I am a final-year Ph.D. student at CISPA Helmholtz Center for Information Security, supervised by Michael Backes. Prior to that, I obtained my bachelor’s (2018) and master’s (2021) degrees from the University of Science and Technology of China (USTC).

Research Interests

• Agentic RL
• LLM Post-Training & Alignment
• Certifiable Robustness Methods

Publications

• Jailbreaking Attacks vs. Content Safety Filters: How Far Are We in the LLM Safety Arms Race?
  Yuan Xin, Dingfan Chen, Linyi Yang, Michael Backes, Xiao Zhang. ACL 2026 Findings.

• Provably Cost-Sensitive Adversarial Defense via Randomized Smoothing
  Yuan Xin, Dingfan Chen, Michael Backes, Xiao Zhang. ICML 2025.

• Inside the Black Box: Detecting Data Leakage in Pre-trained Language Encoders
  Yuan Xin, Zheng Li, Ning Yu, Dingfan Chen, Mario Fritz, Michael Backes, Yang Zhang. ECAI 2024.

• Label Incorporated Graph Neural Networks for Text Classification
  Yuan Xin, Linli Xu, Junliang Guo, Jiquan Li, Xin Sheng, Yuanyuan Zhou. ICPR 2020.

Research Experience

• Summer 2021: NLP Research Intern, Alibaba DAMO
  • Use self-training methods to optimize the machine translation performance
  • Optimize the evaluation metrics for machine translation, let the model evaluate its own performance without external reference.
• Summer 2020: NLP Research Intern, Baidu Talent Intelligence Center
  • Extract hierarchical relations between skills in the JD dataset.
  • Skills representation learning