Hello! I’m Xudong Shen (沈旭东). Friends also call me Oliver. I’m finishing my Ph.D. at the National University of Singapore, advised by Professor Mohan Kankanhalli. My research has focused on the technical social alignment of AI, particularly in fairness, and AI governance. I actively contribute to open-community projects, such as inverse scaling, BIG-bench, and natural language instructions. I also organize the algorithm, law, and policy working group at EAAMO bridges (formerly MD4SG).
My long-term goal is to address catastrophic AI risks, specifically the social risks associated with AGI and superhuman AI. These risks are theorized to arise with more advanced AI that surpasses certain thresholds of capability, yet it is striking how some have already shown signs of reality in today’s AI systems, such as sycophancy, situational awareness, and reward hacking. Current safety and alignment methodologies are not well-equipped to address these risks for future superhuman AI. I am actively seeking opportunities (1) to do grounded work on catastrophic AI risks on actual AI systems and (2) to develop technical methods that are more likely to be effective for superhuman AI. Please do get in touch if you share the same interest or know of any opportunities. I truly appreciate it.
From 2022 to 2023, I interned at Sea AI Lab in Singapore, working with Chao Du and Tianyu Pang. From 2015 to 2019, I was an undergraduate at Zhejiang University, China, majoring in Naval Architect and Ocean Engineering. During this period, I explored topics such as computational fluid dynamics, satellite imaging, and policy research. My cv is here.
If you are interested, I’ve written something more about myself…
Publications on fairness:
Publications on AI safety and benchmarking:
Ian R. McKenzie, Alexander Lyzhov, Michael Martin Pieler, Alicia Parrish, Aaron Mueller, Ameya Prabhu, Euan McLean, Xudong Shen, Joe Cavanagh, Andrew George Gritsevskiy, Derik Kauffman, Aaron T. Kirtland, Zhengping Zhou, Yuhui Zhang, Sicong Huang, Daniel Wurgaft, Max Weiss, Alexis Ross, Gabriel Recchia, Alisa Liu, Jiacheng Liu, Tom Tseng, Tomasz Korbak, Najoung Kim, Samuel R. Bowman, Ethan Perez, “Inverse Scaling: When Bigger Isn’t Better”, TMLR, 2023. [openReview] [arXiv] [GitHub]
Aarohi Srivastava, …, Xudong Shen, … (450 authors), “Beyond the Imitation Game: Quantifying and Extrapolating the Capabilities of Language Models”, JMLR, 2023. [openReview] [arXiv] [Github]
Kaustubh Dhole, …, Xudong Shen, … (125 authors), “NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentation”, Northern European Journal of Language Technology (NEJLT), 2023. [paper] [arXiv] [GitHub]
Yizhong Wang, …, Xudong Shen, … (40 authors), “Benchmarking Generalization via In-Context Instructions on 1,600+ Language Tasks”, EMNLP, 2022. [paper] [arXiv] [data]
Publications on AI governance:
Other publications: