Wanqian Yang

Papers

Total Citations

133

H-Index

About

Wanqian Yang is a researcher at the intersection of cognitive science, artificial intelligence, and Bayesian machine learning. Her work centers on understanding how humans and machines can learn and plan efficiently, with a particular focus on hierarchical representations and structured priors. Yang’s most influential contribution is her 2020 paper, “Discovery of hierarchical representations for efficient planning” (107 citations), which proposes that humans spontaneously organize environments into clusters of states to support hierarchical planning—a mechanism that allows complex problems to be broken down into manageable sub-problems. This work, building on her earlier 2018 study (16 citations), offers a computational account of how people tackle challenging tasks by leveraging abstraction. In parallel, Yang has advanced Bayesian deep learning through her 2019 paper, “Output-Constrained Bayesian Neural Networks” (10 citations), which introduces a novel prior that encodes functional constraints directly in function space, enabling more principled uncertainty quantification. Her research bridges cognitive modeling and practical AI, demonstrating how insights from human cognition can inspire more efficient, interpretable algorithms. With a growing citation record and a focus on foundational problems in planning and representation learning, Yang is a rising voice in the quest for more human-like artificial intelligence.