Fengyuan "Michael" Liu
👋
Postdoctoral Research Associate @ Princeton University · Princeton, USA
Howdy! I am an incoming Presidential Postdoctoral Research Fellow affiliated with the University Center for Human Values at Princeton University. Advised by Prof. Molly Crockett, I am interested in how recent advances in artificial intelligence shape diversity in science. Before joining Princeton, I spent 9 years in Abu Dhabi, where I obtained a Ph.D. in computer science from New York University, supervised by Prof. Talal Rahwan and Prof. Bedoor AlShebli, after obtaining a B.S. in computer science from New York University Abu Dhabi. I was born in Xiangyang and grew up in Wuhan, China.
At its broadest, my research interrogates how scientific knowledge is gatekept, and how to redesign those gates. My dissertation examined gatekeeping in its most visible form (i.e., editors of academic journals). Since then, my work has expanded to include other actors who gatekeep scientific knowledge, including digital technologies (via auditing online platforms and AI tools), science policies (via estimating their causal effect), and ourselves (via studying interaction among scientists). By studying how social institutions and digital technologies shape equitable access to scholarly knowledge, I ask: who controls access? Who is left out? What governance choices expand participation?
With a research agenda deeply rooted in metascience and computational social science, I do not pledge allegiance to any single method. Instead, I triangulate evidence across qualitative, computational, and econometric frameworks. The strength of a multidisciplinary approach lies in producing not just stylized facts, but robust evidence that advances normative debates and informs policy.
I have published in Nature Human Behaviour and PNAS. My work was used by the OECD report on the state of academic careers, and featured in The Times, The Guardian, Scientific American, Times Higher Education, Nature, and Science, among many others.
🔎Research interests
📄Publications
-
Citation manipulation through citation mills and pre-print serversPaper ·
-
Non-White scientists appear on fewer editorial boards, spend more time under review, and receive fewer citations
-
Gender inequality and self-publication are common among academic editors
-
Perception, performance, and detectability of conversational artificial intelligence across 32 university courses
📝Works in progress
-
Current policies governing editorial conflicts of interest are ineffective
-
Causal evidence of racial and institutional biases in accessing paywalled articles and scientific data
🎓Teaching & 🛠️ Service
Teaching
- Teaching Assistant, CS-UH 4001 & CS-UH 4002 - Capstone Project in Computer Science - New York University Abu Dhabi (2022-2024)
- Section Leader, CSCI-GA 1170 Fundamental Algorithms - New York University (Fall, 2020 and Spring, 2021)
- Guest lecture in CS-UH 2219E Computational Social Science - New York University Abu Dhabi (Spring 2023)
Service
- Reviewer: International Conference for Computational Social Science (IC2S2, 2024 & 2025), Nature Human Behaviour (2023)
- Sub-reviewer: Nature (2025), Nature Communications (2025), International Conference on the Science of Science and Innovation (ICSSI, 2023)
- Committees: Graduate Student Council, Inclusion, Equity and Action Committee, and Landscape Master Plan Visioning Committee at NYU Abu Dhabi