Fengyi Zhang

Nov 25, 2024 | opg.optica.org | Yang Liu |Fangmei Liu |Shaolin Ke |Fengyi Zhang

We propose a method to convert fundamental modes into orbital angular momentum (OAM) modes through chiral dynamics induced by gauge fluxes in silicon waveguides. By integrating a trench into a few-mode waveguide, we induce the rotation of TE10 and TE01 modes, naturally generating the gauge flux for the synthesized OAM modes.

Oct 15, 2024 | pubs.rsc.org | Xiaohui Ma |Xi Zhang |Fengyi Zhang |Qiong Wang

High Zn(002)-preferential orientation enabled by proton additive for dendrite-free zinc anode Although zinc-based batteries have long been considered as one of the most promising technologies for large-scale energy storage, its development was still seriously hindered by dendrite formation.

Sep 10, 2024 | pubs.rsc.org | Xiaoyu Zhao |Fengyi Zhang |Yaxin Wang |Yongjun Zhang

Dilute Ferromagnetic ZrO2/Carbon Nanocomposite Derived from Zirconium-Based Metal-Organic Framework for High-Performance Electromagnetic Wave Absorption Metal-organic frameworks (MOF) have been intensively studied for electromagnetic wave absorption (EMA) due to their special structure and rich carbon and metallic sources.

Oct 30, 2023 | thinkchina.sg | Andrew Huang |Fengyi Zhang

Technology China put forth a draft Degree Law recently that includes harsh consequences for degree holders who use artificial intelligence tools to ghostwrite their dissertations. These aggressive measures reflect a conundrum that the country’s academia and wider community finds themselves in: how can China balance between the desire for technological progress and the fear of losing its identity and autonomy?

Jul 4, 2023 | onlinelibrary.wiley.com | Yiming Qian |Fengyi Zhang |zhiyong Liu

Abstract Lifelong reinforcement learning is able to continually accumulate shared knowledge by estimating the inter-task relationships based on training data for the learned tasks in order to accelerate learning for new tasks by knowledge reuse. The existing methods employ a linear model to represent the inter-task relationships by incorporating task features in order to accomplish a new task without any learning.