Shanghai, China

Annsnsna

Advanced
@Dtc7w3PQ

M.S. @ Shanghai AI Lab | LLM & MLLM Research | Constantly Learning and Reflecting Feel free to reach out~

Response-Attack. Official implementation of “Response Attack: Exploiting Contextual Priming to Jailbreak Large Language Models” (AAAI 2026).

37

Visco-Attack. Official implementation of Visco-Attack (EMNLP 2025 Main). An open-source one-click reproduction script is also provided.

31

PRCO. Official implementation of Seeing with You: Perception-Reasoning Co-evolution for Multimodal Reasoning.

30
3
Apply