13 papers
Query-Adaptive Neuron Steering for LLMs
Yanxi Pan, Jingwen Wu, Yan Ma +8
Effectively eliciting the capabilities of large language models (LLMs) at inference time remains an important challenge. Existing studies address this challenge through context eng…
Can Tabular Foundation Models Amortize Statistical Inference?
Kai Ye, Shijin Gong, Hongyi Zhou +2
For decades, statistical inference has largely been developed one problem at a time. Given a scientific target, such as a treatment effect or a regression function, statisticians d…
Detecting LLM-Generated Tokens in Human--LLM Coauthored Text
Yangjun Lu, Hongyi Zhou, Fabian Spill +3
The rise of human-AI collaborative writing has created a growing need for fine-grained detection methods that support localizing likely LLM-generated content in mixed-authorship do…
Kernelized Advantage Estimation: From Nonparametric Statistics to LLM Reasoning
Shijin Gong, Kai Ye, Jin Zhu +3
Recent advances in large language models (LLMs) have increasingly relied on reinforcement learning (RL) to improve their reasoning capabilities. Three types of approaches have been…
Demystifying Group Relative Policy Optimization: Its Policy Gradient is a U-Statistic
Hongyi Zhou, Kai Ye, Erhan Xu +4
Group relative policy optimization (GRPO), a core methodological component of DeepSeekMath and DeepSeek-R1, has emerged as a cornerstone for scaling reasoning capabilities of large…
Learn-to-Distance: Distance Learning for Detecting LLM-Generated Text
Hongyi Zhou, Jin Zhu, Kai Ye +3
Modern large language models (LLMs) such as GPT, Claude, and Gemini have transformed the way we learn, work, and communicate. Yet, their ability to produce highly human-like text r…
A Difference-in-Difference Approach to Detecting AI-Generated Images
Xinyi Qi, Kai Ye, Chengchun Shi +3
Diffusion models are able to produce AI-generated images that are almost indistinguishable from real ones. This raises concerns about their potential misuse and poses substantial c…
Detecting LLM-Generated Text with Performance Guarantees
Hongyi Zhou, Jin Zhu, Ying Yang +1
Large language models (LLMs) such as GPT, Claude, Gemini, and Grok have been deeply integrated into our daily life. They now support a wide range of tasks -- from dialogue and emai…
AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guarantees
Hongyi Zhou, Jin Zhu, Pingfan Su +4
We study the problem of determining whether a piece of text has been authored by a human or by a large language model (LLM). Existing state of the art logits-based detectors make u…
Doubly Robust Alignment for Large Language Models
Erhan Xu, Kai Ye, Hongyi Zhou +3
This paper studies reinforcement learning from human feedback (RLHF) for aligning large language models with human preferences. While RLHF has demonstrated promising results, many…
Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation
Hongyi Zhou, Josiah P. Hanna, Jin Zhu +2
This paper studies off-policy evaluation (OPE) in reinforcement learning with a focus on behavior policy estimation for importance sampling. Prior work has shown empirically that e…
Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach
Jin Zhu, Jingyi Li, Hongyi Zhou +3
This paper focuses on the design of spatial experiments to optimize the amount of information derived from the experimental data and enhance the accuracy of the resulting causal ef…
Robust Reinforcement Learning from Human Feedback for Large Language Models Fine-Tuning
Kai Ye, Hongyi Zhou, Jin Zhu +2
Reinforcement learning from human feedback (RLHF) has emerged as a key technique for aligning the output of large language models (LLMs) with human preferences. To learn the reward…