Paper Citation Record · LEDGER

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

As of 6 August 2026, this Paper Citation Record lists 23 of 23 outbound references and 4 inbound Pith citation observations for arXiv:2602.02557.

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

pith.paper-citation-record.v1
2602.02557 v2

Coverage vector

measured 23 of 23 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-03T06:23:44.133687Z

measured 27 of 27 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-06T06:34:29.942622+00:00

measured 4 of 4 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-13T22:11:30.473038Z

measured 0 of 1 external citation measurements

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

Source: pith, observed 2026-07-04T13:59:52.855148Z

Reference resolution

23 of 23 outbound references displayed

verified exact0
verified fuzzy0
unresolved23
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 8df046ce-da43-4915-936f-f725516a4280 · outbound

This paper cites GPT-4 Technical Report.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer GPT-4 Technical Report

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:41.951699Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 91eecc3d-81ed-42a7-9447-deb39ab92dc3 · outbound

This paper cites Gigaspeech: An evolving, multi-domain asr corpus with 10,000 hours of transcribed audio.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Gigaspeech: An evolving, multi-domain asr corpus with 10,000 hours of transcribed audio

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.118875Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 915656ab-cbcd-4b88-b8c6-f189a0119684 · outbound

This paper cites Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.307716Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c1b9b95b-1e8f-4623-ac72-23544b9dd911 · outbound

This paper cites R., Wang, J.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer R., Wang, J

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.644560Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation de3a5a36-2e3b-4676-a565-384451ba3c6f · outbound

This paper cites GPT-4o System Card.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer GPT-4o System Card

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.840744Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7f86b5fa-fff4-42bf-8459-b32db5da6500 · outbound

This paper cites How do multimodal foundation models encode text and speech? an analysis of cross-lingual and cross-modal representations.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer How do multimodal foundation models encode text and speech? an analysis of cross-lingual and cross-modal representations

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.933572Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 197f8de6-c5d7-4ff5-ab03-f036496473fb · outbound

This paper cites Privacy in Large Language Models: Attacks, Defenses and Future Directions.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Privacy in Large Language Models: Attacks, Defenses and Future Directions

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.063047Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f9931894-d676-413b-8da9-73a5da3e1d54 · outbound

This paper cites Force: Transferable visual jailbreaking at- tacks via feature over-reliance correction.arXiv preprint arXiv:2509.21029,.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Force: Transferable visual jailbreaking at- tacks via feature over-reliance correction.arXiv preprint arXiv:2509.21029,

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.194497Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 46c46491-756c-4f5e-8702-ba3b54a0ed5d · outbound

This paper cites The trojan detection challenge.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer The trojan detection challenge

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.331291Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b71e13e9-7123-458d-aee5-2afb6f987467 · outbound

This paper cites Large language models 10 The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models encode semantics in low-dimensional linear subspaces.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Large language models 10 The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models encode semantics in low-dimensional linear subspaces

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.399901Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0c3ab26a-ad32-4cf4-8c3e-4cdc8e6ff0a7 · outbound

This paper cites Voice Jailbreak Attacks Against GPT-4o.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Voice Jailbreak Attacks Against GPT-4o

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.490072Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0e95eee1-6048-4e13-a200-873848c94e8b · outbound

This paper cites Interactiveomni: A unified omni- modal model for audio-visual multi-turn dialogue.arXiv preprint arXiv:2510.13747,.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Interactiveomni: A unified omni- modal model for audio-visual multi-turn dialogue.arXiv preprint arXiv:2510.13747,

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.671830Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 432e8ebe-e187-4091-93ad-ae6b1c7facf7 · outbound

This paper cites A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.757957Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1716ce80-123e-484e-a4f3-65e6dab0be23 · outbound

This paper cites VioLA: Unified Codec Language Models for Speech Recognition, Synthesis, and Translation.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer VioLA: Unified Codec Language Models for Speech Recognition, Synthesis, and Translation

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.836552Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 283535b4-0802-4c51-8287-ab7fc3714800 · outbound

This paper cites The devil behind the mask: An emergent safety vulnerability of diffusion llms.arXiv preprint arXiv:2507.11097,.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer The devil behind the mask: An emergent safety vulnerability of diffusion llms.arXiv preprint arXiv:2507.11097,

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.923790Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5013eebf-d67e-4b2d-8fe1-35cb12490c24 · outbound

This paper cites Qwen2.5-Omni Technical Report.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Qwen2.5-Omni Technical Report

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:44.002831Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cb9b45af-ff70-45d6-b6c1-fd34ca24822d · outbound

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

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:44.133687Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7854c359-8009-4d2b-8082-fe46114277eb · outbound

This paper cites A wolf in sheep’s clothing: Generalized nested 9 The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models jailbreak prompts can fool large language models eas- ily.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer A wolf in sheep’s clothing: Generalized nested 9 The Alignment Curse: Cross-Modality Jailbreak Transfer in Omni-Models jailbreak prompts can fool large language models eas- ily

Reference 1980

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.725683Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 41e7fcb8-0ca2-4b68-8f33-496b08e11164 · outbound

This paper cites Longcat-flash-omni technical report.arXiv preprint arXiv:2511.00279,.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Longcat-flash-omni technical report.arXiv preprint arXiv:2511.00279,

Reference 2012

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:43.603135Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9bbf1846-fe05-4519-9533-d54f16f67f75 · outbound

This paper cites VoiceBench: Benchmarking LLM-Based Voice Assistants.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer VoiceBench: Benchmarking LLM-Based Voice Assistants

Reference 2021

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.230799Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bcb6ba1c-84b1-4b6b-8aa7-e7424a487c92 · outbound

This paper cites Qwen2-Audio Technical Report.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Qwen2-Audio Technical Report

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.430421Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2a583085-1d1f-418e-bf42-73077b69c3f9 · outbound

This paper cites Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.549386Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 94a21d3a-536c-477f-9fd0-3cbbabd2bcbf · outbound

This paper cites FunAudioLLM: Voice Understanding and Generation Foundation Models for Natural Interaction Between Humans and LLMs.

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer FunAudioLLM: Voice Understanding and Generation Foundation Models for Natural Interaction Between Humans and LLMs

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-03T06:23:42.014632Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation fc922f0f-ad39-464a-9bc9-5eedeb66284a · inbound

On Optimizing Multimodal Jailbreaks for Spoken Language Models cites this paper.

On Optimizing Multimodal Jailbreaks for Spoken Language Models The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

Reference 22

Resolution: unresolved; no resolver link, observed 2026-07-13T22:11:30.473038Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 01295594-4342-46e3-b788-a3c736d88664 · inbound

$D^2$-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation-Aware Routing cites this paper.

$D^2$-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation-Aware Routing The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

Reference 38

Resolution: verified exact; local_arxiv, observed 2026-06-29T22:03:59.818695Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.

Observation 472196aa-c5c3-4adc-9893-58fd2e4bb509 · inbound

Audio Jailbreaks in Large Audio-Language Models: Taxonomy, Attack-Defense Analysis, and Cost-Aware Evaluation cites this paper.

Audio Jailbreaks in Large Audio-Language Models: Taxonomy, Attack-Defense Analysis, and Cost-Aware Evaluation The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

Reference 3

Resolution: metadata mismatch; local_arxiv, observed 2026-06-29T05:53:09.561486Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.

Observation 78cb61bc-e918-4467-a138-e874fcbfe1c4 · inbound

RedVox: Safety and Fairness Gaps in Speech Models Across Languages cites this paper.

RedVox: Safety and Fairness Gaps in Speech Models Across Languages The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

Reference 95

Resolution: verified exact; local_arxiv, observed 2026-07-04T13:59:52.856907Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.