Junyu Liu

Assistant professor @ Pitt, Postdoc @ UChicago/IBM, PhD @ Caltech, working on quantum, AI, crypto.

[John Preskill @preskill]

@TeamOratomic has raised $300M to build the world’s first fault-tolerant quantum computer.

Now we’re growing the team to make it happen. We’re looking for extraordinary people to join us in pursuing one of the defining technological achievements of our generation.

Details: https://t.co/UnTVNoYpsc

[Junyu Liu @junyuliu4]

@Isaac__kim That could potentially make FTQC easier. I am curious that: 1. How about sensitivity to noise? How sensitive the catalyst state has to noise, and how to correct them efficiently? 2. Can this be used to implement QRAM as well, which contains controlled i-swap, which contains T?

[John Preskill @preskill]

I'm excited and honored to receive the 2024 Bell Prize.
https://t.co/dApr6RkMq6

[Junyu Liu @junyuliu4]

@chae_yeun_park @Qottmann A few people are not ignoring, like me. https://t.co/5e95EINr1E we show that classical machine learning has barren plateau as well.

[Jens Eisert @jenseisert]

I am happy to see this work out in the @NatureComms. Warm thanks to the team of @junyuliu4, Minzhao Liu, Jin-Peng Liu, Ziyu Ye, Yuri Alexeev, and Liang Jiang for this great @UofC-@FUB-@FraunhoferHHI-@ChicagoQuantum-@qbraid_official-@argonne-@UCBerkeley-@CQE_MIT-collaboration.

[Junyu Liu @junyuliu4]

@quantum_graeme Actually, I don’t know if classical error correction codes have been demonstrated to a large scale in personal computing systems: they are not needed classically in most cases.

[Junyu Liu @junyuliu4]

@qZoeHolmes @MvsCerezo I believe that our papers have similar spirits, where I introduce a new notion, laziness, and distinguishing it from barren plateaus.

[Junyu Liu @junyuliu4]

@qZoeHolmes @MvsCerezo The paper is very nice. wish to mention that I have a paper with the related idea:
[2206.09313]
In this paper we say that classical MLPs have the barren plateau as well, where 1/N in quantum is replaced by 1/width in classical neural networks.

[Pedram Roushan @PedramRoushan]

Home to great works on quantum sensing, simulation, and communication.

[Junyu Liu @junyuliu4]

@kanav_setia @quantum_graeme Henry, Junyu, Kanav

[Junyu Liu @junyuliu4]

@Isaac__kim One can start from quantum diffusion model and perform their algorithm to train the generative AI for graphics.

[Junyu Liu @junyuliu4]

@CharzKelso Thanks!

[Junyu Liu @junyuliu4]

@quantum_graeme What’s the salary?

[Dan Roberts @danintheory]

Congrats to @em_dinan, @ShoYaida, and @suchenzang on their new work https://t.co/ZT61Hg7Awl, applying the effective theory blueprint of PDLT to the transformer architecture!

[Boris Hanin @BorisHanin]

I’m excited to read this!

[SeQure @TheSeQure]

🎉Congratulations to @junyuliu4, co-founder of @TheSeQure, for being selected as one of the "National Chinese Youth Under 30 Elite List," one of the most prestigious awards in the space.🎉 #Top30under30 #SeQure #Entrepreneurship #AI #Metaverse #NFTGiveaways

[Jin-Peng Liu @JinPengLiu__Sky]

We develop the first quantum algorithm for training classical sparse neural network with end-to-end setting. It shows that fault-tolerant quantum computing can contribute to the scalability and sustainability of the most state-of-the-art, large-scale machine learning models.

[Jens Eisert @jenseisert]

In fact, @junyuliu4 is at the #APSMarch at this moment in time. That picture was taken last week when he was spending time in Berlin.

[Jens Eisert @jenseisert]

The biggest advantage #quantumcomputers can contribute to algorithms for #machinelearning may be in the #training of large-scale #classical models.

https://t.co/MuhRvgF2OJ

[Junyu Liu @junyuliu4]

@JiaqingJ At least Trotter simulation and adiabatic algorithms could work.