Paper Citation Record · LEDGER

Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

As of 9 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 38 inbound Pith citation observations for arXiv:2307.10490.

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

pith.paper-citation-record.v1
2307.10490 v4

Coverage vector

measured 0 of 0 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links

measured 38 of 38 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00

measured 38 of 38 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-08T12:07:35.853723Z

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

0 of 0 outbound references displayed

verified exact0
verified fuzzy0
unresolved0
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

14
arxiv_reference, observed 2026-08-05T02:28:24.338817Z

Outbound references

No outbound reference observations are available for this paper version.

Pith citing papers

Observation 8bcc565e-2fd1-4a78-95d7-86c2094b28be · inbound

Whispers in the Machine: Confidentiality in Agentic Systems cites this paper.

Whispers in the Machine: Confidentiality in Agentic Systems Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 52

Resolution: verified exact; arxiv_id, observed 2026-05-24T04:03:53.924075Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation ba6f02e6-5b5e-4765-a2d5-7ce7ab6c3470 · inbound

AI Safety Landscape for Large Language Models: Taxonomy, State-of-the-art, and Future Directions cites this paper.

AI Safety Landscape for Large Language Models: Taxonomy, State-of-the-art, and Future Directions Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 34

Resolution: verified exact; arxiv_id, observed 2026-05-23T21:55:50.756346Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 2586924b-d8a2-40a5-8ac9-2d56e349ac96 · inbound

Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety cites this paper.

Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 293

Resolution: verified exact; arxiv_id, observed 2026-05-23T04:42:34.252950Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 5fbed452-07e8-4d5e-bcb8-94d749fcab45 · inbound

A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations cites this paper.

A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 91

Resolution: unresolved; no resolver link, observed 2026-08-07T19:45:19.262296Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 72c7d2c3-637d-49f7-9744-6d1eec08e422 · inbound

RedDiffuser: Auditing Multimodal Safety Failures in Vision-Language Models via Reinforced Diffusion cites this paper.

RedDiffuser: Auditing Multimodal Safety Failures in Vision-Language Models via Reinforced Diffusion Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 25

Resolution: verified exact; arxiv_id, observed 2026-05-23T00:15:14.829234Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation e177162d-e68a-4e9f-a1e1-382d99dbc1e8 · inbound

Seven Security Challenges in Cross-domain Multi-agent LLM Systems cites this paper.

Seven Security Challenges in Cross-domain Multi-agent LLM Systems Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 75

Resolution: unresolved; no resolver link, observed 2026-08-07T13:05:34.852642Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e45d494d-1e18-483f-80c2-3892c392dc02 · inbound

Con Instruction: Universal Jailbreaking of Multimodal Large Language Models via Non-Textual Modalities cites this paper.

Con Instruction: Universal Jailbreaking of Multimodal Large Language Models via Non-Textual Modalities Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-07T12:07:58.079163Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 75da2dfd-2a96-4e97-9eda-f29973741e30 · inbound

Normative Conflicts and Shallow AI Alignment cites this paper.

Normative Conflicts and Shallow AI Alignment Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-07T10:40:54.417585Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 80107157-42c9-4510-a17b-5ec0990ce669 · inbound

Prompt Injection 2.0: Hybrid AI Threats cites this paper.

Prompt Injection 2.0: Hybrid AI Threats Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-06T16:34:05.101373Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6cd493d6-d804-492f-b72c-8cde5d1c5bb4 · inbound

Secure Tug-of-War (SecTOW): Iterative Defense-Attack Training with Reinforcement Learning for Multimodal Model Security cites this paper.

Secure Tug-of-War (SecTOW): Iterative Defense-Attack Training with Reinforcement Learning for Multimodal Model Security Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-06T12:09:39.603513Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f12926ff-74ab-4146-9cad-112b3fb9769a · inbound

Layer-Wise Perturbations via Sparse Autoencoders for Adversarial Text Generation cites this paper.

Layer-Wise Perturbations via Sparse Autoencoders for Adversarial Text Generation Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 111

Resolution: unresolved; no resolver link, observed 2026-08-05T20:31:43.258454Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b451794c-2a09-460f-9695-11d6ad1a2e58 · inbound

Invitation Is All You Need! Promptware Attacks Against LLM-Powered Assistants in Production Are Practical and Dangerous cites this paper.

Invitation Is All You Need! Promptware Attacks Against LLM-Powered Assistants in Production Are Practical and Dangerous Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-05T19:39:09.526632Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d2734992-083f-4ae0-b78f-89aaeff0a021 · inbound

Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges cites this paper.

Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 81

Resolution: verified exact; arxiv_id, observed 2026-05-18T03:42:22.045980Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f42a9e4d-5ec1-476b-9dd8-94feee168b2e · inbound

Prevalence of Security and Privacy Risk-Inducing Usage of AI-based Conversational Agents cites this paper.

Prevalence of Security and Privacy Risk-Inducing Usage of AI-based Conversational Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-04T07:01:14.163930Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8463fab9-571f-4e8e-b7dd-dabc3bfac939 · inbound

Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring cites this paper.

Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-05-16T22:31:19.422099Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 34480220-1d5d-4fee-98a5-820723d20d35 · inbound

A Synonymous Variational Perspective on the Rate-Distortion-Perception Tradeoff cites this paper.

A Synonymous Variational Perspective on the Rate-Distortion-Perception Tradeoff Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 20

Resolution: unresolved; no resolver link, observed 2026-07-12T20:09:57.922788Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ba77251d-9df2-4d85-8ca2-bdeb9251ed60 · inbound

Hijacking Large Audio-Language Models via Context-Agnostic and Imperceptible Auditory Prompt Injection cites this paper.

Hijacking Large Audio-Language Models via Context-Agnostic and Imperceptible Auditory Prompt Injection Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 20

Resolution: verified exact; arxiv_id, observed 2026-05-10T11:35:18.867051Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation e7e78990-4ca0-4ece-913f-9385457e017b · inbound

Temporal UI State Inconsistency in Desktop GUI Agents: Formalizing and Defending Against TOCTOU Attacks on Computer-Use Agents cites this paper.

Temporal UI State Inconsistency in Desktop GUI Agents: Formalizing and Defending Against TOCTOU Attacks on Computer-Use Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 16

Resolution: verified exact; arxiv_id, observed 2026-05-11T12:21:04.276860Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f7c90804-d766-4002-a5b8-6af4d28cd976 · inbound

MCP Pitfall Lab: Exposing Developer Pitfalls in MCP Tool Server Security under Multi-Vector Attacks cites this paper.

MCP Pitfall Lab: Exposing Developer Pitfalls in MCP Tool Server Security under Multi-Vector Attacks Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-05-11T14:31:07.944342Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 3d59fd30-10a5-4c33-a868-42015cdb1a8f · inbound

Ghost in the Agent: Redefining Information Flow Tracking for LLM Agents cites this paper.

Ghost in the Agent: Redefining Information Flow Tracking for LLM Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 4

Resolution: verified exact; arxiv_id, observed 2026-05-08T22:39:20.664808Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 86b07d54-192f-42a8-bf58-7f0780903955 · inbound

Semantic Denial of Service in LLM-controlled robots cites this paper.

Semantic Denial of Service in LLM-controlled robots Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-05-11T20:46:14.602650Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 2bf34282-1d7d-4985-99ff-d4341cef654e · inbound

From Prompt to Physical Actuation: Holistic Threat Modeling of LLM-Enabled Robotic Systems cites this paper.

From Prompt to Physical Actuation: Holistic Threat Modeling of LLM-Enabled Robotic Systems Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 25

Resolution: verified exact; arxiv_id, observed 2026-05-12T09:41:26.904379Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 8b73f468-d843-4c22-a309-b934f46880ed · inbound

VisInject: Disruption != Injection -- A Dual-Dimension Evaluation of Universal Adversarial Attacks on Vision-Language Models cites this paper.

VisInject: Disruption != Injection -- A Dual-Dimension Evaluation of Universal Adversarial Attacks on Vision-Language Models Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-05-11T16:56:08.007157Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation dd5b62e4-8956-45c1-8185-8601b0578196 · inbound

Laundering AI Authority with Adversarial Examples cites this paper.

Laundering AI Authority with Adversarial Examples Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 6

Resolution: verified exact; arxiv_id, observed 2026-05-11T17:41:08.017954Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b3141f92-532f-4220-9c54-27138fe399e7 · inbound

Cross-Modal Backdoors in Multimodal Large Language Models cites this paper.

Cross-Modal Backdoors in Multimodal Large Language Models Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-05-11T04:20:55.541381Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation a11d1780-b5a2-4e01-8618-3e765bf64265 · inbound

Hallucination as Exploit: Evidence-Carrying Multimodal Agents cites this paper.

Hallucination as Exploit: Evidence-Carrying Multimodal Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-05-20T09:48:11.621232Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 37c91339-e6ee-4f0d-a953-64b7410774ae · inbound

Hallucination as Exploit: Evidence-Carrying Multimodal Agents cites this paper.

Hallucination as Exploit: Evidence-Carrying Multimodal Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 8

Resolution: verified exact; arxiv_id, observed 2026-05-22T09:01:19.902347Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b2cc4d12-f2ca-4f45-ae83-c5cd23ffa2c7 · inbound

The Surface You Test Is Not the Surface That Breaks cites this paper.

The Surface You Test Is Not the Surface That Breaks Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 4

Resolution: verified exact; arxiv_id, observed 2026-06-29T14:33:31.206657Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation f1131dda-ded6-43fc-b755-4e2713447359 · inbound

HLL: Can Agents Cross Humanity's Last Line of Verification? cites this paper.

HLL: Can Agents Cross Humanity's Last Line of Verification? Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-07-01T22:56:19.607755Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 10cc6aee-624b-410a-a368-a9a0c102c966 · inbound

Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language Models cites this paper.

Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language Models Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 5

Resolution: verified exact; arxiv_id, observed 2026-07-02T03:16:34.719751Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation acb5163e-3814-4a3b-85b2-f480b497ffa0 · inbound

The Self-Correction Illusion: Role Relabeling Gates Explicit Error Flagging in Large Language Models cites this paper.

The Self-Correction Illusion: Role Relabeling Gates Explicit Error Flagging in Large Language Models Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-06-28T01:31:29.301820Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation bfa57969-40fc-4e2e-8e8c-8cc1183fea77 · inbound

The Self-Correction Illusion: Role Relabeling Gates Explicit Error Flagging in Large Language Models cites this paper.

The Self-Correction Illusion: Role Relabeling Gates Explicit Error Flagging in Large Language Models Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-03T02:14:12.243970Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e9c648c7-be23-4f06-805c-ed9d7d789a84 · inbound

Devil in the Lens: Analyzing and Defending Physical Prompt Injection Against Vision-Language Models on Wearable Devices cites this paper.

Devil in the Lens: Analyzing and Defending Physical Prompt Injection Against Vision-Language Models on Wearable Devices Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 1

Resolution: unresolved; no resolver link, observed 2026-07-14T13:02:42.673767Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3ca2b105-d4f9-4efc-a76a-fcd377754eae · inbound

Do Agents Dream of False Memories? Black-box Visual Attacks on Long-term Memory in Multimodal AI Agents cites this paper.

Do Agents Dream of False Memories? Black-box Visual Attacks on Long-term Memory in Multimodal AI Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-01T22:43:50.536410Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9b7750ee-0b58-4992-bd03-9af838bcdeb8 · inbound

Agent Security Needs Redefinition through a Holistic Framework cites this paper.

Agent Security Needs Redefinition through a Holistic Framework Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 295

Resolution: unresolved; no resolver link, observed 2026-08-01T06:04:46.731247Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 293a4c02-1cd1-473b-8c3d-18967550e9d9 · inbound

The Boy Who Cried Wolf: Adversarial Misclassification of Safe Inputs as Unsafe in Multimodal Guardrails cites this paper.

The Boy Who Cried Wolf: Adversarial Misclassification of Safe Inputs as Unsafe in Multimodal Guardrails Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-06T00:20:22.213214Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 646d89a3-18e4-4702-a6e0-1f7169807a60 · inbound

PromptShield Home: Ambient Multimodal Prompt Injection Defense for Smart-Home Agents cites this paper.

PromptShield Home: Ambient Multimodal Prompt Injection Defense for Smart-Home Agents Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-08T12:07:35.853723Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6e5903d9-8340-4ce3-87d2-e302054779fd · inbound

Hijacking Robots with a Piece of Paper: A Systematic Study of Physical Prompt Injection in VLM-Controlled Robots cites this paper.

Hijacking Robots with a Piece of Paper: A Systematic Study of Physical Prompt Injection in VLM-Controlled Robots Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-08T00:38:15.346908Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.