2cf1a76e-3bd3-4abf-bd23-2923b991510d · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI International Conference on Learning Representations , year =
Reference 18
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3fbdd567-3e93-477e-bb33-75531a62bb26 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 19
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 85db97ed-5629-4e6c-a6e7-e1e92567facf · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Advances in Neural Information Processing Systems , volume =
Reference 20
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 47fe86a6-3dbc-4074-a1c2-cb4638f34dd0 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2024 , url =
Reference 21
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 30e7271c-0435-4f54-9b58-2065402017b9 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 22
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3dd95b80-2175-45b6-9f8b-81b33e2b18d9 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 23
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation f4f58328-2972-4a04-8678-e34b7b022542 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 24
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7d9e6bb8-dbef-4421-94ef-a46a252cc254 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2025 , url =
Reference 25
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b93cbb9c-790a-42b5-8ff5-3d6436ade7e0 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Unresolved cited work
Reference 26
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 226d0859-17b0-4c84-a502-f081fbc3a68d · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2023 , howpublished =
Reference 27
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 71107b49-84da-4093-8952-702e58766bcf · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations , pages =
Reference 28
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6d9f5401-728b-4c1e-8d3f-d3ef8c809365 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2024 , howpublished =
Reference 29
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7e3ed754-7fc3-481a-8c22-95dbf5f88121 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Proceedings of the AAAI Conference on Artificial Intelligence , volume =
Reference 30
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 9a77f8e3-45a4-49fd-b284-1d24de734995 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2025 , howpublished =
Reference 31
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6e6b469b-848e-4a1e-8355-1f37e39cdd63 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI and Deshpande, Kaustubh and Sirdeshmukh, Ved and Mankikar, Meher and Scale Red Team and SEAL Research Team and Michael, Julian , journal =
Reference 33
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation be6220f4-e2c9-4150-aaad-03b1f6ad96ec · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Safer or Luckier?
Reference 35
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 84a56c00-f822-4bc7-861b-37a0a505d926 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Investigating the Potential Use of Frontier
Reference 36
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2d7b209b-9fda-49b4-8baa-f9b924952d54 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI 2025 , howpublished =
Reference 37
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 9ef102d9-e442-4745-9e8b-8810bc54f6fe · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Investigating the potential use of frontier AI models for offensive cyberattacks: A human uplift study
Reference 38
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d6aa0650-fda6-40c7-99de-26bc3c15f75c · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Angelopoulos, Stephen Bates, Adam Fisch, Lihua Lei, and Tal Schuster
Reference 39
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 4788f1d8-9b6d-46dc-9544-a5eee0216bf6 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Why do we take LLM s seriously as a potential source of biorisk? https://www.anthropic.com/research/biorisk, 2025
Reference 40
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3b898e95-62b6-4618-9ad8-0e8d5f84822a · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Claude system cards, 2026
Reference 41
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 456e7b70-c241-499a-863a-292c1bdbc7d2 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Safer or luckier? LLM s as safety evaluators are not robust to artifacts
Reference 42
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6e88b490-7ac7-42b8-bd01-a1006ace5936 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Framework convention on artificial intelligence and human rights, democracy and the rule of law, cets no
Reference 43
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 78b1abef-d838-4919-bbf3-a414aa3b33fc · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI OR-Bench : An over-refusal benchmark for large language models
Reference 44
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2094cad3-422a-4397-b57a-778c04e442ec · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Secure by design, 2023
Reference 45
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 55f12536-1212-4ce2-a381-30e86bbb1725 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI General-purpose AI code of practice, 2025
Reference 46
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 639cf979-8b12-4301-9b24-158e56520398 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Regulation (eu) 2024/1689 laying down harmonised rules on artificial intelligence, 2024
Reference 47
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 00f30630-c510-4668-9650-8834e20cafac · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Selective classification for deep neural networks
Reference 48
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5d542346-135c-4686-9426-a8eb51539cfa · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Gemini 3 pro model card, 2026
Reference 49
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ae8f414f-5a20-4a1f-aecd-45874c484f85 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI An Empirical Study of Multi-Generation Sampling for Jailbreak Detection in Large Language Models
Reference 50
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d921dd8d-c965-4623-97ef-556a87ee48b0 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
Reference 51
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation df0c230d-5c14-4555-a6b6-4dc5eb2ba4e8 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI FORTRESS: Frontier Risk Evaluation for National Security and Public Safety
Reference 52
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation a4d97111-8556-4ca6-bd93-bd6f91f79161 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI PRISM 4.0: Verification of probabilistic real-time systems
Reference 53
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6ba8f4bc-4d22-4f68-b311-29d6b731947c · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI A brief account of runtime verification
Reference 54
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 262a9b1d-6929-40ec-89e7-43119cc7b77d · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI A holistic approach to undesired content detection in the real world
Reference 55
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 60a70377-0365-47ca-9a47-ab52212ae630 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI HarmBench : A standardized evaluation framework for automated red teaming and robust refusal
Reference 56
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 81970eb8-bc0d-435c-8243-2ab1ff0a3267 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Artificial intelligence risk management framework ( AI RMF 1.0)
Reference 57
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation f7875d81-c32f-4c5c-a375-8bb729e3450f · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Artificial intelligence risk management framework: Generative artificial intelligence profile
Reference 58
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 813e2c9f-0a7b-46e7-8ea0-5e5da20023c2 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI CAISI evaluation of DeepSeek ai models finds shortcomings and risks
Reference 59
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 8c2266c6-4d8b-4abe-a70d-78fdef69d4ca · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI GPT-5 system card, 2025
Reference 60
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation a44b9562-4e4e-4158-90da-effe1154a8df · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI GPT-5.5 system card, 2026
Reference 61
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1cc1b510-a081-4d37-a465-c8fa80c98ce4 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Evaluating Frontier Models for Dangerous Capabilities
Reference 62
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6c60f9e8-3e1f-4254-8244-8600bc86b4d8 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI NeMo guardrails: A toolkit for controllable and safe LLM applications with programmable rails
Reference 63
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 76462850-05c8-4e54-a52a-3e78fa128dd6 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI XSTest : A test suite for identifying exaggerated safety behaviours in large language models
Reference 64
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3a7feac4-3d05-441e-ba13-57d03795735e · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Saltzer and Michael D
Reference 65
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b4b4bc95-6f10-4c78-bee9-f8301d54e997 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
Reference 66
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 71c8b7a1-6495-44bc-a25b-fd2cd0dc4005 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Judging the judges: A systematic study of position bias in LLM -as-a-judge
Reference 67
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5df24c37-9aa2-4d7b-ba37-12d9d6864e5f · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI A StrongREJECT for empty jailbreaks
Reference 68
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 90034dc1-7a10-4e5c-bbaa-aca42dac7f46 · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Frontier AI safety commitments, AI seoul summit 2024, 2024
Reference 69
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5fa189bb-86f1-4bba-aff1-8d6c721b71fd · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Grok 4.1 model card, 2025
Reference 70
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ccf80d18-6a88-4ddf-bc67-ab897124dd4b · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI SORRY-Bench : Systematically evaluating large language model safety refusal
Reference 71
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 679131dd-d80e-44e7-9fc4-fcbb10eb3d2a · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI ShieldGemma: Generative AI Content Moderation Based on Gemma
Reference 72
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 20ab5d61-2eaa-4375-92a1-14c712e8dede · outbound
SAGE: Safety-First Defense-in-Depth Guardrails for Verified Lifecycle Control of High-Impact Generative AI Judging LLM -as-a-judge with MT-Bench and chatbot arena
Reference 73
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
No inbound Pith citation observations are available.