Miles S

Intuitive video explaining how linear regression can be viewed in a probablisitic framework, including popular regularization choices
www.youtube.com/watch?v=q7se...

A depressing thing I’ve found is that the left is just in denial about this. I’ve met quite a few people where if I mention Fox News they say “I’ve never watched it”. How will you understand your own country if you don’t from time to time?

Somehow the speed up is both quadratic and exponential

Have researchers really ‘de-extincted’ the dire wolf? No, but behind the hype was a genuine breakthrough | Helen Pilcher The pups are cute – and great for PR – but they’re modified grey wolves. The real work is being done with their red cousins, says science writer Helen Pilcher

This is not quantum-related but the parallels with quantum computing are uncanny: a company has declared that it has resurrected an extinct species from the last ice age, but all they actually did was change 14 genes of the grey wolf's DNA. 🧵
www.theguardian.com/commentisfre...

What if Wigner's friend was Schrodinger's cat?

These Mathematical Equations Are Slashing America’s Electric Bills These Mathematical Equations Are Slashing America’s Electric Bills on Simons Foundation

A beautiful story of abstract mathematics extending into physics and engineering and resulting in 33% more efficient lighting. Better algorithms were the key, as they usually are, speeding up calculations by 1000x. www.simonsfoundation.org/2025/03/24/t...

The attitude we desperately need in Tech

I think there could be other reasons to build quantum computers, if more serious investigations were done into them. Power consumption and time to solution mainly. I think supremacy over classical is mostly a dead end, especially for NISQ.

I should add: not just simulate, but to better accuracy than they have reported. And in a scalable way.

Just to be clear, I don’t think we claimed anywhere that D-Wave has no advantage. We see that as something they still have the burden of proving (if it’s even conceivably provable). We just showed it’s possible to simulate a large range of the same protocols.

Actually we just haven’t tried that one yet … maybe it’s hard, maybe not. It’s an open question.

Maybe the question should be why D-Wave’s was timed to land just before APS :^)

Yeah I can ask my coauthors again but I certainly had no personal idea about the timing. It is pretty wild.

Could perhaps have done it sooner but one of the key methods we used was only invented by a different group last fall. I’ve been emphasizing to journalists that it’s a fast-moving, dynamic field!

In fact, we posted when we did because of APS Physics Summit coming up and the first author wanted to speak on it. The timing with Science’s publication was a coincidence. It’s been surprising to us too.

👇 quite worthwhile to read the whole thread

Challenging the Quantum Advantage Frontier with Large-Scale Classical Simulations of Annealing Dynamics Recent demonstrations of D-Wave's annealing-based quantum simulators have established new benchmarks for quantum computational advantage [arXiv:2403.00910]. However, the precise location of the classi...

Another new classical simulation incoming:
arxiv.org/abs/2503.08247

How far we can scale up to very large sizes depends quite a bit on which system 2D vs 3D interactions etc. All of our results appear quite scalable though.

It depends. Since the couplings are random one has to define that limit carefully, and if it’s defined through observables then one can often obtain the same result by averaging many finite systems.

And if similar results can be obtained for other systems, the flexibility and precision of tensor network methods can shed light on important open physics questions.

Despite decades of research, the development of classical algorithms continues at a rapid pace and is really only beginning. More caution is needed when claiming that structured problems, including quantum ones, are too hard for classical methods.

It is sometimes claimed that simulating dynamics is exponentially hard, but not all problems are created equally. Newly efficient methods for contracting tensor networks are bringing more cases of dynamics of 2D & 3D quantum systems within reach.

Quantum dynamics is a frontier at which classical simulation methods have struggled: For example, many Monte Carlo methods encounter a severe sign problem and tensor networks have been limited to short times and low dimensions.

Dynamics of disordered quantum systems with two- and three-dimensional tensor networks Quantum spin glasses form a good testbed for studying the performance of various quantum annealing and optimization algorithms. In this work we show how two- and three-dimensional tensor networks can ...

In a new preprint arxiv.org/abs/2503.05693, led by Joseph Tindall and Antonio Mello at Flatiron CCQ, we simulate annealing of disordered quantum magnets 🧲 ⌛ and in many cases find better accuracy than recent results from D-Wave devices and leading classical methods (c.f. arxiv.org/abs/2403.00910).