Paper Citation Record · LEDGER

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection

As of 19 August 2026, this Paper Citation Record lists 27 of 27 outbound references and 2 inbound Pith citation observations for arXiv:2607.19829.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2607.19829 v1

Coverage vector

measured 27 of 27 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-01T11:40:29.288152Z

measured 29 of 29 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-19T06:32:44.657259+00:00

measured 2 of 2 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-12T20:49:20.302614Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: pith, observed 2026-08-06T10:46:11.686159Z

Reference resolution

27 of 27 outbound references displayed

verified exact0
verified fuzzy0
unresolved27
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 0cd9fb51-16b4-4a21-9632-2539ec7881a4 · outbound

This paper cites Jailbreaking Black Box Large Language Models in Twenty Queries.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Jailbreaking Black Box Large Language Models in Twenty Queries

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.308203Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d88f9d01-4cc3-4048-a3cc-ab972bdb7c47 · outbound

This paper cites Dynaguard: A dynamic guardian model with user-defined policies.arXiv preprint arXiv:2509.02563,.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Dynaguard: A dynamic guardian model with user-defined policies.arXiv preprint arXiv:2509.02563,

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.702036Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 10667c4e-b9a1-4623-8233-fb3168bb72d6 · outbound

This paper cites Toxicchat: Unveiling hidden challenges of toxicity detection in real-world user-ai conversation.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Toxicchat: Unveiling hidden challenges of toxicity detection in real-world user-ai conversation

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.111834Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9490cd8a-54f1-426e-b140-c686b44f30d2 · outbound

This paper cites Autodan-turbo: A lifelong agent for strategy self-exploration to jailbreak llms.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Autodan-turbo: A lifelong agent for strategy self-exploration to jailbreak llms

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.258662Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 221abdd0-51f2-4066-af83-0b87d3abe50a · outbound

This paper cites HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.294658Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7bee6b2c-cff6-44d5-870d-f6020aa95d81 · outbound

This paper cites Can a suit of armor conduct elec- tricity? a new dataset for open book question answering.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Can a suit of armor conduct elec- tricity? a new dataset for open book question answering

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.365443Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 854e725e-9c82-4c49-80bd-72bfce434bb3 · outbound

This paper cites Granite Guardian.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Granite Guardian

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.436778Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9b521d1e-c292-4b2c-9a38-8ad87089f048 · outbound

This paper cites Majic: Markovian adaptive jailbreaking via iterative composition of diverse innovative strategies.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Majic: Markovian adaptive jailbreaking via iterative composition of diverse innovative strategies

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.527425Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c5533a15-3c8c-4db2-8cf6-0ce30afd0641 · outbound

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

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Xstest: A test suite for identifying exaggerated safety behaviours in large language models

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.612489Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6045229e-334d-45e2-9e93-a652524a5b4c · outbound

This paper cites ” do anything now”: Characterizing and evaluating in-the-wild jailbreak prompts on large language models.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection ” do anything now”: Characterizing and evaluating in-the-wild jailbreak prompts on large language models

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.834194Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8970f135-f645-4ec0-b2f6-e59684d0f5b1 · outbound

This paper cites DataShield: Uncovering Risky Fine-Tuning Data Across LLMs Through Consensus Subspace Alignment.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection DataShield: Uncovering Risky Fine-Tuning Data Across LLMs Through Consensus Subspace Alignment

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.995842Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d58ce2cc-fe83-43d2-b06f-990d925cb16b · outbound

This paper cites Dynamic target attack.arXiv preprint arXiv:2510.02422,.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Dynamic target attack.arXiv preprint arXiv:2510.02422,

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:28.214780Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f957e25a-8ef6-4671-90fc-f4f47f355516 · outbound

This paper cites Harmmetric eval: Benchmarking metrics and judges for llm harmfulness assessment.arXiv preprint arXiv:2509.24384,.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Harmmetric eval: Benchmarking metrics and judges for llm harmfulness assessment.arXiv preprint arXiv:2509.24384,

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:28.396101Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fe54d41a-03dd-41e9-8a8b-b3d9702bcb8e · outbound

This paper cites Hotpotqa: A dataset for diverse, explainable multi-hop question answering.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Hotpotqa: A dataset for diverse, explainable multi-hop question answering

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:28.639197Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0dbe886c-34d6-43db-a75a-7fa7827063bc · outbound

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

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection ShieldGemma: Generative AI Content Moderation Based on Gemma

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:28.857968Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d3c2feac-4084-4a7f-9b3a-cb8a85d742be · outbound

This paper cites Evolving Skill-Structured Attack Memory Enhances LLM Jailbreaking.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Evolving Skill-Structured Attack Memory Enhances LLM Jailbreaking

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:29.039928Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 15f55aeb-07aa-46dd-8b40-4b1ee674b554 · outbound

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

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:29.288152Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6d40fd64-6589-481e-9e53-08b0119ad890 · outbound

This paper cites Qwen3Guard Technical Report.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Qwen3Guard Technical Report

Reference 2015

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:29.208494Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 395d7eb0-19b4-49a6-8ddb-268b82d5a129 · outbound

This paper cites YuFeng-XGuard: A Reasoning-Centric, Interpretable, and Flexible Guardrail Model for Large Language Models.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection YuFeng-XGuard: A Reasoning-Centric, Interpretable, and Flexible Guardrail Model for Large Language Models

Reference 2017

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.996387Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5ba3c237-a16f-4fd8-8a2c-e98f075c15ce · outbound

This paper cites Training Verifiers to Solve Math Word Problems.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Training Verifiers to Solve Math Word Problems

Reference 2018

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.463838Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1f6a45b4-2ba2-4116-afa2-68f9d5e552bf · outbound

This paper cites Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge

Reference 2019

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.383606Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 13b206d3-c391-4ac3-ba2d-c0fd199f33f4 · outbound

This paper cites Race: Large-scale reading comprehension dataset from examinations.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Race: Large-scale reading comprehension dataset from examinations

Reference 2020

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.932655Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6f17d31-b5a2-4056-926d-74a0b212b90c · outbound

This paper cites A wolf in sheep’s clothing: Generalized nested jailbreak prompts can fool large language models easily.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection A wolf in sheep’s clothing: Generalized nested jailbreak prompts can fool large language models easily

Reference 2021

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.561319Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c8def011-a8e3-4757-8c17-ccc5f25fcd72 · outbound

This paper cites Autodan: Generating stealthy jailbreak prompts on aligned large language models.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Autodan: Generating stealthy jailbreak prompts on aligned large language models

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:27.181672Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a3fe96a3-5ae9-4aa7-a047-2c6670dbf6b5 · outbound

This paper cites an unresolved cited work.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Unresolved cited work

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.624887Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 70930a7a-5005-46e0-ae70-a2fe46ce03fa · outbound

This paper cites Dualbreach: Efficient dual-jailbreaking via target-driven initialization and multi- target optimization.arXiv preprint arXiv:2504.18564,.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection Dualbreach: Efficient dual-jailbreaking via target-driven initialization and multi- target optimization.arXiv preprint arXiv:2504.18564,

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.781514Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e8aaf750-6a18-4fec-9160-935d9589216e · outbound

This paper cites NonTextual Target Attack.

DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection NonTextual Target Attack

Reference 2026

Resolution: unresolved; no resolver link, observed 2026-08-01T11:40:26.849024Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation afe0bdc5-8d00-441b-b918-89d842d108e4 · inbound

Gradient Immunity: Null-Space Resistance to Malicious Fine-Tuning cites this paper.

Gradient Immunity: Null-Space Resistance to Malicious Fine-Tuning DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection

Reference 97

Resolution: metadata mismatch; local_arxiv, observed 2026-08-06T10:46:11.691488Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 15258ecd-328c-47e5-99ee-d10dba8ba14f · inbound

ProbGuard: Calibrated Safety Risk Estimation from LLM Output Distributions cites this paper.

ProbGuard: Calibrated Safety Risk Estimation from LLM Output Distributions DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-12T20:49:20.302614Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.