Abstract: 

Quantum information technology is advancing at an astonishing pace. What should we do with it?

In this pedagogical talk, I will argue that one of the most powerful and near-term applications of quantum computers lies in quantum sensing. We will show how a digital quantum computer can be used to improve a foundational sensing task: the detection of weak oscillating fields with unknown amplitude and frequency. This task underlies a wide range of real-world applications, from NMR and atomic clocks to gravitational-wave detection and dark-matter searches. By integrating quantum computation with sensing, we develop quantum computing–enhanced sensing protocols that achieve the ultimate performance bound—the Grover–Heisenberg limit. The central idea is to digitize a continuous, analog signal into a discrete quantum operation that can be coherently processed within a quantum algorithm, enabling nonlinear signal processing beyond what is possible with conventional strategies. We will discuss what makes this approach qualitatively distinct from traditional entanglement-enhanced quantum sensing, and propose proof-of-principle implementations based on solid-state spins. Together, these results establish quantum computation as a new and practical resource for pushing the fundamental limits of sensing.

Bio:

Soonwon Choi was born in Seoul and grew up in Daejeon, South Korea. He completed his undergraduate study in physics at Caltech in 2012, got his PhD degree in physics from Harvard University under the supervision of Prof. Mikhail Lukin in 2018. Then he worked as a Miller Postdoctoral Fellow at UC Berkeley before joining MIT as an assistant professor in July 2021. 

Abstract: 
A popular selling point of diffusion large language models (dLLMs) is their capacity for parallelism: the ability to generate sequences of text far more efficiently than autoregressive models, which require one forward pass per token. Yet among the many competing paradigms for dLLMs, from masked to uniform to Gaussian diffusion, principled understanding of how these different proposals compare in parallelism remains limited. In this talk, I will show that (1) all three can sample in a number of forward passes that scales with the dual total correlation of the underlying distribution, yet (2) there are natural families of distributions for which uniform and Gaussian diffusion quadratically outperform masked diffusion. Contrary to popular intuition that masked diffusions are harder to parallelize because they must commit to token values, the latter separation instead comes from the fact that the “critical windows” / “speciation transitions” in masked diffusion sampling are quadratically sharper than those in uniform and Gaussian diffusion sampling.

Bio:
Sitan Chen is an Assistant Professor of Computer Science at Harvard SEAS, where he is a member of the Theory of Computation and the Harvard Quantum Initiative, and an associate faculty at the Kempner Institute. Previously, he was an NSF math postdoc at UC Berkeley, after completing his PhD in EECS at MIT in 2021. His work has been recognized with an NSF CAREER award, an ICML Outstanding Paper Award, and the Harvard Dean’s Competitive Fund for Promising Scholarship.

TBD

TBD