邀请人：扈鸿业

报告摘要：
A central challenge in quantum computing is to identify tasks where quantum devices can outperform classical computers while keeping that advantage both verifiable and experimentally realistic. I will discuss peaked circuit sampling as one route toward this goal. Unlike standard sampling proposals, peaked circuits contain a high-probability output string that can serve as a directly checkable witness, while still admitting strong evidence for classical hardness. I will present recent results on the complexity of generating peaked distributions, including hardness results and circuit-complexity lower bounds,as well as hardware-oriented constructions and experiments on Quantinuum and IBM devices. I will also discuss the emerging classical attacks on these constructions and how circuit recompilation can hide the local history of how a peak was planted without changing the computation.

报告人简介：
Yuxuan Zhang is an incoming Assistant Professor in Computer Science and Physics at the National University of Singapore. He is currently a postdoctoral researcher at EPFL and Princeton University with Dmitry Abanin, and received his Ph.D. in Physics from UT Austin under Scott Aaronson and Andrew Potter. His research spans quantum advantage, many-body dynamics, and machine learning for quantum computing.