mlsanityData to 7 Oct 2026

VLA-Mark: A cross modal watermark for large vision-language alignment model

arXiv 2507.14067 ↗First posted 18 Jul 2025cs.CV cs.AIcited by 11

On the map: Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness

also in 1 more
Abstract (arXiv)

Vision-language models demand watermarking solutions that protect intellectual property without compromising multimodal coherence. Existing text watermarking methods disrupt visual-textual alignment through biased token selection and static strategies, leaving semantic-critical concepts vulnerable. We propose VLA-Mark, a vision-aligned framework that embeds detectable watermarks while preserving semantic fidelity through cross-modal coordination. Our approach integrates multiscale visual-textual alignment metrics, combining localized patch affinity, global semantic coherence, and contextual attention patterns, to guide watermark injection without model retraining. An entropy-sensitive mechanism dynamically balances watermark strength and semantic preservation, prioritizing visual grounding during low-uncertainty generation phases. Experiments show 7.4% lower PPL and 26.6% higher BLEU than conventional methods, with near-perfect detection (98.8% AUC). The framework demonstrates 96.1% attack resilience against attacks such as paraphrasing and synonym substitution, while maintaining text-visual consistency, establishing new standards for quality-preserving multimodal watermarking

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 21 of its 32 references that are in the map. First posted on arXiv 18 Jul 2025.

Builds on

21 of 32 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025SafeEraser: Enhancing Safety in Multimodal Large Language Models through Multimodal Machine Unlearning18 Feb

Other references · 20

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  3. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,151
  4. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  5. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,669
  6. 2024DeepSeek-VL: Towards Real-World Vision-Language Understandingcited by 718
  7. 2023Provable Robust Watermarking for AI-Generated Textcited by 259
  8. 2024CogVLM2: Visual Language Models for Image and Video Understandingcited by 214
  9. 2023Who Wrote this Code? Watermarking for Code Generationcited by 116
  10. 2023A Robust Semantics-based Watermark for Large Language Model against Paraphrasingcited by 71
  11. 2024An Entropy-based Text Watermarking Detection Methodcited by 61
  12. 2022HiURE: Hierarchical Exemplar Contrastive Learning for Unsupervised Relation Extractioncited by 32
8 more
  1. 2024Watermarking Techniques for Large Language Models: A Surveycited by 20
  2. 2024Revisiting the Robustness of Watermarking to Paraphrasing Attackscited by 19
  3. 2025CoheMark: A Novel Sentence-Level Watermark for Enhanced Text Qualitycited by 17
  4. 2024Waterfall: Framework for Robust and Scalable Text Watermarking and Provenance for LLMscited by 14
  5. 2025Do BERT-Like Bidirectional Models Still Perform Better on Text Classification in the Era of LLMs?cited by 8
  6. 2025Unlocking Speech Instruction Data Potential with Query Rewritingcited by 4
  7. 2024Evaluating Durability: Benchmark Insights into Multimodal Watermarkingcited by 3
  8. 2025TreeHop: Efficient Embedding-Level Query Rewritercited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 10.

11 papers.

  1. 2025DocPruner: A Storage-Efficient Framework for Multi-Vector Visual Document Retrieval via Adaptive Patch-Level Embedding Pruningcited by 19
  2. 2026Vision-Language Introspection: Mitigating Overconfident Hallucinations in MLLMs via Interpretable Bi-Causal Steeringcited by 8
  3. 2026Sculpting the Vector Space: Towards Efficient Multi-Vector Visual Document Retrieval via Prune-then-Merge Frameworkcited by 8
  4. 2026Distilling the Thought, Watermarking the Answer: A Principle Semantic Guided Watermark for Large Reasoning Modelscited by 3
  5. 2026A Visual Semantic Adaptive Watermark grounded by Prefix-Tuning for Large Vision-Language Modelcited by 3
  6. 2026Decoding by Perturbation: Mitigating MLLM Hallucinations via Dynamic Textual Perturbationcited by 3
  7. 2026SIF: Semantically In-Distribution Fingerprints for Large Vision-Language Modelscited by 2
  8. 2026AGMark: Attention-Guided Dynamic Watermarking for Large Vision-Language Modelscited by 1
  9. 2026Latent-Mark: An Audio Watermark Robust to Neural Codec Compressioncited by 1
  10. 2026Adversarial Attacks for Good: A Survey of Proactive Protection across the Visual Content Lifecyclecited by 0
  11. 2026SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Breaking Semantic-Aware Watermarks via LLM-Guided Coherence-Preserving Semantic Injection
  2. 2025Removing Watermarks with Partial Regeneration using Semantic Information
  3. 2025Watermarking Degrades Alignment in Language Models: Analysis and Mitigation
  4. 2026WMVLM: Evaluating Diffusion Model Image Watermarking via Vision-Language Models
  5. 2024The Brittleness of AI-Generated Image Watermarking Techniques: Examining Their Robustness Against Visual Paraphrasing Attacks
  6. 2023WatME: Towards Lossless Watermarking Through Lexical Redundancy
  7. 2026FeatMark: Feature-level Watermark Protection against Mimicry Attacks with Diffusion Models
  8. 2024SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models