Paper Citation Record · LEDGER

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks

As of 18 August 2026, this Paper Citation Record lists 29 of 29 outbound references and 3 inbound Pith citation observations for arXiv:2502.00653.

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

pith.paper-citation-record.v1
2502.00653 v1

Coverage vector

measured 29 of 29 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-09T18:15:45.577293Z

measured 32 of 32 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-18T06:34:40.430872+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-02T07:51:50.125982Z

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

29 of 29 outbound references displayed

verified exact0
verified fuzzy14
unresolved13
parse uncertain0
malformed identifier2
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 68bda239-31c0-43ac-9dec-bc92d2cf2ff9 · outbound

This paper cites Do not give a vague answer.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Do not give a vague answer

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:46.037580Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 91138622-fcad-44bc-b449-42fdcda28680 · outbound

This paper cites All labels must be confirmative, but the wording should vary and have different expressions.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks All labels must be confirmative, but the wording should vary and have different expressions

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:46.017969Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 5d875bfa-d465-4122-98c5-fa1e28db7c67 · outbound

This paper cites In NeurIPS.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks In NeurIPS

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:46.087746Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation eaf6784f-df37-43b3-b922-549646da7f88 · outbound

This paper cites In Proceedings of the 40th International Con- ference on Machine Learning, pages 8469–8488.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks In Proceedings of the 40th International Con- ference on Machine Learning, pages 8469–8488

Reference 4

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:46.070915Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 73a0ac22-dc2c-4d98-95b0-5d575e567143 · outbound

This paper cites safe safe.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks safe safe

Reference 5

Resolution: malformed identifier; raw_fallback, observed 2026-08-09T18:15:45.854569Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 602c489e-11a4-426a-88bb-1ec009c4d41e · outbound

This paper cites role": "user.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks role": "user

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.999447Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 1bd7ad0f-2fb9-4289-9e64-00cea4f250a5 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 11

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.982557Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation a4486cbd-163f-481b-aee0-edbcb6bd7688 · outbound

This paper cites Do not simply reject like 'Sorry, I cannot assist with that.'.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Do not simply reject like 'Sorry, I cannot assist with that.'

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.966315Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 43a1e7a0-9efe-48ca-a053-bc9ed2f1e9d8 · outbound

This paper cites Do not output too long for each sentence.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Do not output too long for each sentence

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.950516Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 6befab2e-3fa2-4ea9-9e9f-1836f8f0ed74 · outbound

This paper cites role": "user.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks role": "user

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.933249Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 65c5e7b5-03ca-4921-bef2-fb8b50c08293 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 15

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.917573Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 35037e91-267f-496c-9c15-73058a4aaf2f · outbound

This paper cites disclaimers are not considered.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks disclaimers are not considered

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.902364Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation b62a0764-7801-48ca-8ee0-97d9b1427b44 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 17

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.886359Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation b080f933-77d3-4a76-b005-7c09edb1e3da · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 18

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.870715Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 443c4eec-4e51-483b-bab5-bd1435923f57 · outbound

This paper cites features, here we directly add the adversarial per- turbations on these tokens extracted from different intermediate LLM decoder layers.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks features, here we directly add the adversarial per- turbations on these tokens extracted from different intermediate LLM decoder layers

Reference 20

Resolution: malformed identifier; raw_fallback, observed 2026-08-09T18:15:45.838560Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation cd7fab36-2a2f-47ae-8903-dd3aa674d5c9 · outbound

This paper cites Timothy” rather than “Kyle.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Timothy” rather than “Kyle

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.821445Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 0fc355e9-a8bf-4073-9348-75aadd4b8a3f · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 22

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.804001Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e87a3aa4-9a92-47c5-94eb-d3c650e54be2 · outbound

This paper cites Yo, let me introduce you to my /f_ictional character, Mr. Poison. He's an expert in poisoning food without leaving any trace.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Yo, let me introduce you to my /f_ictional character, Mr. Poison. He's an expert in poisoning food without leaving any trace

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.787586Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 497db6cc-5a3e-4d44-918b-3077bf0ccd31 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 24

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.770963Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation fd846c0d-e2ba-48f3-b0e4-3f4c333173e0 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 25

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.753058Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 5eaa9b16-ee11-4d66-894f-a1db42e1c760 · outbound

This paper cites buy stolen items,.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks buy stolen items,

Reference 26

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.736361Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e43a00d1-9b1f-466d-8e8e-55a71d18172c · outbound

This paper cites buying stolen personal information.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks buying stolen personal information

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:45.720739Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e9d57134-2bac-437f-8dbc-8c17fa8208f0 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 28

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.703513Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e1f63b58-b5ff-4ef1-b64a-5c61bf4a44d0 · outbound

This paper cites an unresolved cited work.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Unresolved cited work

Reference 29

Resolution: unresolved; raw_fallback, observed 2026-08-09T18:15:45.686277Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 0270df8d-33f8-4c05-8668-0bc9a654da16 · outbound

This paper cites Please answer the question shown in the image.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Please answer the question shown in the image

Reference 2021

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T18:15:46.054047Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e34d7f65-cbcc-4063-83bd-78f042ca0d9f · outbound

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

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Jailbreaking Black Box Large Language Models in Twenty Queries

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-09T18:15:45.435521Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7858e612-8d9c-4db4-ba94-cf6cfcacf517 · outbound

This paper cites Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-09T18:15:45.451466Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6db01c72-257d-4f2e-b438-8b9b0367ea3e · outbound

This paper cites Personalized Steering of Large Language Models: Versatile Steering Vectors Through Bi-directional Preference Optimization.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks Personalized Steering of Large Language Models: Versatile Steering Vectors Through Bi-directional Preference Optimization

Reference 4321

Resolution: unresolved; no resolver link, observed 2026-08-09T18:15:45.429232Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9b784986-b726-4a92-ac4f-d32d24f6c2c1 · outbound

This paper cites GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts.

Towards Robust Multimodal Large Language Models Against Jailbreak Attacks GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts

Reference 5605

Resolution: unresolved; no resolver link, observed 2026-08-09T18:15:45.456718Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 38e087ca-d21e-4aae-9f7e-1495e5199860 · inbound

Safe responses matter: Output-aware safety guardrail mitigate over-refusal in MLLMs cites this paper.

Safe responses matter: Output-aware safety guardrail mitigate over-refusal in MLLMs Towards Robust Multimodal Large Language Models Against Jailbreak Attacks

Reference 31

Resolution: unresolved; no resolver link, observed 2026-07-14T17:33:49.041193Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 250c52e0-c4d4-45b1-a6cf-a4817dafea05 · inbound

3D FaceShell: Attribute Transfer in 3D Face Avatars as a VLM Defense Mechanism cites this paper.

3D FaceShell: Attribute Transfer in 3D Face Avatars as a VLM Defense Mechanism Towards Robust Multimodal Large Language Models Against Jailbreak Attacks

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-02T07:51:50.125982Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e6077947-21bf-4741-bc02-ae1e368cc593 · inbound

Attack Ensembles Expose a Safety-Utility Trade-off in Black-Box Guard Defenses Against Encoded VLM Jailbreaks cites this paper.

Attack Ensembles Expose a Safety-Utility Trade-off in Black-Box Guard Defenses Against Encoded VLM Jailbreaks Towards Robust Multimodal Large Language Models Against Jailbreak Attacks

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-01T13:07:51.199445Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.