Lattice AI Research

Research focus

Multimodal Models (2)Red-Teaming & Adversarial Robustness (2)Constitutional AI & AI Ethics (1)Eval Frameworks & Benchmarks (1)

Frequent co-authors

Zhiheng Li (1)Zongyang Ma (1)Yuntong Pan (1)Ziqi Zhang (1)

Papers (2)

Apr 8, 2026

UWApr 8, 2026·also CAS, Hellogroup, Jilin, ShanghaiTech +1

Making MLLMs Blind: Adversarial Smuggling Attacks in MLLM Content Moderation

MLLMs can be tricked into missing 90% of harmful content simply by encoding it in images that humans can easily read.

Zhiheng Li, Zongyang Ma, Yuntong Pan +6

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

Mar 31, 2026

Zhiqian Zhang +5Mar 31, 2026·also Hellogroup

Xuanwu: Evolving General Multimodal Models into an Industrial-Grade Foundation for Content Ecosystems

You don't need a massive model to beat Gemini-2.5-Pro in real-world content moderation: Xuanwu VL-2B achieves superior recall on policy-violating text using only 2B parameters.

Zhiqian Zhang, Xu Zhao, Xiaoqing Xu +3

Eval Frameworks & Benchmarks Multimodal Models Red-Teaming & Adversarial Robustness

Search

Xiaolei Lv

Research focus

Frequent co-authors

Papers (2)