mlsanityData to 14 Oct 2026

Adversarial Attacks on Image Generation With Made-Up Words

arXiv 2208.04135 ↗First posted 4 Aug 2022cs.CV cs.CL cs.CRcited by 32

Not on the map. Why?

Abstract (arXiv)

Text-guided image generation models can be prompted to generate images using nonce words adversarially designed to robustly evoke specific visual concepts. Two approaches for such generation are introduced: macaronic prompting, which involves designing cryptic hybrid words by concatenating subword units from different languages; and evocative prompting, which involves designing nonce words whose broad morphological features are similar enough to that of existing words to trigger robust visual associations. The two methods can also be combined to generate images associated with more specific visual concepts. The implications of these techniques for the circumvention of existing approaches to content moderation, and particularly the generation of offensive or harmful images, are discussed.

Key facts. Cited by 32 papers in the map (data through 14 Oct 2026). Builds on 14 of its 21 references that are in the map. First posted on arXiv 4 Aug 2022.

Builds on

14 of 21 references.

References · 14

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  3. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,585
  4. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,355
  5. 2013Intriguing properties of neural networkscited by 5,155
  6. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,918
  7. 2022Scaling Autoregressive Models for Content-Rich Text-to-Image Generationcited by 1,295
  8. 2019Universal Adversarial Triggers for Attacking and Analyzing NLPcited by 722
  9. 2017Adversarial Examples for Evaluating Reading Comprehension Systemscited by 571
  10. 2021Multimodal datasets: misogyny, pornography, and malignant stereotypescited by 327
  11. 2021Bad Characters: Imperceptible NLP Attackscited by 71
  12. 2022Discovering the Hidden Vocabulary of DALLE-2cited by 43
2 more
  1. 2021Code-Mixing on Sesame Street: Dawn of the Adversarial Polyglotscited by 23
  2. 2022Deep Learning and Synthetic Mediacited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 10; 2024: 10; 2025: 6; 2026: 4.

32 papers.

Sort:

By area:

  1. 2023SneakyPrompt: Jailbreaking Text-to-image Generative Modelscited by 189
  2. 2023Sociotechnical Safety Evaluation of Generative AI Systemscited by 157
  3. 2024Latent Guard: a Safety Framework for Text-to-image Generationcited by 71
  4. 2025Safety at Scale: A Comprehensive Survey of Large Model and Agent Safetycited by 67
  5. 2024Adversarial Attacks and Defenses on Text-to-Image Diffusion Models: A Surveycited by 35
  6. 2023SurrogatePrompt: Bypassing the Safety Filter of Text-to-Image Models via Substitutioncited by 34
  7. 2022Exploiting Cultural Biases via Homoglyphs in Text-to-Image Synthesiscited by 29
  8. 2023Discovering Failure Modes of Text-guided Diffusion Models via Adversarial Searchcited by 25
  9. 2023Evaluating the Robustness of Text-to-image Diffusion Models against Real-world Attackscited by 24
  10. 2024Adversarial Nibbler: An Open Red-Teaming Method for Identifying Diverse Harms in Text-to-Image Generationcited by 20
  11. 2022Does CLIP Know My Face?cited by 19
  12. 2025Differentially Private Steering for Large Language Model Alignmentcited by 15
20 more
  1. 2023Harnessing LLM to Attack LLM-Guarded Text-to-Image Modelscited by 12
  2. 2024UPAM: Unified Prompt Attack in Text-to-Image Generation Models Against Both Textual Filters and Visual Checkerscited by 12
  3. 2024Revealing Vulnerabilities in Stable Diffusion via Targeted Attackscited by 9
  4. 2024A Survey on Responsible Generative AI: What to Generate and What Notcited by 8
  5. 2023Prompts have evil twinscited by 7
  6. 2024Autonomous LLM-Enhanced Adversarial Attack for Text-to-Motioncited by 7
  7. 2024DP-TLDM: Differentially Private Tabular Latent Diffusion Modelcited by 5
  8. 2025Unified Prompt Attack Against Text-to-Image Generation Modelscited by 5
  9. 2023Hypernymy Understanding Evaluation of Text-to-Image Models via WordNet Hierarchycited by 3
  10. 2023The Challenges of Image Generation Models in Generating Multi-Component Imagescited by 3
  11. 2024Discovering Universal Semantic Triggers for Text-to-Image Synthesiscited by 3
  12. 2025Erased, But Not Forgotten: Erased Rectified Flow Transformers Still Remain Unsafe Under Concept Attackcited by 3
  13. 2024Adversarial Attacks on Parts of Speech: An Empirical Study in Text-to-Image Generationcited by 2
  14. 2025DIVE: Inverting Conditional Diffusion Models for Discriminative Taskscited by 2
  15. 2025Towards Effective Prompt Stealing Attack against Text-to-Image Diffusion Modelscited by 2
  16. 2026MacPrompt: Maraconic-guided Jailbreak against Text-to-Image Modelscited by 2
  17. 2026PC2: Politically Controversial Content Generation via Jailbreaking Attacks on GPT-based Text-to-Image Modelscited by 1
  18. 2026Model Hypnosis: Strong control of AI via additive subliminal effectscited by 1
  19. 2023Trustworthy Large Models in Vision: A Surveycited by 0
  20. 2026Concept Score Relearning: A Unified Cross-Architecture Attack on Concept Erasurecited by 0