#Role of machine learning Particle Characterization

[qrovidcore]

hey what’s up tumblr i’ve now seen hbo’s watchmen all the way through Three Fucking Times and i very well may go for a fourth if given an excuse whoops and apparently i can’t stop thinking about Laurie’s joke in She Was Killed By Space Junk, no i’m not the first person to analyze this and i’m sure i won’t be the last but i sure do have some Thoughts^TM,  so here’s some meta let’s go.

major spoilers ahead for the entire series:

Hey, it’s me again. I’ve got a joke. Stop me if you’ve heard this one. There’s this guy, he’s a bricklayer. He’s really good at it. He’s a real master of his craft. Because he’s precise. Every brick has its place. Anyway this guy has a daughter and he’s gonna teach her to be a bricklayer because after all, all a man has is his legacy. So dad decides to build a barbecue in the backyard. He does the math. He figures out exactly what he needs and he shows the daughter how to do everything. Step by step. And when he finishes, it’s a beauty. It’s a perfect barbecue. Just the way he drew it in blueprints. Only one problem. There’s a brick left over. One single brick. The guy freaks out. He must have done something wrong. He’s gonna have to start all over again. So he picks up his sledgehammer to knock the thing to pieces and his daughter suddenly says ‘daddy wait! I have an idea.’ She picks up the orphan brick and throws it up into the air as high as she can. And then…shit. Messed it up.

Okay forget that joke. Can I tell you another one?

As I said, I’m not the first to break down that Laurie is referring to specific people who have an influence on the story, there’s plenty of meta posts online that’ll say the same thing. I just think this is a Really Clever way to introduce us to her, to the major players in this story, and to the events from the comic that are going to end up being referenced. Anyhow, the bricklayer here is The Comedian. Laurie’s father. I’ll get back to this and how it connects later, but given that one of Watchmen’s major themes is the concept of legacy - who carries it and how, and what happens when that legacy is painful - this is a neat little hook into that idea. Laurie’s dad’s legacy. What she’s done with it, what she’s going to do with it, how she feels about it. Again, coming back to that.

Okay. Forget the brick. New joke. Three heroes die and they all show up at the pearly gates. God’s there and he’s going to decide what their eternal fate shall be: heaven or hell. Our first hero is dressed up like a big owl. God says to him “I gifted you the ability to make fantastic inventions. What did you do with this amazing talent?” Owl guy says “I made this really awesome flying ship and lots of cool outfits and weapons so I could bring peace to the city.” God asks, “So how many people did you kill?” Owl guy seems offended. He says “Zero. I didn’t take a single life.” God frowns. “Sorry owl guy, your heart’s in the right place but you’re just too soft.” God snaps his fingers and the hero goes to hell.

I'm not super into the comic so it took me a while to get that she's referencing Nite Owl. I think this is strange since he doesn't appear in the show himself, whereas everyone else she talks about does, but I suppose it gives a more rounded-out view of the different approaches to heroism, and what exactly constitutes it, and also ties in another one of the original Minutemen. They did cut this over her arrest of Mr. Shadow in the bank, which makes me wonder about his role and why he appeared, and I still find it strange that this part of the joke wasn't about someone who had more of a presence in the show. (Though that being said, DC making fun of Batman, their own big-ticket character? 10/10 thank you for this).

Where was I? The pearly gates await our next hero in line for Almighty judgment. Our hero number two is confident he can game this out because that’s his God-given talent: smarts. Some might even say he’s the smartest man in the world. “So what did you do with that big brain I gave you?” asks God. “As a matter of fact, I saved humanity, ”says Smarty Pants. “Well how’d you do that,” asks God.” “Well I dropped a giant alien squid on New York and everybody was so afraid of it they stopped being afraid of each other.” “OK,” says God. “How many people did you kill?” Smarty Pants smiles. “Three million, give or take. But you can’t make an omelet without breaking a couple of eggs. “Christ,” God says. “You’re a fucking monster.”  “Am not,” says Smarty Pants. God snaps his fingers and our hero goes to hell.

GOD YES PLEASE DRAG OZYMANDIAS. GET THIS FUCKER’S ASS. Though the line that’s sticking out to me here is “You can’t make an omelet without breaking a couple of eggs.” Watchmen’s got an egg motif - and that’s an entire post on its own - and wow this is a place to drop it. I find it interesting that it’s given to Adrien here. Especially since it comes back later, when Will tells Angela that that’s what Jon said in justification of giving his life to stop the 7th K/Cyclops and Trieu. Eggs are used for a lot of things, but this line ties the motif solidly to a value of life here - how Adrien is the way he is because he refuses to value other peoples’, and maybe how Jon is the way he is because, when you can see the future laid out before you and live knowing how you’re going to die, how do you learn to value your own?

Okay. We’re down to the nitty gritty now. One hero left. God cracks his knuckles ready to administer the final reckoning. Now Hero Number 3 is pretty much a god himself. So for the sake of telling them apart, he’s blue and he likes to stroll around with his dick hanging out. He can teleport, he can see into the future, he blows shit up. He’s got actual superpowers. Regular God asks Blue God what have you done with these gifts?” Blue God says “I fell in love with a woman, I walked across the sun, and then I fell in love with another woman. I won the Vietnam War. But mostly I just stopped giving a shit about humanity.” God sighs. “Do I even need to ask how many people you’ve killed?” Blue guy shrugs. “A live body and a dead body have the same number of particles so it doesn’t matter. And it doesn’t matter how I answer your question because I know you’re sending me to hell.” “How do you know that?” asks God. Blue God sounds very sad when he softly says “Because I’m already there.” And so, a mere piston in the inevitable of time and space God does what he did and will do. He snaps his fingers and the hero goes to hell.

And now, we’ve got Jon. Dr. Manhattan. It's a neat moment of insight into his actions, motives, and how those are perceived by others (namely Laurie), and it's a nice thread of introduction to his previous actions to drop for audiences who haven't read the comics (actually, I can make this point about Adrien’s part of the joke too). Especially because most of what we get of Jon in-show is his relationship with Angela, his entire character arc really revolves around her and we don't see him portrayed as the contentious, unfeeling figure the world sees him as. So this sort of contrast between him as a figure and him as a person is very telling, doubly so coming from someone who it's clear knew him. And I really appreciate that there’s just as much stiffness as there is warmth to the Jon we the audience see - he’s kind, he’s loving, but he’s also very matter-of-fact and deterministic, and that bit of characterization really spans the gap between these two versions of him.

And so it’s been a long day at the pearly gates. All the heroes have gone to hell. His work done, God’s packing up to go home and then he notices someone waiting. But it’s not a hero, it’s just a woman. “Where did you come from?” asks God. “Oh I was just standing behind those other guys the whole time, you just didn’t see me.” “Did I give you a talent,” God asks. “No, none to speak of,” says the woman.  God gives her a good long look. “I’m so sorry. I’m embarrassed. Seriously, this almost never happens but I don’t know who you are.” And the woman looks at God and she quietly says “I’m the little girl who threw the brick in the air.” And a sound from above, something falling: the brick. God looks up but it’s too late. He never saw it coming. It hits him so hard, his brains shoot out his nose. Game over. He’s dead. And where does God go when he dies? He goes to hell. 

Into some Thoughts^TM that I haven’t seen anyone theorize yet(?): I think God is meant to be Lady Trieu, and even if Laurie wouldn’t know this yet that’s some brilliant fucking foreshadowing. It's not as exact, but enough parallels are there that I think they're purposeful. It makes Trieu out as the ultimate judge of everyone - and in a way, she is. She sees herself as the most deserving of power of everyone, and it's her who kills Dr. Manhattan - sends him to hell, you could say, and he knows she's going to do it. It also hints at how she's going to die too, crushed by her machine falling from the sky like the brick, because she didn't expect anyone would be capable of stopping her. And where does God go when he dies? He goes to hell. Trieu isn't ultimately above the others, and she's subject to their justice as they are to hers. 

Fitting too that Laurie is involved with the plan to stop Trieu, since, as I said I’d come back to, the girl who threw the brick is Laurie herself. Her depiction of herself in this way is representative, perhaps, of Laure's own feelings on vigilantism and what justice is, and that she's the force that's going to bring down these overblown personalities and their many incorrect uses of their abilities. Given this, it's interesting to think how the "failed" joke at the beginning connects, given that Laurie's dad is the bricklayer, and he's definitely... not a good person, or at least not in this continuity. But I wonder if it's indicative of what Laurie mentions about her parents training her up to do vigilante stuff (especially since she’s based in part(?) on a member of the Minutemen from the comic), and how she feels about her father and his work. If the brick is symbolic of his work as a vigilante, is Laurie throwing the brick in the air, and ultimately taking down the threat at the top, meant to indicate how she sees herself using what she learned from him, or - maybe and - a disrespect for his work based on her justified hatred of him?

Roll on snare drum. Curtains. Good joke. 

[dorcasrempel]

A scientist turns to entrepreneurship

Like the atomic particles he studies, Pablo Ducru seems constantly on the move, vibrating with energy. But if he sometimes appears to be headed in an unexpected direction, Ducru, a doctoral candidate in nuclear science and computational engineering, knows exactly where he is going: “My goal is to address climate change as an innovator and creator, whether by pushing the boundaries of science” through research, says Ducru, or pursuing a zero-carbon future as an entrepreneur.

It can be hard catching up with Ducru. In January, he returned to Cambridge, Massachusetts, from Beijing, where he was spending a year earning a master’s degree in global affairs as a Schwarzman Scholar at Tsinghua University. He flew out just days before a travel crackdown in response to Covid-19.

“This year has been intense, juggling my PhD work and the master’s overseas,” he says. “But I needed to do it, to get a 360-degree understanding of the problem of climate change, which isn’t just a technological problem, but also one involving economics, trade, policy, and finance.”

Schwarzman Scholars, an international cohort selected on the basis of academic excellence and leadership potential, among other criteria, focus on critical challenges of the 21st century. While all the students must learn the basics of international relations and China’s role in the world economy, they can tailor their studies according to their interests.

Ducru is incorporating nuclear science into his master’s program. “It is at the core of many of the world’s key problems, from climate change to arms controls, and it also impacts artificial intelligence by advancing high-performance computing,” he says.

A Franco-Mexican raised in Paris, Ducru arrived at nuclear science by way of France’s selective academic system. He excelled in math, history, and English during his high school years. “I realized technology is what drives history,” he says. “I thought that if I wanted to make history, I needed to make technology.” He graduated from Ecole Polytechnique specializing in physics and applied mathematics, and with a major in energies of the 21st century.

Creating computational shortcuts

Today, as a member of MIT’s Computational Reactor Physics Group (CRPG), Ducru is deploying his expertise in singular ways to help solve some of the toughest problems in nuclear science.

Nuclear engineers, hoping to optimize efficiency and safety in current and next-generation reactor designs, are on a quest for high-fidelity nuclear simulations. At such fine-grained levels of modeling, the behavior of subatomic particles is sensitive to minute uncertainties in temperature change, or differences in reactor core geometry, for instance. To quantify such uncertainties, researchers currently need countless costly hours of supercomputer time to simulate the behaviors of billions of neurons under varying conditions, estimating and then averaging outcomes.

“But with some problems, more computing won’t make a difference,” notes Ducru. “We have to help computers do the work in smarter ways.” To accomplish this task, he has developed new formulations for characterizing basic nuclear physics that make it much easier for a computer to solve problems: “I dig into the fundamental properties of physics to give nuclear engineers new mathematical algorithms that outperform thousands of times over the old ways of computing.”

With his novel statistical methods and algorithms, developed with CRPG colleagues and during summer stints at Los Alamos and Oak Ridge National Laboratories, Ducru offers “new ways of looking at problems that allow us to infer trends from uncertain inputs, such as physics, geometries, or temperatures,” he says.  

These innovative tools accommodate other kinds of problems that involve computing average behaviors from billions of individual occurrences, such as bubbles forming in a turbulent flow of reactor coolant. “My solutions are quite fundamental and problem-agnostic — applicable to the design of new reactors, to nuclear imaging systems for tumor detection, or to the plutonium battery of a Mars rover,” he says. “They will be useful anywhere scientists need to lower costs of high-fidelity nuclear simulations.”

But Ducru won’t be among the scientists deploying these computational advances. “I think we’ve done a good job, and others will continue in this area of research,” he says. “After six years of delving deep into quantum physics and statistics, I felt my next step should be a startup.”

Scaling up with shrimp

As he pivots away from academia and nuclear science, Ducru remains constant to his mission of addressing the climate problem. The result is Torana, a company Ducru and a partner started in 2018 to develop the financial products and services aquaculture needs to sustainably feed the world.

“I thought we could develop a scalable zero-carbon food,” he says. “The world needs high-nutrition proteins to feed growing populations in a climate-friendly way, especially in developing nations.” 

Land-based protein sources such as livestock can take a heavy toll on the environment. But shrimp, on the other hand, are “very efficient machines, scavenging crud at the bottom of the ocean and converting it into high-quality protein,” notes Ducru, who received the 2018 MIT Water Innovation Prize and the 2019 Rabobank-MIT Food and Agribusiness Prize to help develop his aquaculture startup (then called Velaron).

Torana is still in early stages, and Ducru hopes to apply his modeling expertise to build a global system of sustainable shrimp farming. His Schwarzman master thesis studies the role of aquaculture in our future global food system, with a focus on the shrimp supply chain.

In response to the Covid-19 pandemic, Ducru relocated to the family farm in southern France, which he helps run while continuing to follow the Tsinghua masters online and work on his MIT PhD. He is tweaking his business plans, and putting the final touches on his PhD research, including submitting several articles for publication. While it’s been challenging keeping all these balls in the air, he has supportive mentors — “Benoit Forget [CRPG director] has backed almost all my crazy ideas,” says Ducru. “People like him make MIT the best university on Earth.”

Ducru is already mapping out his next decade or so: grow his startup, and perhaps create a green fund that could underwrite zero-carbon projects, including nuclear ones. “I don’t have Facebook and don’t watch online series or TV, because I prefer being an actor, creating things through my work,” he says. “I’m a scientific entrepreneur, and will continue to innovate across different realms.”

A scientist turns to entrepreneurship syndicated from https://osmowaterfilters.blogspot.com/

[kathleenseiber]

Machine learning pushes quantum computing forward

Researchers have created a machine learning framework to precisely locate atom-sized quantum bits in silicon.

It’s a crucial step for building a large-scale silicon quantum computer, the researchers report. 

Here, Muhammad Usman and Lloyd Hollenberg of the University of Melbourne explain their research and what it means for the future of quantum computers:

Quantum computers are expected to offer tremendous computational power for complex problems—currently intractable even on supercomputers—in the areas of drug design, data science, astronomy, and materials chemistry among others.

The high technological and strategic stakes mean major technology companies as well as ambitious start-ups and government-funded research centers are all in the race to build the world’s first universal quantum computer.

Qubits and quantum computers

In contrast to today’s classical computers, where information is encoded in bits (0 or 1), quantum computers process information stored in quantum bits (qubits). These are hosted by quantum mechanical objects like electrons, the negatively charged particles of an atom.

Quantum states can also be binary and can be put in one of two possibilities, or effectively both at the same time���known as quantum superposition—offering an exponentially larger computational space with an increasing number of qubits.

This unique data crunching power is further boosted by entanglement, another magical property of quantum mechanics where the state of one qubit is able to dictate the state of another qubit without any physical connection, making them all 1’s for example. Einstein called it a “spooky action at distance.”

Different research groups in the world are pursuing different kinds of qubits, each having its own benefits and limitations. Some qubits offer potential for scalability, while others come with very long coherence times, that is the time for which quantum information can be robustly stored.

Qubits in silicon are highly promising as they offer both. Therefore, these qubits are one of the front-runner candidates for the design and implementation of a large-scale quantum computer architecture.

One way to implement large-scale quantum computer architecture in silicon is by placing individual phosphorus atoms on a two-dimensional grid.

The single and two qubit logical operations are controlled by a grid of nanoelectronic wires, bearing some resemblance to classical logic gates for conventional microelectronic circuits. However, key to this scheme is ultra-precise placement of phosphorus atoms on the silicon grid.

What’s holding things back? 

However, even with state-of-the-art fabrication technologies, placing phosphorus atoms at precise locations in silicon lattice is a very challenging task. Small variations, of the order of one atomic lattice site, in their positions are often observed and may have a huge impact on the efficiency of two qubit operations.

The problem arises from the ultra-sensitive dependence of the exchange interaction between the electron qubits on phosphorus atoms in silicon. Exchange interaction is a fundamental quantum mechanical property where two subatomic particles such as electrons can interact in real space when their wave functions overlap and make interference patterns, much like the two traveling waves interfering on water surface.

Exchange interaction between electrons on phosphorus atom qubits can be exploited to implement fast two-qubit gates, but any unknown variation can be detrimental to accuracy of quantum gate. Like logic gates in a conventional computer, the quantum gates are the building blocks of a quantum circuit.

For phosphorus qubits in silicon, even an uncertainty in the location of qubit atom of the order of one atomic lattice site can alter the corresponding exchange interaction by orders of magnitude, leading to errors in two-qubit gate operations.

Such errors, accumulated over the large-scale architecture, may severely impede the efficiency of quantum computer, diminishing any quantum advantage expected due to the quantum mechanical properties of qubits.

Pinpointing qubit atoms

So in 2016, we worked with the Center for Quantum Computation & Communication Technology researchers at the University of New South Wales, to develop a technique that could pinpoint exact locations of phosphorus atoms in silicon.

The technique, reported in Nature Nanotechnology, was the first to use computed scanning tunneling microscope (STM) images of phosphorus atom wave functions to pinpoint their spatial locations in silicon.

The images were calculated using a computational framework which allowed electronic calculations to be performed on millions of atoms utilizing Australia’s national supercomputer facilities at the Pawsey supercomputing center.

These calculations produced maps of electron wave function patterns, where the symmetry, brightness, and size of features was directly related to the position of a phosphorus atom in silicon lattice, around which the electron was bound.

The fact that each donor atom positions led to a distinct map, pinpointing of qubit atom locations, known as spatial metrology, with single lattice site precision was achieved.

The technique worked very well at the individual qubit level. However, the next big challenge was to build a framework that could perform this exact atom spatial pinpointing with high speed and minimal human interaction coping with the requirements of a universal fault tolerant quantum computer.

Machine learning to the rescue

Machine learning is an emerging area of research which is revolutionizing almost every field of research, from medical science to image processing, robotics, and material design.

A carefully trained machine learning algorithm can process very large data sets with enormous efficiency.

One branch of machine learning is known as convolutional neural network (CNN)—an extremely powerful tool for image recognition and classification problems. When a CNN is trained on thousands of sample images, it can precisely recognize unknown images (including noise) and perform classifications.

Recognizing that the principle underpinning the established spatial metrology of qubit atoms is basically recognizing and classifying feature maps of STM images, we decided to train a CNN on the computed STM images. The work is published in the NPJ Computational Materials journal.

The training involved 100,000 STM images and achieved a remarkable learning of above 99% for the CNN. We then tested the trained CNN for 17600 test images including blurring and asymmetry noise typically present in the realistic environments.

The CNN classified the test images with an accuracy of above 98%, confirming that this machine learning-based technique could process qubit measurement data with high-throughput, high precision, and minimal human interaction.

This technique also has the potential to scale up for qubits consisting of more than one phosphorus atoms, where the number of possible image configurations would exponentially increase. However, machine learning-based framework could readily include any number of possible configurations.

In the coming years, as the number of qubits increase and size of quantum devices grow, qubit characterization via manual measurements is likely to be highly challenging and onerous.

This work shows how machine learning techniques such as developed in this work could play a crucial role in this aspect of the realization of a full-scale fault-tolerant universal quantum computer—the ultimate goal of the global research effort.

Source: University of Melbourne

The post Machine learning pushes quantum computing forward appeared first on Futurity.

Machine learning pushes quantum computing forward published first on https://triviaqaweb.weebly.com/