About me

Hi! I’m a third year Ph.D. student in the Department of Computer Science and Engineering at UC San Diego. I’m fortunate to be advised by Prof. Tsui-wei (Lily) Weng. My major research interest is in trustworthy machine learning. I hope to combine empirical success of deep networks and theoretical soundness of traditional machine learning methods to produce robust and trustworthy AI models. My CV could be found here.

Education

Ph. D. in the Department of Computer Science and Engineering at UC San Diego. [Sep. 2023 - Current]

M.S. in the Department of Electrical and Computer Engineering at UC San Diego. [Sep. 2021 - Mar. 2023]

B.S. in the School of Mathematical Science at Peking University. [Sep. 2017 - Jun. 2021]

Publications

Provably robust conformal prediction with improved efficiency.
Ge Yan, Yaniv Romano, Tsui-Wei (Lily) Weng, ICLR 2024.

VLG-CBM: Training Concept Bottleneck Models with Vision-Language Guidance.

Divyansh Srivastava^*, Ge Yan^*, Tsui-Wei (Lily) Weng, NeurIPS 2024.
^* Equal contribution.

ReflCtrl: Controlling LLM Reflection via Representation Engineering,
Ge Yan, Chung-En Sun, Tsui-Wei (Lily) Weng, NeurIPS 2025 MI workshop (Spotlight)

Faithful and Stable Neuron Explanations for Trustworthy Mechanistic Interpretability
Ge Yan, Tuomas Oikarinen, Tsui-Wei (Lily) Weng, NeurIPS 2025 MI workshop

ThinkEdit: Interpretable Weight Editing to Mitigate Overly Short Thinking in Reasoning Models
Chung-En Sun, Ge Yan, Tsui-Wei (Lily) Weng, EMNLP 2025.

Evaluating neuron explanations: A unified framework with sanity checks
Tuomas Oikarinen, Ge Yan, Tsui-Wei (Lily) Weng, ICML 2025.

Interpretable Generative Models through Post-hoc Concept Bottlenecks
Akshay Kulkarni, Ge Yan, Chung-En Sun, Tuomas Oikarinen, Tsui-Wei (Lily) Weng, CVPR 2025.

Multimodal Concept Bottleneck Models
Tongqing Shi, Ge Yan, Tsui-Wei (Lily) Weng, NeurIPS 2025 MI workshop.

ReFIne: A Framework for Trustworthy Large Reasoning Models with Reliability, Faithfulness, and Interpretability
Chung-En Sun, Ge Yan, Akshay Kulkarni, Tsui-Wei (Lily) Weng, NeurIPS 2025 MI workshop.

Beyond Top Activations: Efficient and Reliable Crowdsourced Evaluation of Automated Interpretability
Tuomas Oikarinen, Ge Yan, Tsui-Wei (Lily) Weng, NeurIPS 2025 MI workshop.

Internship

Applied Scientist Intern, Amazon, San Diego. [Jun. 2024 - Sep. 2024]

Data Scientist Intern, DiDi Technology, Beijing. [Jun. 2023 - Aug. 2023]

Data Analyst Intern at Shanghai Stock Exchange. [Aug. 2020 - Sep. 2020]