cv | Xiaonan Luo

Basics

Name	Xiaonan Luo
Label	Ph.D. Student in Computer Science
Email	xluo6@nd.edu
Url	https://xiaonan-debug.github.io
Summary	Ph.D. student at the University of Notre Dame, focusing on large language models with emphasis on post-training, evaluation, multi-agent systems, and scientific AI applications.

Education

2025.01 - Present

Notre Dame, IN, USA
Ph.D.

University of Notre Dame

Computer Science and Engineering
2020.09 - 2024.06

Hong Kong
Bachelor of Engineering

Hong Kong University of Science and Technology (HKUST)

Computer Science

Publications

2026.01.01

Better datasets start from refinelab: Automatic optimization for high-quality dataset refinement

AAAI Conference on Artificial Intelligence

Luo*, Xiaonan and Huang*, Yue and He, Ping and Zhang, Xiangliang. High-quality datasets are crucial for training effective machine learning models. This paper presents RefineLab, an automated framework for optimizing dataset refinement processes to improve data quality and model performance.
2025.12.01

ChemOrch: Empowering LLMs with Chemical Intelligence via Synthetic Instructions

Advances in Neural Information Processing Systems (NeurIPS)

Huang, Yue and Jiang, Zhengzhe and Luo, Xiaonan and Guo, Kehan and Zhuang, Haomin and Zhou, Yujun and others. ChemOrch presents a novel approach to enhancing large language models with specialized chemical knowledge through carefully designed synthetic instructions, enabling more accurate chemical reasoning and prediction.
2025.12.01

AdaReasoner: Adaptive Reasoning Enables More Flexible Thinking

Advances in Neural Information Processing Systems (NeurIPS) - Spotlight

Wang, Xiangqi and Huang, Yue and Wang, Yanbo and Luo, Xiaonan and Guo, Kehan and Zhou, Yujun and Zhang, Xiangliang. AdaReasoner introduces an adaptive reasoning framework that enables large language models to dynamically adjust their reasoning strategies based on problem complexity and context.
2025.07.01

Torpor: GPU-Enabled Serverless Computing for Low-Latency, Resource-Efficient Inference

USENIX Annual Technical Conference (ATC)

Yu, Mingyu and Wang, Ao and Chen, Dong and Yu, Haoxuan and Luo, Xiaonan and Li, Zhuzhong and others. Torpor presents a GPU-enabled serverless computing system that achieves low-latency, resource-efficient inference by intelligently managing GPU memory and computation resources.

Skills

	Machine Learning
	Large Language Models
	Data-Centric AI
	Deep Learning
	PyTorch
	TensorFlow
	Hugging Face Transformers

	Programming
	Python
	C/C++
	Java
	JavaScript
	Git
	Docker
	Linux
	CUDA

Languages

	Chinese
	Native speaker

	English
	Fluent

Interests

	Research
	LLM Post-Training & Evaluation
	Multi-Agent Systems
	Scientific AI

Basics

Education

Ph.D.

University of Notre Dame

Computer Science and Engineering

Bachelor of Engineering

Hong Kong University of Science and Technology (HKUST)

Computer Science

Publications

Better datasets start from refinelab: Automatic optimization for high-quality dataset refinement

AAAI Conference on Artificial Intelligence

Luo*, Xiaonan and Huang*, Yue and He, Ping and Zhang, Xiangliang. High-quality datasets are crucial for training effective machine learning models. This paper presents RefineLab, an automated framework for optimizing dataset refinement processes to improve data quality and model performance.

ChemOrch: Empowering LLMs with Chemical Intelligence via Synthetic Instructions

Advances in Neural Information Processing Systems (NeurIPS)

AdaReasoner: Adaptive Reasoning Enables More Flexible Thinking

Advances in Neural Information Processing Systems (NeurIPS) - Spotlight

Wang, Xiangqi and Huang, Yue and Wang, Yanbo and Luo, Xiaonan and Guo, Kehan and Zhou, Yujun and Zhang, Xiangliang. AdaReasoner introduces an adaptive reasoning framework that enables large language models to dynamically adjust their reasoning strategies based on problem complexity and context.

Torpor: GPU-Enabled Serverless Computing for Low-Latency, Resource-Efficient Inference

USENIX Annual Technical Conference (ATC)

Yu, Mingyu and Wang, Ao and Chen, Dong and Yu, Haoxuan and Luo, Xiaonan and Li, Zhuzhong and others. Torpor presents a GPU-enabled serverless computing system that achieves low-latency, resource-efficient inference by intelligently managing GPU memory and computation resources.

Skills

Languages

Interests

Luo, Xiaonan and Huang, Yue and He, Ping and Zhang, Xiangliang. High-quality datasets are crucial for training effective machine learning models. This paper presents RefineLab, an automated framework for optimizing dataset refinement processes to improve data quality and model performance.