Kai Ye

PhD student in the Department of Statistics at LSE, researching reinforcement learning and alignment. CTO at Stats-Powered AI.

About me

I am a PhD student in the Department of Statistics at The London School of Economics and Political Science (LSE), advised by Chengchun Shi.
I am also the Chief Technology Officer (CTO) at Stats-Powered AI.
Prior to this, I completed an MSc in Applicable Mathematics at LSE.
Before that, I received a BSc in Mathematics with Finance from The University of Liverpool and Xi'an Jiaotong-Liverpool University.

AWS Academy Educator

Research

My current research focuses on the intersection of reinforcement learning and large language models, with particular interest in:

Reinforcement learning and policy optimization
Confounded and partially observable environments (POMDPs)
Human feedback alignment and preference-based learning
Doubly robust estimation and variance-reduction techniques
High-dimensional statistics and distributional analysis

Selected preprints

2026
Research

Can Tabular Foundation Models Amortize Statistical Inference?

Ye, K., Gong, S., Zhou, H., Zangirolami, V. and Shi, C.

Preprint 2026

TabCon: amortized confidence intervals with tabular foundation models.
2026
Research

Detecting LLM-Generated Tokens in Human–LLM Coauthored Text

Lu, Y., Zhou, H., Spill, F., Ye, K., Shi, C. and Zhu, J.

Preprint 2026

Token-level detection in text coauthored by humans and LLMs.
READER: Reasoning-Enhanced AI-Generated Text Detection

Su, P^*., Ye, K^*., Gong, S., Xu, E., Zhu, J., Livieri, G. and Shi, C.

Preprint 2026

READER.
Kernelized Advantage Estimation: From Nonparametric Statistics to LLM Reasoning

Gong, S^*., Ye, K^*., Zhu, J., Zhang, X., Zhou, H. and Shi, C.

Preprint 2026

KAE.
2026
Research

Demystifying Group Relative Policy Optimization: Its Policy Gradient is a U-Statistic

Zhou, H^*., Ye, K^*., Xu, E., Zhu, J., Yang, Y., Gong, S. and Shi, C.

Preprint 2026

Demystifying GRPO
Code ↗

Selected publications

Robust Reinforcement Learning from Human Feedback for Large Language Models Fine-Tuning

Ye, K^*., Zhou, H^*., Zhu, J^*., Quinzan, F. and Shi, C.

AOAS 2026 · Accepted

VRPO.
Code ↗
BASIS: Batchwise Advantage Estimation from Single-Rollout Information Sharing for LLM Reasoning

Gong, S^*., Xu, E^*., Ye, K^*., Livieri, G., Quinzan, F. and Shi, C.

EMNLP (Findings) 2026 · Accepted

BASIS: single-rollout information sharing for LLM reasoning.
A Difference-in-Difference Approach to Detecting AI-Generated Images

Qi, X^*., Ye, K^*., Shi, C., Yang, Y., Zhu, J. and Zhou, H.

CVPR 2026

DID detection for AI-Generated Images.
Code ↗
Learn-to-Distance: Distance Learning for Detecting LLM-Generated Text

Zhou, H^*., Zhu, J^*., Xu, E., Ye, K., Yang, Y. and Shi, C.

ICLR 2026

Learn-to-Distance.
Doubly Robust Alignment for Large Language Models

Xu, E^*., Ye, K^*., Zhou, H^*., Zhu, L., Quinzan, F. and Shi, C.

NeurIPS 2025

DRPO4LLM.
Code ↗
AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guarantees

Zhou, H^*., Zhu, J^*., Su, P., Ye, K., Yang, Y., Gavioli-Akilagun, S. and Shi, C.

NeurIPS 2025

AdaDetectGPT.
Code ↗

All publications and preprints →