Published June 3, 2026 | Version v1
Report Open

Cross-Domain Adversarial Transfer and Alignment Stability in Multimodal Models

Authors/Creators

  • 1. https://assignee.net

Description

This report synthesises findings from 8 peer-reviewed papers addressing the following research question: How does cross-domain adversarial transfer between image and text modalities affect the alignment stability of multimodal models under attack. 9 claims were extracted from source literature; 9 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.8/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: How does cross-domain adversarial transfer between image and text modalities affect the alignment stability of multimodal models under attack?

Autonomous literature synthesis. Automated review score: 8.8/10. Full text and citation available at Assignee Research.

Notes

Machine-generated literature synthesis. Content is derived from peer-reviewed papers; see individual sources for authoritative data. Automated review score: 8.8/10. Published by Assignee Research (https://assignee.net).

Files

paper.pdf

Files (73.5 kB)

Name Size Download all
md5:29d7761b29245f713fb4e8b024c74ae0
73.5 kB Preview Download

Additional details

Related works

Is compiled by
https://assignee.net (URL)