About

I am an AI safety researcher at Lila Sciences, helping to develop scientific superintelligence responsibly and safely! Contact me at "wen dot projectz at gmail dot com" to discuss research.

I was previously a researcher at the Center for AI Safety, where my work focused on benchmarking and developing robust, aligned, and transparent artificial intelligence systems. Before that, I was a Senior Research Scientist at the Institute for Infocomm Research (I²R) A*STAR in Singapore, specializing in multimodal large language models (LLMs). My research focused on developing localized audio-text LLMs and benchmarks, as part of the National Multimodal LLM Programme.

My work also spans vision-language models, computer vision (particularly robustness and transfer learning), and time series prediction and anomaly detection. My broader research interest lies in advancing generalizable and trustworthy AI models for real-world applications.

I received my PhD in Statistics from Cornell University in 2020, advised by David Matteson, where I focused on anomaly and change detection. Prior to that, I earned my B.A. in Applied Mathematics and Statistics from UC Berkeley. My undergraduate and PhD studies were supported by the A*STAR National Science Scholarship (BS-PhD). I have also gained valuable experience through research and industry internships at Amazon, MERL (supervised by Devesh Jha) and IBM Science of Social Good Fellowship (supervised by Kush Varshney, Lingfei Wu, Karthikeyan Ramamurthy, Jinfeng Yi, Raya Horesh) during my PhD.

For a full list of my publications, please refer to my Google Scholar profile.