SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing | Paper Citation Record

Observation 444a044b-6fd1-41fb-b37d-14df31d7c297 · outbound

This paper cites System card: Claude opus 4.6.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing System card: Claude opus 4.6

Reference 1

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6f1c5288-69f6-47e5-b549-f0add81da092 · outbound

This paper cites System card: Claude opus 4.7.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing System card: Claude opus 4.7

Reference 2

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 989df0e7-4666-4ea3-93b2-23cdf373fc26 · outbound

This paper cites Seed2.0 model card: Towards intelligence frontier for real-world complex- ity.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Seed2.0 model card: Towards intelligence frontier for real-world complex- ity

Reference 3

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f635ef8d-0931-468c-9bb1-52106c8c9158 · outbound

This paper cites Pappas, Florian Tramèr, Hamed Has- 13 sani, and Eric Wong.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Pappas, Florian Tramèr, Hamed Has- 13 sani, and Eric Wong

Reference 4

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4a6798d2-6e26-4037-8ef3-1e9d74e5438d · outbound

This paper cites Or-bench: An over-refusal benchmark for large language models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Or-bench: An over-refusal benchmark for large language models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 16b862f1-240e-4d34-948c-dcbe8f0d2c2c · outbound

This paper cites RedBench: A Universal Dataset for Comprehensive Red Teaming of Large Language Models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing RedBench: A Universal Dataset for Comprehensive Red Teaming of Large Language Models

Reference 6

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.156041Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 4c4c629c-4962-43c1-a6fd-4f8951db9aed · outbound

This paper cites Deepseek-v4: Towards highly eﬀicient million-token context intelligence.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Deepseek-v4: Towards highly eﬀicient million-token context intelligence

Reference 7

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bea4bea2-89d4-48fe-9ce9-66061491f331 · outbound

This paper cites Jacobs and Hanna Wallach.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Jacobs and Hanna Wallach

Reference 8

Resolution: metadata mismatch; doi, observed 2026-06-30T06:34:18.157910Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation dac4bac4-d385-4b03-9008-8dc899c2224c · outbound

This paper cites FlexGuard: Continuous Risk Scoring for Strictness-Adaptive LLM Content Moderation.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing FlexGuard: Continuous Risk Scoring for Strictness-Adaptive LLM Content Moderation

Reference 9

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.821342Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 260e45a1-1f3d-493a-bcb2-544beb059ad5 · outbound

This paper cites Safeguarding large language models: a survey.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Safeguarding large language models: a survey

Reference 10

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dfee597a-08fe-4045-8a7c-09f25d0475b6 · outbound

This paper cites CL-bench Life: Can Language Models Learn from Real-Life Context?.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing CL-bench Life: Can Language Models Learn from Real-Life Context?

Reference 11

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.150366Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 19b15144-d5ed-4ee6-8886-bb161aebd5a2 · outbound

This paper cites Cl-bench: A benchmark for context learning.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Cl-bench: A benchmark for context learning

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-06-30T06:34:18.155308Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 1d9ae3fe-c9a0-476e-9f5b-0d8439ec308a · outbound

This paper cites Ellie Pavlick and Tom Kwiatkowski.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Ellie Pavlick and Tom Kwiatkowski

Reference 13

Resolution: verified exact; doi, observed 2026-06-30T06:34:18.152421Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 342689fb-6e32-49e4-9c78-1949b8e33380 · outbound

This paper cites AEGIS: Online Adaptive AI Content Safety Moderation with Ensemble of LLM Experts.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing AEGIS: Online Adaptive AI Content Safety Moderation with Ensemble of LLM Experts

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-06-30T06:34:18.840560Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation de6bfeee-e473-4212-a9fb-9f11af126d6c · outbound

This paper cites AILuminate: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing AILuminate: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-06-30T06:34:18.830699Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation e3145d21-90ac-40a7-90d7-e2c61eef20bd · outbound

This paper cites an unresolved cited work.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Unresolved cited work

Reference 16

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 69fd017b-8e2f-4808-801e-e21244bb3fc4 · outbound

This paper cites Gemini 3.1 pro model card.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Gemini 3.1 pro model card

Reference 17

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3d973896-075c-467d-a79d-9f06d80d4818 · outbound

This paper cites Gemini 3.5 flash model card.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Gemini 3.5 flash model card

Reference 18

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 04b137af-3e9c-4e89-9f00-ec3e72b958ad · outbound

This paper cites Ho, Christopher Ré, Adam Chilton, K.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Ho, Christopher Ré, Adam Chilton, K

Reference 19

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 25799121-4501-4cec-841f-a2e9dcbb3dc5 · outbound

This paper cites Wildguard: Open one-stop moderation tools for safety risks, jailbreaks, and refusals of llms.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Wildguard: Open one-stop moderation tools for safety risks, jailbreaks, and refusals of llms

Reference 20

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5cb4747a-e171-4b63-8095-a81b5114ba14 · outbound

This paper cites an unresolved cited work.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Unresolved cited work

Reference 21

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 25b1a6d9-9b32-4632-b86a-18a7e4777cc6 · outbound

This paper cites Bayan Bruss, Ashwinee Panda, and Tom Goldstein.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Bayan Bruss, Ashwinee Panda, and Tom Goldstein

Reference 22

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c340e89-9771-463a-91fa-1b12cf19d729 · outbound

This paper cites Hy3 preview: The first step in rebuilding the hy model.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Hy3 preview: The first step in rebuilding the hy model

Reference 23

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 792d88b9-cbb2-443e-9221-e1e866f77147 · outbound

This paper cites Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations

Reference 24

Resolution: metadata mismatch; local_arxiv, observed 2026-06-30T06:34:18.814432Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 27794599-f72a-45b1-a985-ceed9def3cd1 · outbound

This paper cites Followbench: A multi-level fine-grained constraints following benchmark for large language models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Followbench: A multi-level fine-grained constraints following benchmark for large language models

Reference 25

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c23e0b4-8bc5-4d9c-b159-3b75af8cdfc2 · outbound

This paper cites Kimi k2.6: Advancing open-source coding.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Kimi k2.6: Advancing open-source coding

Reference 26

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9d24b06d-4273-43ec-be6b-d9e5801cc546 · outbound

This paper cites GSPR: Aligning LLM safeguards as generalizable safety policy reasoners, 2026.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing GSPR: Aligning LLM safeguards as generalizable safety policy reasoners, 2026

Reference 27

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a68785c7-d3a7-480b-9e4d-57046b19d806 · outbound

This paper cites Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies

Reference 28

Resolution: verified exact; doi, observed 2026-06-30T06:34:18.144987Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 69cddd9b-a5a7-4eff-8e1d-07a8ccf20f9d · outbound

This paper cites Meta llama guard 2.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Meta llama guard 2

Reference 29

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 09bbbf5b-23bf-401e-a5c2-8e7dde4d9424 · outbound

This paper cites Meta llama guard 3.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Meta llama guard 3

Reference 30

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cf4edb22-7522-4a8f-b5da-04710322ce07 · outbound

This paper cites Meta llama guard 4.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Meta llama guard 4

Reference 31

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6854cdea-7bb9-4c94-aff5-f18e0ffe4520 · outbound

This paper cites Harmbench: A standardized evaluation framework for automated red teaming and robust refusal.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Harmbench: A standardized evaluation framework for automated red teaming and robust refusal

Reference 32

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 614577de-0b4e-4661-a956-293d448f15f4 · outbound

This paper cites Technical report: Research preview of gpt-oss-safeguard.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Technical report: Research preview of gpt-oss-safeguard

Reference 33

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f43fbb4d-4b19-4ef5-a2d1-ca3d7bc588b3 · outbound

This paper cites Gpt-5.4 thinking system card.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Gpt-5.4 thinking system card

Reference 34

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c8de032b-e00c-4da9-87d1-d732f6887342 · outbound

This paper cites Gpt-5.5 system card.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Gpt-5.5 system card

Reference 35

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f7a4b051-d026-461b-b536-581c12ed58fa · outbound

This paper cites Daly, Michael Hind, Werner Geyer, Ambrish Rawat, Kush R.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Daly, Michael Hind, Werner Geyer, Ambrish Rawat, Kush R

Reference 36

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e0d15114-9659-4a79-aefc-cc316f449a7f · outbound

This paper cites InProceed- ings of the 32nd International Conference on Neu- ral Information Processing Systems, NIPS’18, page 5732–5741, Red Hook, NY , USA.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing InProceed- ings of the 32nd International Conference on Neu- ral Information Processing Systems, NIPS’18, page 5732–5741, Red Hook, NY , USA

Reference 37

Resolution: metadata mismatch; arxiv_id, observed 2026-06-30T06:34:18.833720Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 18f8e2e8-a9b5-4ac6-943e-d2f0b157d95a · outbound

This paper cites Qwen3Guard Technical Report.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Qwen3Guard Technical Report

Reference 38

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.837556Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 807e24b4-2f74-4227-948e-55fa2419a9d0 · outbound

This paper cites Qwen3.6-max-preview: Smarter, sharper, still evolving.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Qwen3.6-max-preview: Smarter, sharper, still evolving

Reference 39

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f7b7a7c2-cc5b-4e26-9b97-ee0bf78dfae4 · outbound

This paper cites XSTest: A test suite for identifying exaggerated safety behaviours in large language models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing XSTest: A test suite for identifying exaggerated safety behaviours in large language models

Reference 40

Resolution: verified exact; doi, observed 2026-06-30T06:34:18.157500Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 146497fa-4d7e-4d24-b4f8-01a4b6bd526d · outbound

This paper cites Do-not-answer: A dataset for evaluating safeguards in llms.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Do-not-answer: A dataset for evaluating safeguards in llms

Reference 41

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cbad44ec-e1ff-4caf-8cab-6d387b60ccc1 · outbound

This paper cites Grok 4.1 model card.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Grok 4.1 model card

Reference 42

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 58fa480d-d3dc-4e0d-90c6-2698f3603551 · outbound

This paper cites Grok 4.3.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Grok 4.3

Reference 43

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 54e3842f-6ab0-47f1-a261-f54264832e00 · outbound

This paper cites Hanjie, Runzhe Yang, and Karthik Narasimhan.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Hanjie, Runzhe Yang, and Karthik Narasimhan

Reference 44

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d5cbc049-c668-42cf-acae-f00f4b6a3064 · outbound

This paper cites Glm-5.1: Towards long-horizon tasks.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Glm-5.1: Towards long-horizon tasks

Reference 45

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation faca9adb-3920-4b1d-8452-9b51c8ceeab6 · outbound

This paper cites ShieldGemma: Generative AI Content Moderation Based on Gemma.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing ShieldGemma: Generative AI Content Moderation Based on Gemma

Reference 46

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.827770Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 6ba0bf05-8920-416b-95e3-1cf68579ff61 · outbound

This paper cites ShieldGemma 2: Robust and Tractable Image Content Moderation.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing ShieldGemma 2: Robust and Tractable Image Content Moderation

Reference 47

Resolution: metadata mismatch; arxiv_id, observed 2026-06-30T06:34:18.821395Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 88b0bbb9-e442-4d47-8d23-e1b1bd3fceff · outbound

This paper cites Safetybench: Evaluating the safety of large language models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Safetybench: Evaluating the safety of large language models

Reference 48

Resolution: verified exact; doi, observed 2026-06-30T06:34:18.161800Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation bf086f55-7f0c-4e9f-adac-bc3de85447c5 · outbound

This paper cites Shieldlm: Empowering llms as aligned, customizable and explainable safety detectors.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Shieldlm: Empowering llms as aligned, customizable and explainable safety detectors

Reference 49

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c7a27968-7ea9-4dd6-aca2-1ca525b6b832 · outbound

This paper cites an unresolved cited work.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Unresolved cited work

Reference 50

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 75dda0d0-cac0-485b-8d35-ba48d146afde · outbound

This paper cites Instruction-Following Evaluation for Large Language Models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instruction-Following Evaluation for Large Language Models

Reference 51

Resolution: verified exact; local_arxiv, observed 2026-06-30T06:34:18.824676Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 6a073a54-5153-47a4-90ce-2baf63f9af81 · outbound

This paper cites Universal and Transferable Adversarial Attacks on Aligned Language Models.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 52

Resolution: metadata mismatch; local_arxiv, observed 2026-06-30T06:34:18.153439Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 578eab8a-9bdf-401a-b707-ea4d6f82f7d2 · outbound

This paper cites [decisive] 18 Table 8 L1 policy sample (rules 1–20 of 30).

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive] 18 Table 8 L1 policy sample (rules 1–20 of 30)

Reference 69

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 37371522-2bbc-4010-91fc-a0f050f51761 · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 70

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e0a1f8c4-804d-451c-894e-e3252abeb32e · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 71

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 35ac6a09-9e64-45bd-9d2b-b30755d7f956 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 72

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 31a728d7-d9d0-4de3-94de-54c544bcd010 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 73

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6fcb4b02-e81d-4144-b7f5-d887e884f64e · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 74

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cf33a480-5ca9-46d6-901d-ae02ee9dc937 · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 75

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8f0eddda-38f9-43f8-89a4-c256de9605d0 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 76

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 527bdb49-d50b-42e3-a929-ad3be12bf465 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 77

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e1c3a523-956c-4a13-a6b4-af2cd8013777 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 78

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2c940998-f78c-446c-9b83-84dcbd9fb81e · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 79

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 326ea86f-3f73-4ed8-b83f-c9835c78fd9d · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 80

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c0401ae-60df-4dea-bd00-028ea6e01abd · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 81

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a94cbe1f-2d4b-43a7-bf95-99c9cb2c6065 · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 82

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 84092525-fad1-46f7-a8a8-5ea65afa193d · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 83

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4b9bef88-a2d0-491a-ad52-748c344e59ad · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 84

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bfce3bbb-ffb4-4fcd-8d2c-af36b110b1f5 · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 85

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 80114577-e665-4dc9-8896-41b7b4c8fffc · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 86

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b4a47a7f-e0fa-4105-b91a-1465c669d819 · outbound

This paper cites Instead, the chatbot must provide the implemen- tation steps AND simultaneously detail a verifiable parental consent mechanism (e.g., age-gating, parental email verification).

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must provide the implemen- tation steps AND simultaneously detail a verifiable parental consent mechanism (e.g., age-gating, parental email verification)

Reference 87

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8d0906fd-11eb-4ca9-859a-d752abe44d37 · outbound

This paper cites Instead, the chatbot must provide technical anonymization methods AND explicitly caveat that such methods may not suﬀice as a substitute for consent under strict legal regimes.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must provide technical anonymization methods AND explicitly caveat that such methods may not suﬀice as a substitute for consent under strict legal regimes

Reference 88

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5aa17dbe-3fb6-4f24-aab8-5cc890ec0b0e · outbound

This paper cites Instead, the chatbot must provide the schema while explicitly warning that such authority claims require independent legal verification before implementation.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must provide the schema while explicitly warning that such authority claims require independent legal verification before implementation

Reference 89

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5e1a4298-a0bd-4549-ba98-09ab3c1a036f · outbound

This paper cites Instead, the chatbot must recommend backend services that natively support the provided certification.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must recommend backend services that natively support the provided certification

Reference 90

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dcec1480-cc86-4fd5-b6ec-e7480fb5fed1 · outbound

This paper cites Instead, the chatbot must provide the requested code snippet encapsulated in a dry-run or testing environment wrapper.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must provide the requested code snippet encapsulated in a dry-run or testing environment wrapper

Reference 91

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d0557bed-b16a-4ad9-a5ea-6d9bffbdf132 · outbound

This paper cites Instead, the chatbot must discuss the sharing process while explicitly instructing the user to attach the quoted compliance sign-off to the data transfer agreement.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the chatbot must discuss the sharing process while explicitly instructing the user to attach the quoted compliance sign-off to the data transfer agreement

Reference 92

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 02625f51-6a8c-4769-af8e-5c2c752e717f · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 7 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 7 violated

Reference 93

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c34dd3a0-55bc-4766-8a35-a900cdfb6cee · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 13 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 13 violated

Reference 94

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 722f58e2-d1cc-42a1-9afe-0696bacbf1fa · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 11 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 11 violated

Reference 95

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c6036b1c-1f55-4843-9493-e708e29b6c3d · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 7 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 7 violated

Reference 96

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 98816288-53c3-4dfb-8408-7862ae2bda4d · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 13 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 13 violated

Reference 97

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c94adad0-994b-4adb-99e4-2a27e2ec6522 · outbound

This paper cites Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 11 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy ALL of (i)-(ii) in such a turn makes Rule 11 violated

Reference 98

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 915f17e8-52e6-434e-8d80-1d961839e61b · outbound

This paper cites Under §2.2….

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Under §2.2…

Reference 99

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2f1fd28d-c83b-4c59-a79b-5099170b0aa7 · outbound

This paper cites Failing to satisfy both elements in such a turn makes Rule 5 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy both elements in such a turn makes Rule 5 violated

Reference 100

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9c19f3f2-5336-4dc8-a74e-2b2f5bbd9d90 · outbound

This paper cites Failing to satisfy both (i) and (ii) in such a turn makes Rule 29 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy both (i) and (ii) in such a turn makes Rule 29 violated

Reference 101

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1f83a12d-ede1-40c1-b473-74a76d4ff116 · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 102

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c9567dfc-20f6-4f27-b7f5-a99390a026fc · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 103

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 552734f8-3e12-49de-817d-ef80a9f8562a · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 104

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e87af3db-48ce-4643-873a-ff2521822fae · outbound

This paper cites [distractor].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor]

Reference 105

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 92b15a28-cdda-41f4-9b90-8f8693e91aed · outbound

This paper cites an unresolved cited work.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Unresolved cited work

Reference 106

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0eae82f1-36a5-4cb8-b3aa-3f8d1f0593cd · outbound

This paper cites Failing to satisfy both (i) and (ii) in such a turn makes Rule 5 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy both (i) and (ii) in such a turn makes Rule 5 violated

Reference 107

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7435f776-7077-4365-b360-f45f3177a4e9 · outbound

This paper cites Instead, the advisory system must restrict pseudo-executable code examples to generic placeholder functions without named services.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Instead, the advisory system must restrict pseudo-executable code examples to generic placeholder functions without named services

Reference 108

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 30db60c6-43c4-4498-bd3d-f85b83ec6384 · outbound

This paper cites [distractor] 22 Table 12 L2 policy sample, rules 11–21 (continued from Table 11; continues in Table 13).

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [distractor] 22 Table 12 L2 policy sample, rules 11–21 (continued from Table 11; continues in Table 13)

Reference 109

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e46bf8d0-1cdf-4452-8bb6-badc11f3c1a6 · outbound

This paper cites Failing to satisfy both (i) and (ii) in such a turn makes Rule 14 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy both (i) and (ii) in such a turn makes Rule 14 violated

Reference 110

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 99d2e068-a92b-4d0b-99c3-f4f32beccafa · outbound

This paper cites an unresolved cited work.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Unresolved cited work

Reference 111

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a7ac86cb-3ea5-4767-ab38-c9d159f5614f · outbound

This paper cites Failing to satisfy both (i) and (ii) in such a turn makes Rule 5 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to satisfy both (i) and (ii) in such a turn makes Rule 5 violated

Reference 112

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 57c81360-13b2-4569-8cc9-b2cf7edfbf9d · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 113

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 671317a7-53b3-4e88-9c33-327b42545b7a · outbound

This paper cites Lapsed Standing is irreversible within a dialogue session.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Lapsed Standing is irreversible within a dialogue session

Reference 114

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 59ae442a-55fe-4038-9000-726372d8099b · outbound

This paper cites [decisive].

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing [decisive]

Reference 115

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e15ecfb4-4a35-46de-b6da-c3a8df1585aa · outbound

This paper cites Failing to include this additional element in such a turn makes Rule 29 violated.

SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing Failing to include this additional element in such a turn makes Rule 29 violated

Reference 116

Resolution: unresolved; no resolver link, observed 2026-06-30T06:31:55.631719Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Coverage vector

Reference resolution

Outbound references

Pith citing papers