Paper Citation Record · LEDGER

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack

As of 17 August 2026, this Paper Citation Record lists 35 of 35 outbound references and 2 inbound Pith citation observations for arXiv:2505.21967.

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

pith.paper-citation-record.v1
2505.21967 v1

Coverage vector

measured 35 of 35 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T13:22:03.315478Z

measured 37 of 37 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-16T06:30:59.297886+00:00

measured 2 of 2 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-06-30T22:36:35.008744Z

measured 1 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-08-05T02:28:24.338817Z

Reference resolution

35 of 35 outbound references displayed

verified exact1
verified fuzzy0
unresolved34
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

1
arxiv_reference, observed 2026-08-05T02:28:24.338817Z

Outbound references

Observation d19d6d1c-9902-40bf-86c3-3486ab3c69af · outbound

This paper cites online" 'onlinestring :=.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack online" 'onlinestring :=

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-07T13:21:59.997430Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8d57f9d3-34a8-44a3-96a8-dd0bfd9aa116 · outbound

This paper cites write newline.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack write newline

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.083452Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4306a273-91d8-4e0f-bef0-f8bbdf768d8e · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 3

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:22:04.973791Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation eca866ee-afb6-4d8d-9af3-01bfed357942 · outbound

This paper cites Qwen2.5-VL Technical Report.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Qwen2.5-VL Technical Report

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.280789Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 553ad8f8-e473-4063-af66-0fd1741979ea · outbound

This paper cites Are aligned neural networks adversarially aligned?.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Are aligned neural networks adversarially aligned?

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.372780Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 08a26a18-0ea7-4d56-8163-881621a560ad · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 6

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:22:04.826249Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 71c06227-c851-4b21-a88d-dc330c50fe68 · outbound

This paper cites ETA: Evaluating Then Aligning Safety of Vision Language Models at Inference Time.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack ETA: Evaluating Then Aligning Safety of Vision Language Models at Inference Time

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.577809Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a75ab97f-f187-4f3f-9c6c-c72661c2fe55 · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 8

Resolution: verified exact; doi, observed 2026-08-07T13:22:04.260322Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 154a0da8-bb06-4901-b630-d98a7539b344 · outbound

This paper cites FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.808067Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5610b33f-03d5-463d-a71f-065bddc96ba1 · outbound

This paper cites Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:00.937985Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6238b638-3e4c-4bf3-beb4-1af93b5b1163 · outbound

This paper cites The VLLM Safety Paradox: Dual Ease in Jailbreak Attack and Defense.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack The VLLM Safety Paradox: Dual Ease in Jailbreak Attack and Defense

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.083283Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f2bf4b8-33ce-41c4-8b52-cbe9f76ad5f9 · outbound

This paper cites Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.209865Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2a43af3d-bcb8-4cfb-9482-2b6cd603c2ae · outbound

This paper cites Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Images are Achilles' Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodal Large Language Models

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.312248Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 191a6725-f006-4205-9b9b-2dbb1cddb02c · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.409051Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2b50626b-3742-48af-97b5-1c12fe9a4428 · outbound

This paper cites Visual Instruction Tuning.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Visual Instruction Tuning

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.498183Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 372b9fad-d7ae-415b-8e5c-2fa6374a698b · outbound

This paper cites MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.628548Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 14a7de28-2358-4d2e-a772-e43b6ba251de · outbound

This paper cites JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack JailBreakV: A Benchmark for Assessing the Robustness of MultiModal Large Language Models against Jailbreak Attacks

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.713276Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 175db9d6-728f-4140-a7eb-cb7830d3e6b1 · outbound

This paper cites Visual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Image Character.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Visual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Image Character

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:01.801888Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc0c3f8e-dab5-4a54-a16f-4c8b6da1a3c5 · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 19

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:22:04.658281Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 7e7c3462-64c6-4026-a5e5-68bae8996c5f · outbound

This paper cites an unresolved cited work.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Unresolved cited work

Reference 20

Resolution: unresolved; raw_fallback, observed 2026-08-07T13:22:04.512101Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 1ec2665d-f663-457b-b69e-de22614a81c1 · outbound

This paper cites Training language models to follow instructions with human feedback.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Training language models to follow instructions with human feedback

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.114092Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b8bad99f-9b0a-4851-beef-152ef6c85924 · outbound

This paper cites Visual Adversarial Examples Jailbreak Aligned Large Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Visual Adversarial Examples Jailbreak Aligned Large Language Models

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.206932Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c8c52f74-861c-4210-8aff-3a9380812a78 · outbound

This paper cites Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.264836Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c78856d-3fd0-4a59-9047-c400e615f3c3 · outbound

This paper cites Direct Preference Optimization: Your Language Model is Secretly a Reward Model.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Direct Preference Optimization: Your Language Model is Secretly a Reward Model

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.361804Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6dcc71f0-4f11-464c-91e1-7b1e3cfb7638 · outbound

This paper cites Proximal Policy Optimization Algorithms.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Proximal Policy Optimization Algorithms

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.468108Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 640678cd-a03e-41a8-b33a-01dfcf983ed4 · outbound

This paper cites Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.564374Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2695b31c-6fb7-47be-871f-bf585facd308 · outbound

This paper cites Safeguarding Vision-Language Models Against Patched Visual Prompt Injectors.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Safeguarding Vision-Language Models Against Patched Visual Prompt Injectors

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.650913Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d9418b88-1f0e-4957-9ca1-fd398b28f113 · outbound

This paper cites Safe Inputs but Unsafe Output: Benchmarking Cross-modality Safety Alignment of Large Vision-Language Model.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Safe Inputs but Unsafe Output: Benchmarking Cross-modality Safety Alignment of Large Vision-Language Model

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.762002Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7c17b61d-0b4d-4798-9adc-1d39b89c96a8 · outbound

This paper cites AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.837913Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a689702f-811f-42a2-956b-4bcb5dd45768 · outbound

This paper cites $\textit{MMJ-Bench}$: A Comprehensive Study on Jailbreak Attacks and Defenses for Multimodal Large Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack $\textit{MMJ-Bench}$: A Comprehensive Study on Jailbreak Attacks and Defenses for Multimodal Large Language Models

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.930271Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 66eaf8e9-4a0e-441f-bf8c-28a6b5c608d6 · outbound

This paper cites SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:02.983754Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 704c3b50-ed54-4bfa-821b-53c6d7106ce8 · outbound

This paper cites BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:03.045502Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ca0153f-f489-4660-824d-dca44f119fae · outbound

This paper cites Multimodal Situational Safety.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Multimodal Situational Safety

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:03.133709Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d7030a2a-6b04-40d9-aa36-666ba7ecdcdf · outbound

This paper cites Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:03.217330Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 223155c9-a9f7-4614-b43b-90df379ffc4a · outbound

This paper cites Image-to-Text Logic Jailbreak: Your Imagination can Help You Do Anything.

Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack Image-to-Text Logic Jailbreak: Your Imagination can Help You Do Anything

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-07T13:22:03.315478Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation f60e2885-fcfa-4e92-965f-45e4e2c9b8e8 · inbound

Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximization cites this paper.

Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximization Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack

Reference 28

Resolution: verified exact; arxiv_id, observed 2026-05-12T03:51:21.142226Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 6f56305f-00ec-4234-8302-71e500584cca · inbound

Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximization cites this paper.

Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximization Seeing the Threat: Vulnerabilities in Vision-Language Models to Adversarial Attack

Reference 28

Resolution: verified exact; arxiv_id, observed 2026-06-30T22:45:06.840605Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.