Di Wang

Provable Responsible AI and Data Analytics (PRADA) Lab, King Abdullah University of Science and Technology

Papers on Lattice

Total citations

Topics

Publication activitypapers/week, last 8 weeks

Research focus

Constitutional AI & AI Ethics (1)Multimodal Models (1)Red-Teaming & Adversarial Robustness (1)

Frequent co-authors

Shu Yang (1)Lijie Hu (1)

Papers (1)

Mar 9, 2026

Provable Responsible AI and Data3d ago·also KAUST, MBZUAI

Visual Self-Fulfilling Alignment: Shaping Safety-Oriented Personas via Threat-Related Images

Fine-tuning VLMs on threat-related images alone can significantly improve safety without any explicit safety labels, revealing a surprising visual pathway for alignment.

Shu Yang, Lijie Hu, Di Wang

Constitutional AI & AI Ethics Multimodal Models Red-Teaming & Adversarial Robustness

Search

Di Wang

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (1)