M.S. Student
Machine Learning Department
Abu Dhabi, UAE
Hi, I am Dylan Xinming Hou, an M.S. student in Machine Learning at Mohamed bin Zayed University of Artificial Intelligence, where I work with Gus Xia at Music X Lab. I also work with Hua Shen at BiAlign Lab, NYU Shanghai. Previously, I received my B.E. in Artificial Intelligence from the Gaoling School of Artificial Intelligence, Renmin University of China, where I worked with Wayne Xin Zhao on NLP-related research.
My previous research has explored a range of machine learning topics, including LLMs, digital agents, interpretability, AI safety, mind-inspired (world) models, and human-AI interaction. These projects gave me a technical foundation spanning AI capabilities, behavior, and interaction, while gradually leading me toward broader, more interdisciplinary questions about how humans and AI shape one another.
My current work across Music X Lab and BiAlign Lab builds on this foundation, focusing particularly on designing AI systems inspired by the human mind and responsive to human intentions, while remaining aligned with diverse and evolving human values and supporting human creativity, cognition, and agency. Together, these efforts shape my broader research vision of human-AI co-evolution.
My research explores how AI and humanity can co-evolve, with the goal of ensuring that increasingly capable AI systems remain aligned with diverse human values while helping to expand individual potential and foster collective flourishing.
As AI systems become more capable and deeply embedded in human society, it is essential that their development remain aligned with the diverse and evolving values of individuals, communities, and cultures.
I am interested in how AI systems can learn, represent, and respond to human values, intentions, and preferences, while remaining trustworthy, safe, and beneficial as their capabilities continue to advance.
Beyond automating existing tasks, I am interested in how AI can become a human-centered partner that strengthens human creativity, cognition, and agency.
I seek to understand how people and AI systems can learn, create, and collaborate in ways that preserve what makes us human by empowering individuals and communities to explore new possibilities, pursue their aspirations, and realize goals that are meaningful to them.
Dylan Xinming Hou, Juntian Zhang, Xu Gu, Yichen Wu, Nils Lukas, Gus Xia, Xiuying Chen, Yuhan Liu
A survey of AI-generated video detection from both visual and language perspectives, with a curated taxonomy and resource list.
Dylan Xinming Hou
This work studies evidence-calibrated synthetic users for predicting human outcomes in interaction design.
Dylan Xinming Hou, Xin Shu, Zhuojiang Cai, Yuting Wu, Yueqing Hu, Enkelejda Kasneci, Haiyang Geng, Gus Xia
This work studies clinically grounded affective interaction in a music-based therapeutic system.
Jizhi Zhang, Dylan Xinming Hou, Zhaoyi Li, Muhammad Asif Ali, Gus Xia, Lijie Hu
A benchmark for studying rule-level knowledge editing behavior in large language models.
Zhuoqi He, Zhaoyi Li, Dylan Xinming Hou, Daniil Ognev, Gus Xia, Lijie Hu
An analysis of whether edited knowledge systematically generalizes in multi-model large language models.
Xinming Hou, Mingming Yang, Wenxiang Jiao, Xing Wang, Zhaopeng Tu, Wayne Xin Zhao
This work studies autonomous agent collaboration through a global-local hierarchy that combines centralized
coordination with local specialization.
BiAlign Lab at NYU Shanghai, working with Hua Shen
Working on bidirectional human-AI alignment, value alignment, and human-AI co-evolution.
Music X Lab, advised by Gus Xia
Working on human-AI affective interaction and mind-inspired models.
RUC AI BOX, advised by Wayne Xin Zhao
Worked on machine learning, natural language processing, factuality control, prompt pipelines for reasoning tasks, and interpretability.
Tencent AI Lab, advised by Wenxiang Jiao and Xing Wang
Explored human-inspired organization and collaboration mechanisms for multi-agent systems and LLM-based agents.
Mohamed bin Zayed University of Artificial Intelligence, advised by Gus Xia at Music X Lab
Current graduate study focused on machine learning, human-centered AI, affective interaction, and mind-inspired (world) models.
Gaoling School of Artificial Intelligence, Renmin University of China
B.E. in Artificial Intelligence with honors.
Shandong Experimental High School
High school education in Jinan, Shandong Province, China.