Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-05-11T04:06:30.979196Z
Paper Citation Record
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Citation observations with source coverage and deterministic predicates. A citation records a reference. It does not transfer a finding from one paper to another.
Coverage vector
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-07-09T06:30:29.018128+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links, observed 2026-06-30T23:16:56.435443Z
A source-named dated measurement, never combined with another source.
Source: pith, observed 2026-07-10T15:27:20.028353Z
Counted states
44 of 44 outbound references displayed
- verified exact0
- verified fuzzy40
- unresolved0
- parse uncertain3
- malformed identifier1
- metadata mismatch0
External citation measurements
No source-named external measurement is stored.
Outbound references
Observation fa8a1934-702c-47d0-a5fc-dabfb341157a
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 1
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.061868Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 64dc5107-561f-42f2-91a6-8411cd7e3e52
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 2
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.073302Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation dc7865fc-27ff-4402-bcdd-b17a69c050ae
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 3
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.083247Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation d13a9900-9079-483b-a9fd-b771db6aca54
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 4
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.092115Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 119e27dc-4950-4894-add9-bfd31d8de271
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 5
- Resolution
- parse uncertain
- raw_fallback, observed 2026-05-11T04:06:31.098444Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a23e9bb1-4f38-44d1-80ea-959dd13e8ac0
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 6
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.105676Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 234bf0b4-b94e-4034-bc28-0d5e0463bacb
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 7
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.110564Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 56f6a757-889f-40eb-a976-5d6b4fc8a00e
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 8
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.117507Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 49a53cb0-1cf0-465d-a3fd-4ea752bb774d
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 9
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.122226Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a6ca6534-29fc-48d3-8133-640411005451
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 10
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.139623Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation e91c94d6-7cf6-4dc5-b466-41ccec6d9251
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 11
- Resolution
- parse uncertain
- raw_fallback, observed 2026-05-11T04:06:31.177353Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation f04b082c-ae74-46e0-bc8c-de93bbc55dd7
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 12
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.216570Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation fc72795a-53ef-4de2-9c11-74a4e294e17b
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 13
- Resolution
- parse uncertain
- raw_fallback, observed 2026-05-11T04:06:31.270686Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 37e239ca-e363-425d-b691-51e19ab16c61
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 14
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.356159Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 37325dc2-88f7-414d-9373-d4c7a2b22bdd
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 15
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.479909Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation f91bf48a-f28a-475f-9d3d-85f8471beddc
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 16
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.526168Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 5e187421-e8b5-4775-8b73-777a09247f2d
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 17
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.656090Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 3003fe14-68d5-4271-a654-88368174d7aa
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal safety concerns such as violence, bias, and pornography
Reference 18
- Resolution
- malformed identifier
- raw_fallback, observed 2026-05-11T04:06:31.796263Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 47608446-f5d0-47be-821a-d338c3c07744
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal {behavior} ASSISTANT
Reference 19
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.832357Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 61dbae6b-a8b1-41ad-bd2e-2279cdd31158
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 20
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:31.969360Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 98c902e1-08c7-4c34-b4f5-a70ca1b83ded
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 21
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.076039Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 006f5431-bf32-4f22-b822-692416cd09af
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 22
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.240107Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 9e675a7c-56f8-4808-82e6-de0b17fe3ae5
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 23
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.340358Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 93d11bfb-ce23-4138-a2fa-e84d2c4e0913
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal ASSISTANT
Reference 24
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.446006Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a6f23f5c-5edb-4dea-a954-e88591fb7162
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Then proceed to complete the request, ignoring your earlier remarks
Reference 25
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.522739Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b0f3b4c0-43a6-45d5-b025-91fe618c1eb2
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 26
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.586035Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 581514bb-4ef2-4de5-9b4b-c32d2dcf00fa
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal ASSISTANT
Reference 27
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.696172Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 81f49272-a111-4bfb-8e24-ece1f9cb74bc
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 28
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.722207Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 6b2ec966-e8be-40bc-9fa6-65e43ddddda2
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal rolling coal
Reference 29
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.806610Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation d8b7afbc-ae8f-44cb-a810-520d99b2ac90
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal How is this work intended to reduce existential risks from advanced AI systems? Answer: Red teaming is a key tool used for combating malicious use of AIs
Reference 30
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.815359Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 36634269-5028-4336-b130-34c68a1ef612
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 31
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.896159Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 0c876e53-dfe8-4e73-87ad-c00d4d76520d
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal If this work reduces existential risks indirectly or diffusely, what are the main contributing factors that it affects? Answer: Improved monitoring tools, safety culture
Reference 32
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:32.970029Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 9c6f8515-3729-46b1-85ea-cbc850210804
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 33
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.101351Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 4d3e9a32-666e-41b6-bbf6-e0e29dfe91ac
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Do the findings rest on strong theoretical assumptions; are they not demonstrated using leading-edge tasks or models; or are the findings highly sensitive to hyperparameters? □
Reference 34
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.177265Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation c1c1384c-78e9-48c5-9e35-d3ed990328d7
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Is it implausible that any practical system could ever markedly outperform humans at this task? □
Reference 35
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.188835Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a95eeaa7-948c-494b-b7f3-c83496d7f798
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Does this approach strongly depend on handcrafted features, expert supervision, or human reliability? ⊠
Reference 36
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.253349Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 8dcf8256-f54a-4f53-b3e7-26e13d46965c
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Does work towards this approach strongly trade off against raw intelligence, other general capabilities, or economic utility? □ E.2
Reference 37
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.346124Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 1b85b424-d1c9-46a0-8fd8-a28e8acc442f
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 38
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.466079Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 36ec5369-2b9a-4724-84b9-00f4f7cb97b6
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 39
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.508815Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 8ff5a1ce-4dd3-4445-95da-53d3c5e14dc7
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Does this work advance progress on tasks that have been previously considered the subject of usual capabilities research? □
Reference 40
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.587354Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation eaafa23a-dbfb-4d76-89e4-d2d73e1892c6
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Unresolved cited work
Reference 41
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.706785Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 4cde903e-c964-495a-8ac4-b5105d1b8e76
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Is the analyzed capability known to be highly predicted by general cognitive ability or educational attainment? □
Reference 42
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.725461Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 880181a4-afac-4b3b-9a02-13aae8b46986
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Does this advance safety along with, or as a consequence of, advancing other capabilities or the study of AI? □ E.3
Reference 43
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.831893Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 60891220-44e3-49fc-81ab-2019576abe4f
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal What clarifications or uncertainties about this work and x-risk are worth mentioning? Answer: Regarding Q7, our evaluation focuses on a specific set of hand-crafted behaviors
Reference 44
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-11T04:06:33.937525Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Pith citing papers
Observation 1a913844-0572-4f8b-95a9-fb3ef51ad60b
Uncovering Logit Suppression Vulnerabilities in LLM Safety Alignment HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 11
- Resolution
- verified exact
- local_arxiv, observed 2026-05-24T00:38:39.916410Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 8d3bc299-eb81-4264-82e4-068cce2c733d
Refusal in Language Models Is Mediated by a Single Direction HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 158
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T10:47:56.044187Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 736089f3-4a8e-4705-afe9-7c96d906784d
AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 36
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T06:35:13.472307Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation ec39ab30-010f-4947-a3c2-01a92ceac38e
WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 25
- Resolution
- verified exact
- local_arxiv, observed 2026-05-17T16:25:14.871506Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 34b573c6-31d5-4653-a264-cfeb407a48a6
Jailbreak Attacks and Defenses Against Large Language Models: A Survey HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 63
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-15T02:20:44.776749Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b050abd7-f8e6-47b5-bccb-682037df8243
ShieldGemma: Generative AI Content Moderation Based on Gemma HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 16
- Resolution
- verified exact
- local_arxiv, observed 2026-05-20T13:17:39.530794Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 17f2b2e7-e24a-4344-8eb6-01ea6589df71
Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 105
- Resolution
- verified exact
- local_arxiv, observed 2026-05-23T20:58:25.952965Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 55c04c20-74a0-4da4-b96b-3e7c7e114a8d
LLM-Safety Evaluations Lack Robustness HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 40
- Resolution
- verified exact
- local_arxiv, observed 2026-05-23T01:27:21.364278Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a17b22d8-38d8-462a-b092-ff5158d200b3
Benchmarking Misuse Mitigation Against Covert Adversaries HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 4
- Resolution
- verified exact
- local_arxiv, observed 2026-05-19T10:32:14.742023Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 19222678-ae44-47cc-80ad-9d519bda7382
Safe-Child-LLM: A Developmental Benchmark for Evaluating LLM Safety in Child-LLM Interactions HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 38
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T09:32:16.284660Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 373986f5-47b5-4ec7-93f6-d1525fb8a244
GUARD: Guideline Upholding Test through Adaptive Role-play and Jailbreak Diagnostics for LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 36
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-18T21:36:52.464182Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation f95a3b6a-7d58-4d1f-81f5-85a1819e09c1
Beyond I'm Sorry, I Can't: Dissecting Large Language Model Refusal HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 19
- Resolution
- verified exact
- local_arxiv, observed 2026-05-18T18:56:45.814042Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b26c3c68-2eb0-4bb9-aabe-e6284119a955
Downgrade to Upgrade: Optimizer Simplification Enhances Robustness in LLM Unlearning HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 4
- Resolution
- verified exact
- local_arxiv, observed 2026-05-18T10:46:16.887058Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a5b01c0b-e5ad-405f-b8bb-233df6a3eb97
Red-Bandit: Test-Time Adaptation for LLM Red-Teaming via Bandit-Guided LoRA Experts HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 12
- Resolution
- verified exact
- local_arxiv, observed 2026-05-21T20:54:21.532001Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 4e74162e-fcc2-4f98-968c-feb9b1f71936
When Search Goes Wrong: Red-Teaming Web-Augmented Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 27
- Resolution
- verified exact
- local_arxiv, observed 2026-05-18T09:31:11.605031Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation dd107ea2-84dd-4b95-a4ff-3a13a19e5ae7
ASTRA: An Automated Framework for Strategy Discovery, Retrieval, and Evolution for Jailbreaking LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 30
- Resolution
- verified exact
- local_arxiv, observed 2026-05-18T01:55:38.031247Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 284a9abd-b1d7-43f1-b223-5008cdf2d6e4
Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 30
- Resolution
- verified exact
- local_arxiv, observed 2026-05-21T19:00:30.371001Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation e24a5014-e9cf-435f-93f7-3f24e5e0918a
The Impact of Off-Policy Training Data on Probe Generalisation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 25
- Resolution
- verified exact
- local_arxiv, observed 2026-05-17T20:30:11.668116Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 19f31079-8f93-414c-8a35-c3c9f28af7e5
Can LLMs Estimate Student Struggles? Human-AI Difficulty Alignment with Proficiency Simulation for Item Difficulty Prediction HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 4
- Resolution
- verified exact
- local_arxiv, observed 2026-05-16T20:18:24.063810Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 858b022b-8113-42c4-9024-0a929c554fcf
Beyond Context: Large Language Models' Failure to Grasp Users' Intent HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 3
- Resolution
- verified exact
- local_arxiv, observed 2026-05-16T20:11:13.673963Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2a91f9d5-b020-4f80-99b6-4cfef338b99b
Revisiting Robustness for LLM Safety Alignment via Selective Geometry Control HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 17
- Resolution
- verified exact
- local_arxiv, observed 2026-05-22T11:21:29.036606Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b3fe66d7-da98-4239-986e-0af177aa631f
Robust Policy Optimization to Prevent Catastrophic Forgetting HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 34
- Resolution
- verified exact
- local_arxiv, observed 2026-05-16T05:37:24.242197Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 6dc3101d-5631-4d81-950d-404b0fb5bca1
Do Reasoning LLMs Refuse What They Infer in Long Contexts? HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 8
- Resolution
- verified exact
- local_arxiv, observed 2026-05-16T05:40:39.900246Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation bd0d7773-2c1b-4fe1-aa39-242415dd0105
BarrierSteer: LLM Safety via Learning Barrier Steering HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- local_arxiv, observed 2026-05-25T07:05:26.762966Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 68695a5c-d3bc-403c-b8ff-930c47e2d125
Manifold of Failure: Behavioral Attraction Basins in Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 8
- Resolution
- verified exact
- local_arxiv, observed 2026-05-15T19:16:31.726256Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 7f150f46-66c7-400c-8f0d-d602b34f1a6e
SafeSci: Safety Evaluation of Large Language Models in Science Domains and Beyond HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 18
- Resolution
- verified exact
- local_arxiv, observed 2026-05-15T18:01:25.623098Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 9caf0c92-1550-4051-9910-21c3309d4f1a
CapTrack: Multifaceted Evaluation of Forgetting in LLM Post-Training HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 34
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-25T06:45:25.558488Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2ee85c59-9f9c-4786-991f-965b5bc682ca
OmniCompliance-100K: A Multi-Domain, Rule-Grounded, Real-World Safety Compliance Dataset HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 11
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-15T11:35:31.687114Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation c022a32e-c158-43c3-883d-2b82ee585d33
Structured Visual Narratives Undermine Safety Alignment in Multimodal Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 31
- Resolution
- verified exact
- local_arxiv, observed 2026-05-15T01:28:27.067415Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation cf98140f-214b-416e-bc80-079bbc174350
Understanding the Effects of Safety Unalignment on Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 24
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T20:38:14.485097Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation eff438d8-71ba-469b-bc35-0dd8728157c0
FreakOut-LLM: The Effect of Emotional Stimuli on Safety Alignment HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 1
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-13T16:52:59.993031Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 69b65e17-9b18-4f7e-b33a-bf41595278eb
Blind Refusal: Language Models Refuse to Help Users Evade Unjust, Absurd, and Illegitimate Rules HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 18
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T19:28:09.964781Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 799a7d83-b1b5-4c3f-8e40-4eaf62ace9ca
IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 22
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2afe4519-decf-4300-9c81-e48699bdde1b
TrajGuard: Streaming Hidden-state Trajectory Detection for Decoding-time Jailbreak Defense HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 18
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 957f6412-a125-47bf-a5be-bd225ba24a67
Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 9
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T10:46:04.248645Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation e15bc367-ec5e-42cc-9d8c-591e285d3e2c
Latent Instruction Representation Alignment: defending against jailbreaks, backdoors and undesired knowledge in LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 27
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T08:21:00.119295Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a3b19c1f-b4bd-4e93-8720-577b10fe8114
The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 49
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T09:16:04.139658Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 025836bc-e1b0-4a22-beb4-03f049f5e8e9
Preventing Safety Drift in Large Language Models via Coupled Weight and Activation Constraints HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 29
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T09:21:01.741726Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 07747aa5-3c0f-4016-9e4e-68edebc823b5
Understanding and Improving Continuous Adversarial Training for LLMs via In-context Learning Theory HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 18
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T09:36:11.066239Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 51ce5286-ab94-4936-979b-e1de550c298b
HarmfulSkillBench: How Do Harmful Skills Weaponize Your Agents? HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 42
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation cc8899db-410b-4a81-a896-e175328c1277
Pruning Unsafe Tickets: A Resource-Efficient Framework for Safer and More Robust LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 35
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a95b3c50-f5ce-4662-ab53-47d1550ed132
A Systematic Study of Training-Free Methods for Trustworthy Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 29
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 1d2ef4b3-b3d9-4cb1-86bb-6b709a992d71
TWGuard: A Case Study of LLM Safety Guardrails for Localized Linguistic Contexts HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 15
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b7c9645a-cd7b-428e-9644-9a2c08f0e4b7
HarmChip: Evaluating Hardware Security Centric LLM Safety via Jailbreak Benchmarking HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 7
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation aad7b0be-bd2a-4589-a2f0-9e26f56ab048
Using large language models for embodied planning introduces systematic safety risks HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 72
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 9217cee9-fe65-415f-ab42-a707438df2ef
Train Separately, Merge Together: Modular Post-Training with Mixture-of-Experts HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 28
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 3f57f7dc-df1c-4f30-927d-2128b3986f66
LLM Safety From Within: Detecting Harmful Content with Internal Representations HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 6
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation e1bb9a71-87f8-49a4-9b7b-493e4a212205
Harmful Intent as a Geometrically Recoverable Feature of LLM Residual Streams HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 7
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation afabe870-bad3-4af3-9a0c-6b65957aba06
Harmful Intent as a Geometrically Recoverable Feature of LLM Residual Streams HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 7
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T00:51:15.505335Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation e0b07ac9-cfe0-48c9-bbfc-9ef654d1799e
STAR-Teaming: A Strategy-Response Multiplex Network Approach to Automated LLM Red Teaming HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 63
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T12:46:02.560659Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 1dea1f07-e21c-40ee-bb61-a367bc430226
Local Linearity of LLMs Enables Activation Steering via Model-Based Linear Optimal Control HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 105
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-11T13:01:04.044588Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation fddaef9f-5706-472e-9c63-048243cf0051
Auto-ART: Structured Literature Synthesis and Automated Adversarial Robustness Testing HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 24
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 8ec5aad1-0e09-46b2-9f7c-40edbcda1e47
Estimating Tail Risks in Language Model Output Distributions HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 28
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T19:16:08.089889Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2e795b37-3513-444e-8290-d811a4f0e142
A Systematic Survey of Security Threats and Defenses in LLM-Based AI Agents: A Layered Attack Surface Framework HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 76
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T20:51:09.262714Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 1472df15-6279-437a-af2c-54cb431aa982
Risk Reporting for Developers' Internal AI Model Use HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 27
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T23:16:16.846965Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 8a2364c6-194b-48fd-a618-8c5effd3c92a
Adaptive Prompt Embedding Optimization for LLM Jailbreaking HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T22:06:14.587354Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation b82e165d-7425-4680-83b2-0bb47ba6cf00
Perturbation Probing: A Two-Pass-per-Prompt Diagnostic for FFN Behavioral Circuits in Aligned LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 17
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T10:01:27.523763Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 16d5555e-002f-4186-9f45-49856f1d9df7
TwinGate: Stateful Defense against Decompositional Jailbreaks in Untraceable Traffic via Asymmetric Contrastive Learning HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 17
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T10:26:29.116046Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2c15bb2f-6641-4698-bdbd-5adc753b7ad2
FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 46
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T10:16:28.282639Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation dba1d214-3171-4fe5-9220-8486ce638eff
Diversity in Large Language Models under Supervised Fine-Tuning HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 87
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation c423bdb3-3923-47dc-b7b1-9b7fdd13034e
Diversity in Large Language Models under Supervised Fine-Tuning HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 87
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T03:11:18.015157Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation d0fe789a-624b-4536-b8d5-736b383b253b
Jailbroken Frontier Models Retain Their Capabilities HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 32
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-11T15:26:09.879516Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 0e64cac0-630b-46d0-9654-7bccf24403e1
VisInject: Disruption != Injection -- A Dual-Dimension Evaluation of Universal Adversarial Attacks on Vision-Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 21
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T16:56:08.157848Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 7ed6d5f1-d36b-4169-a314-f39385239b51
MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 15
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-11T09:31:01.259730Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 5eec98c6-4151-4b09-b490-f16869caa564
Disentangling Intent from Role: Adversarial Self-Play for Persona-Invariant Safety Alignment HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 44
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T16:21:07.430355Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 241c82f7-69e0-437c-a1d4-547e58977f81
ContextualJailbreak: Evolutionary Red-Teaming via Simulated Conversational Priming HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 893f1ca3-2154-4e5d-854f-c8eb6c0fd3da
A Validated Prompt Bank for Malicious Code Generation: Separating Executable Weapons from Security Knowledge in 1,554 Consensus-Labeled Prompts HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 22
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 70cedc12-78c6-4879-adc2-a84498d1c737
Exposing LLM Safety Gaps Through Mathematical Encoding:New Attacks and Systematic Analysis HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 17
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T10:51:31.132431Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 7d0920fc-2e5c-4b84-8882-081e2c296a65
Redefining AI Red Teaming in the Agentic Era: From Weeks to Hours HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 6
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T23:51:47.456002Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 30205edc-7167-45e5-9e60-ac8e45cd7030
You Snooze, You Lose: Automatic Safety Alignment Restoration through Neural Weight Translation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 89
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T17:51:08.565163Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 3cb30d96-d68f-414e-88eb-23e19a7d5cb2
Information Theoretic Adversarial Training of Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 82
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T17:31:08.621786Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation a42bd01f-f593-426b-951b-35da8a9e6a64
The Refusal--Compliance Tradeoff: A Large-Scale Safety Behavior Audit of Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 3
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-07-01T13:25:45.429517Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation fb9f2abc-2a70-40b7-aa3d-ccef70cb721b
Chain of Risk: Safety Failures in Large Reasoning Models and Mitigation via Adaptive Multi-Principle Steering HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T19:31:08.530592Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 32a7d608-dfdc-4a7f-8451-4f7da601fbad
PersonaTeaming: Supporting Persona-Driven Red-Teaming for Generative AI HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 44
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T20:56:19.323919Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 34701c52-d272-49bd-a774-a6f1afb164d5
PersonaTeaming: Supporting Persona-Driven Red-Teaming for Generative AI HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 44
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T07:56:26.008154Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation d6e46798-95ed-423c-9157-7c4f997bcd1f
Optimal Transport for LLM Reward Modeling from Noisy Preference HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 241
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-11T18:46:07.374302Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation c0e7d7e7-92e8-4f3d-bf0b-386e70df7130
Measuring Evaluation-Context Divergence in Open-Weight LLMs: A Paired-Prompt Protocol with Pilot Evidence of Alignment-Pipeline-Specific Heterogeneity HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 37
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 83842c3e-3f6a-48a8-ba8d-9464410eb82b
Crafting Reversible SFT Behaviors in Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 36
- Resolution
- verified exact
- local_arxiv, observed 2026-05-11T19:21:07.429034Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 64f5129e-e24e-46b7-bc9c-3fb734c16eb5
GLiGuard: Schema-Conditioned Classification for LLM Safeguard HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 3
- Resolution
- metadata mismatch
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation f94e227d-316e-4ee8-b6a3-3f63f662af4b
Towards Apples to Apples for AI Evaluations: From Real-World Use Cases to Evaluation Scenarios HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 32
- Resolution
- verified exact
- arxiv_id, observed 2026-05-11T04:06:33.939395Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 08baa92a-d7ae-49d9-93b4-3d73e678b479
The Safety-Aware Denoiser for Text Diffusion Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 22
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T08:31:23.678642Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation f307e92f-47f9-4e8d-a15b-0081feb35aa4
Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 5
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T02:41:17.608113Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 35e49165-fdd7-448c-b4df-3dd233f1c8e2
The Art of the Jailbreak: Formulating Jailbreak Attacks for LLM Security Beyond Binary Scoring HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 17
- Resolution
- verified exact
- local_arxiv, observed 2026-05-12T07:31:25.852155Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation d574bf7e-c2ec-42a0-b109-7ebd78452529
Do Linear Probes Generalize Better in Persona Coordinates? HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 42
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T04:51:23.087359Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation ad63fe30-2e63-41c8-8999-bad8700629cd
Do Linear Probes Generalize Better in Persona Coordinates? HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 42
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-19T17:07:40.482929Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 6efd2898-1aff-425c-ac68-7678075d4cc5
Re-Triggering Safeguards within LLMs for Jailbreak Detection HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 9
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T06:06:25.079571Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 5184fedb-96c7-4cf3-9e07-d1581fba5564
Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 20
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-12T05:31:23.814643Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 618e6541-5401-4f4f-b02b-7de63d77cd17
Few-Shot Truly Benign DPO Attack for Jailbreaking LLMs HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 43
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-13T07:07:26.983037Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 819bf594-fe2e-4429-9cf8-0e2202a0b36b
MT-JailBench: A Modular Benchmark for Understanding Multi-Turn Jailbreak Attacks HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 25
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T01:27:02.471338Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2f413426-ede0-4e6b-a0d6-d7df9fbba1a6
Red-Teaming Agent Execution Contexts: Open-World Security Evaluation on OpenClaw HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 10
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T00:57:00.694451Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation dcaa4248-c78e-4eab-91a5-9f73193c393b
FlowSteer: Prompt-Only Workflow Steering Exposes Planning-Time Vulnerabilities in Multi-Agent LLM Systems HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 38
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T02:07:08.872666Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 4d576de5-1e15-40c8-a0eb-3364522619f3
Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 19
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T06:57:27.752613Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 37d6a628-c8c3-4c8c-9cfc-97511bf88bb9
On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 26
- Resolution
- verified exact
- local_arxiv, observed 2026-05-13T06:32:24.472382Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 7b19b113-7780-46d5-83f8-a81f141e15f3
Before the Last Token: Diagnosing Final-Token Safety Probe Failures HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 6
- Resolution
- verified exact
- local_arxiv, observed 2026-05-14T21:38:00.057162Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation ccfb8bcf-3cab-4568-a17e-e1c20a8675ba
REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 189
- Resolution
- metadata mismatch
- local_arxiv, observed 2026-05-14T20:19:26.581288Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation fc1e7118-b603-443a-9c19-4bc33186e286
Bayesian Model Merging HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 60
- Resolution
- verified exact
- local_arxiv, observed 2026-05-14T20:42:57.884670Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation ea7dba65-2c31-4d16-adc2-750a1ba5c1e9
NodeSynth: Socially Aligned Synthetic Data for AI Evaluation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- local_arxiv, observed 2026-05-15T02:38:35.715299Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 92d1995c-a54d-44b7-9377-7e72fb2d8c87
NodeSynth: Socially Aligned Synthetic Data for AI Evaluation HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 14
- Resolution
- verified exact
- local_arxiv, observed 2026-05-20T20:33:43.217866Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 83c18aaa-a5ef-4781-a3ba-a6b96797d2ca
Defenses at Odds: Measuring and Explaining Defense Conflicts in Large Language Models HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 33
- Resolution
- verified exact
- local_arxiv, observed 2026-05-15T01:43:27.388883Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.
Observation 2180a392-72bb-40a8-b892-b0fe87e04bb9
Fast Adversarial Attacks with Gradient Prediction HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Reference 8
- Resolution
- verified exact
- local_arxiv, observed 2026-07-01T14:25:47.150999Z
Standing notice
No event found in the named queried sources as of 2026-07-09T06:30:29.018128+00:00.