Chunlin Gong(巩春林)

Chunlin Gong(巩春林)

University of Minnesota

Hi, my name is Chunlin Gong (巩春林). I am currently a third-year undergraduate (senior) at the University of Minnesota, in the Department of Computer Science and Engineering, majoring in Computer Science, advised by Prof. Mattia Fazzini.

My primary research interests are Trustworthy Machine Learning and AI Safety. I study the security of internal routing mechanisms in large language models from a neuron-level perspective, and I explore alignment methods—including fine-tuning and RLHF—to better align models with safety principles.

I also had the privilege of interning at the Institute of Automation, Chinese Academy of Sciences (CASIA) and Shanghai AI Lab. I am deeply grateful for the guidance and support from Prof. Shu Wu, Dr. Xingcheng Xu, and Dr. Zhao Tong.

🔥 News

  • 2025.12:  🎉I joined the Shanghai AI Lab to conduct research on safety alignment strategies in collaboration with CASIA.
  • 2025.08:  🎉I joined the software engineering research group at the University of Minnesota, to study safetyissues related to logs.
  • 2025.05:  🎉 I joined the Institute of Automation, Chinese Academy of Sciences (CASIA), to research content Safety in social media.
  • 2024.9: 🏠Thanks to Dr. Yin Wang, School of Control Science and Engineering, Shandong University. Our project has been approved by the Shandong Provincial Natural Science Foundation! This will be the starting point for my research.

📝 Publications (* Equal Contribution)

Preprint
sym

CoT is Not the Chain of Truth: An Empirical Internal Analysis of Reasoning LLMs for Fake News Generation

Zhao Tong*,Chunlin Gong*, Yiping Zhang, Haichao Shi, Qiang Liu, Xingcheng Xu, Shu Wu, Xiao-Yu Zhang

  • Our paper shows that in fake-news generation, reasoning LLMs can still produce unsafe, deceptive content inside their chain-of-thought even when the final answer is a refusal, and proposes a layer/head-level Jacobian spectral analysis to localize the safety-critical routing mechanisms driving that divergence.
Preprint
sym

Group-Adaptive Adversarial Learning for Robust Fake News Detection Against Malicious Comments

Zhao Tong*, Chunlin Gong*, Yimeng Gu, Qiang Liu, Shu Wu, Haichao Shi, Xiao-Yu Zhang

  • This paper demonstrates that fake-news detectors are highly vulnerable to psychological-based malicious adversarial comments and proposes group-adaptive adversarial training to substantially improve robustness.
Preprint
sym

OptFormer: Optical Flow-Guided Attention and Phase Space Reconstruction for SST Forecasting

Yin Wang*, Chunlin Gong*, Zhuozhen Xu, Lehan Zhang

  • This paper combines phase-space reconstruction with optical-flow–guided attention to better track dynamic regions.

🎖 Honors and Awards

  • 2025 University of Minnesota UROP Scholarship💴
  • 2024 Shandong Provincial Natural Science Foundation💴, Funded (Sole student member).

  • 2023 National Bronze Medal🥉, The International Collegiate Programming Contest (ACM-ICPC).

  • 2022 Second Prize🥈, Shandong Division, National Olympiad in Informatics(NOIP).

📖 Educations

  • 2023-2026, University of Minnesota, Twin Cities.
  • 2020-2023, Shandong Experimental High School.

💻 Research Experience

LLM Research Intern · Shanghai AI Lab

December 2025 – May 2026 · Shanghai, China · Mentors: Dr. Xingcheng Xu.

  • Collaborate with CASIA to analyze and develop defense strategies for multimodal large-scale safety issues.


Research Assistant · University of Minnesota

August 2025 – May 2026 · Minneapolis, United States · Mentors: Prof. Mattia Fazzini

  • Research on software engineering security focuses on anomaly detection in log files.


Algorithm Intern · Institution of Automation, Chinese academy of science

May 2025 – May 2026 · Beijing, China · Mentors: Prof. Shu Wu and Dr. Zhao Tong

  • Content Safety in social media and LLM safety