Cookie Consent by Free Privacy Policy Generator
Close this search box.

Superposition Guy’s Podcast — Winners of the QuEra Challange at the MIT iQuHack Hackathon

“The Superposition Guy’s Podcast”, hosted by Yuval Boger, CMO at QuEra Computing

A team of undergraduates who excelled at MIT’s Quantum Hackathon are interviewed by Yuval Boger. They discuss their journey, the challenge they tackled—optimizing the “maximum independent set problem” on a quantum computer for machine learning applications—and their personal backgrounds and aspirations in quantum computing. They all exhibit a shared interest in leveraging quantum computing to solve complex problems and reflect on mentorship, teamwork, future career paths, and much more.

Listen on Spotify — here

Full Transcript

Yuval: Hello Kevin, hello Ishika, hello Zero, hello Alan, hello Jeff. Thanks for being with me today.

Responsive Image

Kevin: Happy to be here.

Ishika: Thanks for having us.

Zirui: Hello.

Yuval: So, who are you and what do you do? Why don’t we start with you, Kevin?

Kevin: My name is Kevin. I’m a junior at Brown. I’m studying chemical physics. I think, like most of us, I’m here because I did pretty well at MIT’s Quantum Hackathon this year.

Yuval: I think you did very well and the team did. And Ishika, who are you and what do you do?

Ishika: Yep, so my name is Ishika. I’m also a junior, currently studying at Brown University and I’m majoring in physics and computer science. I’m also very interested in quantum computing, like the rest of us, and was part of this team.

Yuval: And Zirui, how about you?

Zirui: Well, I’m Zirui. I’m from Purdue University. I’m currently a third year computer science major. I guess, similar to everyone here, I’m quite interested in quantum computing.

Yuval: Jeff?

Jeff: Yeah, I’m also a junior at Brown University studying chemical physics and I’m here because I’m obviously interested in quantum computing. Probably want to pursue that post college.

Yuval: And last, Alan?

Alan: Hello, I’m Alan. I’m a third year undergrad at MIT and I’m studying electrical engineering and computer science. I’m interested in the intersection between quantum computing and machine learning.

Yuval: Wonderful. So you are all part of a team that won a prize in the QuEra-specific challenge at MIT QHack this year. I think that when you arrived, you could have chosen other challenges, right? There are many companies that present the challenges. What made you choose this particular one from QuEra?

Kevin: I think something we were drawn to was the specificness of the challenge because some of the challenges suggested an open-ended solution, whereas QuEra had a very, it appeared like it was going to have a pretty rigid type of answer that it was looking for.

Ishika: Yeah, I was actually just going to add to that that I actually, I was pretty excited about the idea of maybe working on a neutral atom interface just because the mainstream in industry is usually superconducting qubits and other as such. And the idea of actually using the neutral atom architecture to create some sort of solution to a problem seemed very exciting to me personally.

Yuval: What was the challenge? What did you, what were you asked to do?

Alan: All right. So basically, QuEra’s quantum computer is this, they basically have a grid of Rydberg atoms, it’s called a Rydberg atom array, I believe. And so their quantum hardware is really like excellently suited to solve this specific problem called maximum independent set problem. So maximum independent set, if you imagine a graph with like nodes that could be connected to each other, the question is how do you choose the maximum number of nodes such that no two nodes are neighbors with each other? That is, they don’t share an edge. And so the way that they have their atoms set up in a grid with like attraction and repulsive forces within the actual quantum computer makes it really easy to like, it’s a very natural problem to solve for this computer. And so we were, the problem was to basically investigate how this problem could be solved on this quantum computer.

Yuval: I think the trigger for this problem was a JP Morgan article that was also using that computer. Can anyone tell me a little bit about what that article was and what were you trying to investigate?

Jeff: Yeah, so I don’t remember reading the article, but I do know it was like based off of like one paper they wrote. But the big thing is that when you have, you have a certain radius for which nodes are allowed to connect by an edge. And if you set that radius to be like three, which is also equal to another set of constants, it turns out that a quantum computer might have, or classical computers tend to be really slow when they do that problem, specifically for like a radius of three. And so the idea is that, you know, using a quantum computer, you could find a quantum advantage, like for this very specific case of the maximally independent set problem.

Yuval: How did you come up with the solution?

Ishika: I guess I could add in a little and if you want to like follow up, feel free. So basically the way we approached this problem was, well, it was set in three parts. There was like the theory aspect, then kind of like coding it and then also finding potential applications for solving this problem. So like more of the business side. So in terms of the more theoretical and experimenting with different algorithm side, we mainly tried to implement two different types of algorithms, which was the QAOA and also an adiabatic variational quantum algorithm. And I think what really helped us was the fact that we were able to kind of compare and find advantages and maybe like disadvantages to both algorithms and where they performed better for this sort of problem. And then using all of that, we experimented with parameters like the detuning, the Rabi indices and also all sorts of pulse schemes after which that was like one set of the problem. And then we had the business side where we kind of used a machine learning application to show why solving this NP-hard problem was actually useful.

Yuval: If you were able to solve this problem, what practical applications does it have? Why would someone want to use a computer to solve this problem?

Alan: I can take this one. Honestly, finding use cases for the maximum independent set problem is a little tricky because usually like natural problems that are in the real world require you to find the least amount of some resource that covers the entire area. That seems like a very natural problem. So this problem is more like finding the maximum number of resources to cover the area such that the resources are like diverse enough or like spread apart enough. We discovered an application in machine learning where we could reduce the size of a training set so that instead of using, for example, 5,000 training images, we could use 300 of the most diverse ones. And hopefully they encapsulate some of the same information so that there’s not much loss in accuracy, but there is a drastic reduction in compute time, training time. So the process of our application was we had a database of basically pictures of skin moles and the question is, is this cancerous and if so, what type of cancer is this? We convert all of these images using a very general embedding vector. And what these embedding vectors are is this compact way to represent everything. And it represents everything in the same 1536 dimensional vector. So once we have these images in this vector space, we use principal component analysis, which basically is a way to take your 1536 dimension vector and drag it all the way down to two dimensions, hopefully preserving most of the information, which we found that it did. And so now we have, instead of like training images, we basically had points on a grid where each point corresponds to one of the images. And this is a natural use case of the maximum independent set problem because now if we connect the closest points as neighbors, then we run the maximum independent set problem and we pick out the images such that there’s no two images that are very similar. And so what we found at the end of the day was that if we run this whole algorithm, that the training accuracy after training on 300 images was almost converging to the same accuracy as training on all 5,000 images, which was very impressive.

Yuval: So you use the quantum computer as a preprocessor to down select the number of training images. Is that correct?

Kevin: Yeah, exactly.

Yuval: And what was the division of labor inside the team? Maybe Zirui, maybe you can take that. Who did what?

Zirui: Well, it’s kind of hard to exactly say who did what, but I think the most obvious was Alan did most of the marketing and the application specific elements. The rest of us mostly tackled the algorithmic side, especially dissecting the theory and how to apply it to the actual quantum computer. Most of us spread out into various sections such as what specific parts of the algorithm we should tackle. I also, at the end of the day, I personally tackle the QAOA part. I personally tackle the post-processing elements. So yeah, it’s hard to say who exactly did what as if everyone kind of did their own thing. We kind of just combined our effort to dissect the algorithm and just implement it basically.

Yuval: Is this your first hackathon or have you done many hackathons, I mean individually or collectively?

Kevin: So at MIT’s Quantum Hackathon last year, Jeff and Alan and I were on a team together. And that was kind of the reason why we’re on a team this time, which is pretty cool and kind of one of the reasons why we wanted to go for this project compared to the other challenges. I think the others, Ishika and Zirui, you guys also have some experience too, right?

Zirui: I guess, admittedly, I’m the one with the least amount of hackathon experience. I’ve only attended one other hackathon prior to this.

Ishika: Oh no, that’s pretty much, I think the same as me. So I’ve attended a hackathon before this, but it wasn’t obviously a Quantum Hackathon. So this was much more interesting for me.

Yuval: So after this weekend, what did you feel that you learned other than how to eat cold pizza?

Zirui: I guess I’ll start. So for this specific problem, I have to say this is my first time working with a neutral atom platform. So of course it gave me some perspective into a different system besides just gate-based quantum computers. So I think that’s the most immediate knowledge I’ve gained. And I personally, I think I want to bring this neutral atom platform back to my school and just start working with my team around this new platform since it seems quite interesting in what it’s capable of.

Jeff: Yeah. Another fun thing, I think a lot of us are in the same boat. I didn’t know any Julia going into this and it is similar enough to Python, but it is slightly different and needing to learn very quickly on the fly how to use Julia, that was something I took out of it. I know how to code Julia now.

Ishika: I think for me, I kind of learned how to think about problems with the mindset of like, “Oh, what’s the computational cost of this?” Or “What are the benefits of this algorithm?” Whereas my previous experience with research papers and research in general was very theoretical physics concepts. Whereas this is more of an applied setting where I got to first theorize maybe some of these algorithms and then also apply it on hardware, experiment with it. And so I got a bit of the full process and I thought that was pretty interesting and useful and combined my physics and CS interest.

Alan: I definitely-

Yuval: How active were the mentors and what did they contribute to the team?

Jeff: Yeah. So, do you mind if I take this one? So we kind of shot ourselves in the foot a little bit at the very beginning of the hackathon. So Alan, he goes to MIT, so we were like, “Hey, rather than being this really noisy room, why don’t we go to a classroom till like five or six and then we’ll go from there?” And it turns out that the entire time we were in these classrooms, the mentors were going around helping everyone a lot with different issues that they were having with the code or just conceptualizing things. Meanwhile, we were doing the entire thing by ourselves and we were a little behind when we got back to the room. But luckily the mentors were there until eight and we had like half an hour, an hour with one of them. And then Pedro, one of the mentors, incredible, incredible guy. He is the GOAT. He said that he would show up to the MIT museum at 6.30 in the morning and then he did. And that was at the point where we were kind of like, I don’t know, kind of like flattening out like what, you know, we’re finally struggling a little bit with like what we could do and what we, you know, how we were going to finish up. And he showed up at 6.30 in the morning. He gave us a bunch of advice and we ended up finishing around like 9, 9.30 in the morning, right? Just in time for the 10 a.m. like deadline.

Alan: Yeah, he was so proactive.

Ishika: Oh, sorry. Sorry. Do you want to go ahead?

Jeff: No, you’re good.

Alan: Yeah, I was going to say I remember that one night that we came back to the main hall of the main room and like Pedro came up to us and he’s like, I’m sorry, I have to excuse myself. It’s my wife’s birthday. I’m like, please go ahead. Like he was very proactive. He was like definitely trying to help us out. Maybe at like his own expense really.

Kevin: Yeah, we definitely got a lot of help from Pedro when we were asking for it. But I think looking back, we might have had an advantage of forging our path by ourselves in the beginning because when we arrived back to the space after working by ourselves and we talked with the mentors, we realized, I think we made progress and we figured out how to apply the QAOA algorithm and the mentors were like, really? Like you got that one to work? Oh, like we’ve been working with everyone else and like we’ve kind of just been focusing on the adiabatic. And that was pretty cool to be like, oh, we’re struggling with that one and we’re kind of in the middle of in the weeds with the adiabatic algorithm. But the QAOA is working great. Like we’re getting awesome results with this. I don’t know if we would have gotten that otherwise.

Yuval: What would you like to do after college? Are you looking for an academic career? Are you looking to go into industry? Are you going to do something completely outside quantum computing? What would you hope to be doing? Kevin why don’t you go first and then Ishika and then Zirui and then Jeff and then Alan.

Kevin: I think I need a little bit more experience before I know if I want to specify or I guess focus on one a more industry setting versus an academic setting. But I do hope to go to graduate school and kind of specialize in  a niche so that I can work in the quantum computing field in some way and lead projects and kind of forge my own ideas. I think it would be really rewarding to be in a setting where I can contribute to research and then see kind of the fruit of my labor in terms of the outcomes. I know  the space is really working towards that right now and I don’t know how much time it’s going to take for real world applications to be accessible and really contributing to changing everyday life. But I would like to be a part of that for sure.

Ishika: I think for me it’s a lot of different options that I’m kind of juggling and trying to figure out what I actually want to do. One of the main ones I guess or more obvious ones is that I would like to go to a PhD program after graduating. Specifically something within quantum research maybe on the material side or maybe condensed matter. Not really sure. Another option is to go into industry, maybe work in some sort of tech setting or I don’t know graduate and do something entirely unexpected. I did want to go to art school for the longest time so maybe I’ll just do something creative instead. But yeah it’s a lot of different options and I feel like there’s pressure to decide especially just being an international student and not having a whole lot of time to plan ahead. But I’m very excited for whatever it is.

Zirui: Personally I think or I guess perhaps for I guess all of us here what we need is time not specifically just to make a decision but rather time for this industry to evolve to a point where it’s not only more accessible but has more applications. Now as for me I personally taught quite a few students here at my institution and I think what I want to do with quantum computing at least is to bring it to a field where it can actually be applied and be usable in some way. Where it’s not just this niche tool limited to perhaps just us who have more knowledge in it. Now whether that be grad school or perhaps in some industry it doesn’t matter as much to me. But what I want to create is this system or this usability for quantum computing and whether that takes time or more knowledge. Well I guess only time knows.

Jeff: Yeah and then for me I think the biggest thing is I really want to teach. I really enjoy teaching. I also really like research so I guess the best way to do both would probably be to be a professor somewhere. Like I tutor a lot like I tutor high schoolers I tutor intro chemistry here. I also teach like Kevin and I started a quantum computing club at Brown. Well Kevin founded it and I kind of like co-founded I guess TACON at the end but I teach like the educational workshops there and I’ve been teaching them so teaching people quantum computing has been really fun. And you know I know that 50 years from now quantum computing might not be taught I guess depending on how the field goes or maybe I won’t be the lucky professor that gets to teach it but I like teaching it. I like doing quantum computing research. So yeah probably get a PhD after this.

Alan: Yeah I have a personal philosophy that a lot of people probably will disagree with which is that I don’t really make my life plan ahead of time. I really just like seeing what opportunities come my way and just grab at whatever seems the most appealing because I feel like if I channel myself towards something specific then I’m only I’m narrowing my field of what I’m looking for. I try to keep things very open in terms of like what comes up. So I mean if I get a great grad school offer great. If I get a great intern like industry position great you know. So I try to be very flexible with what I do.

Yuval: And what is it in your history that made you get into quantum? Is it a is your physics teacher in high school is it a book or a movie that you saw? Maybe we go in reverse order. Alan, Jeff, Zirui, Ishika, and Kevin.

Alan: Yes so I think the way I entered the quantum computing field is also just as much of a fluke as anything else that’s really happened to me. So last year I had the great privilege and honor of hosting Kevin and Jeff as like you know well oh maybe not Jeff maybe just Kevin last year. But like I hosted him in my dorm room and so we got to know each other. We went on this team for IQ hack last year. I think they call it IQ hack. So really I just wanted to see what the field was about when I entered that. I had absolutely no prerequisite knowledge of quantum computing. And so we ended up winning the social impact award. And what happened was they sent us to Abu Dhabi for a different hackathon. Like it was beautiful. It was free of charge. Like they covered everything. And Sana Oda the person who hosts that hackathon is a really amazing person. She really put on a great hackathon in Abu Dhabi. And so we ended up well we got separated. They made the teams randomly. And Kevin, Jeff, and I we all got separated from each other. And I guess my team like had a really really great mentor because they put mentors on every individual team. So we won that hackathon. And then they invited us to talk about quantum computing in Geneva for this JESDA conference. And I that’s when I really had to get my act together with quantum. I took classes. I started learning about it really profusely. And this year we came back to IQUAC after one year to do instead of this open-ended challenge with some social flavoring to it we went to a really like industry specific problem and I thought that was a great like cycle back a great growth over the year from being like not knowing anything about quantum competing to like winning first place in a very technical problem.

Jeff: Yeah so for me like I’ve always kind of been interested in chemistry and physics. And up until like I got to Brown I kind of had a little imposter syndrome and I was like oh like I can’t study physics. I’m not smart enough. And then I went to like our intro gen chem class and it started off by talking about particle in a box and like intro quantum mechanics. And I was like wow I don’t care if I’m not smart enough. This is really cool. So from that moment on I was like ask her what I’m chemical physics. And then Kevin and I had the privilege of working with Brenda Rubenstein for like an internship over the summer. She’s a professor at Brown University. She works a lot of alternative computing stuff and she had an independent study in quantum computing like our sophomore fall. And so we took it. It was awesome. And then naturally that kind of led to saying hey how do we prove that we know about quantum computing. We go to a quantum hackathon. So we went there and seeing how like interdisciplinary everyone and how open everyone was to like I don’t know kind of accepting people into the field. Like it really really made me happy. Like I really wanted to do it after that. Knowing that like chemists, engineers, computer scientists, like mathematicians, like everyone’s in this field trying to play their part made me want to do my part if that makes sense. And then also something that like reaffirmed me wanting to do quantum computing is you know last year we didn’t do the best in terms you know of like sorry the way I want to say this is the amount that I learned over the span of one year is incredible. And the fact that like last year we went from code that did not place and I was a little confused on like you know how exactly do I do these kinds of algorithms. How do I do these kinds of things to like actually implementing them is crazy. Like I feel like I’ve learned so much more and it just reaffirmed my my want to be in this field.

Zirui: As for me I guess perhaps I’m the odd one out. I went into college not really knowing what I want to do. So as a college freshman I tried basically everything in across various fields and quantum computing was just so happened to be one of them. I guess I never really stuck with them my freshman year. It wasn’t until my sophomore year this is this was the fall of September sorry around like September and November of twenty twenty two. That’s when chat GPT exploded. And I thought to myself if I want to catch the next big thing before it becomes popular I think I have to stick with quantum computing. So in the end of course I stuck with it. Here I am. I’ve taught like hundreds of students now. And I’m still not I want to say I’m still competent enough in quantum computing but I’m slowly getting there I think.

Ishika: I think for me it kind of started in my sophomore spring where I’d kind of I’d heard this term of quantum computing several times by now and I was pretty deep into like experimental condensed matter research and I was like this feels like a big buzzword. What is it? It’s usually a viable physics research area. I want to explore it. Now’s the time. And so on a whim I just applied for a quantum research internship at Fermilab for the summer and I happened to get it and I was like you know what I’m going to explore and just try this out. And so I did. And I had one of the best times that summer working with people and researchers at Fermilab and I got paired up with a mentor at Rigetti and it was one of the coolest experiences working on the startup and also academic side. Just seeing how this like very niche research area had so many different people of different backgrounds different research areas. And I really enjoyed the sort of like optimization problems interesting like physics hardware problems as well. Also the material side was really connected to my previous knowledge in condensed matter physics. And so I found it this very cool combination of my interest in just physics and CS and in general my interest in just solving interesting weird problems. And so I kind of stuck on that and continued collaborating with people at Rigetti working over the year with them and that really helped me like develop my sort of foundations in this field. And you know it’s a continuous learning process and I just decided to go to this hackathon see what it’s about. And I met some brilliant people there and had a whole lot of fun.

Kevin: For me, Jeff and I’s paths have been pretty intertwined. We took gen chem together our freshman fall and we met because we were assigned as lab partners for the lab for that course and then it’s history ever since. But I can relate to Zirui with the not really knowing  what I wanted to do when I arrived to Brown. When I was applying to college I was fully set on being premed like wanting to go to medical school to be a doctor to kind of have impact through that way. And then by the time I arrived I realized that thinking through I maybe this wasn’t the path for me. But just in case I went through the intro chemistry sequence and I took some CS courses as well just to kind of test things out. And so the spring semester of my freshman year some of us in the previous course in general chemistry in the fall there was a head T.A. as a graduate student or he was a postdoc named Matthew Church and he offered for anyone who was successful in the gen chem course to take a graduate level statistical mechanics course in the spring. And this is pretty unheard of. This is a pretty, I don’t, I don’t think anything like that’s happened since for it to be like an open invitation to anyone who was successful in this 400 person class. But Jeff convinced me to shop the course with him, to go at the beginning of spring semester and then kind of seeing what the high level theoretical chemistry and physical chemistry looks like. I kind of by the end of semester we were, we were going to every single office hours offered, it was like three times a week we went to office hours and constantly asking questions and learning, and felt like I was learning for the sake of learning for my own interest and it was so rewarding. I think that was really, that was one of the roots took hold for me just for this general getting into this theory. And then as Jeff mentioned over the summer we worked with Brenda Rubenstein and our project was we wrote a Python program to calculate the entanglement of molecules using like a Hubbard model. And going into it I was like I want to practice these theories that I’ve been thinking about in this chemistry course that I took. And then we realized oh like this is, this is relevant for quantum computing and that’s like an interesting burgeoning technology. That’s something that you know you could work on now and it can have, it could just go some place which I just find really exciting. And then from there I just continued to explore the interest and now here I am.

Yuval: And as our time today together comes to a close I wanted to ask each of you a hypothetical. If you could have dinner with one of the quantum greats dead or alive who would that person be? Maybe Kevin and then Jeff, Ishika, Alan and Zirui.

Kevin: I think I would love to have dinner with Isaac Chuang with a caveat that I’d have to do so much preparation to get my questions and my understandings of the backgrounds for my questions in order because he’s such a brilliant man. And I know that I would probably be grilled in return for grilling him on the questions I have.

Jeff: Yeah I think I’d have to go with Isaac Chuang as well. I think so at last year’s hackathon Kevin and I like we really wanted to just shake his hand. It was like a big thing. We like he spoke at the event and we saw him sitting there and he was asking all these questions and he was you know like answering these questions just so articulate like it was just incredible to see him like in the space doing his thing. And we were just like we need to shake his hand. We need to talk to him. Like we have to like interact with this person. And we remember asking him like what drove you to make a quantum computing textbook 20 years ago? You know because at the time that wasn’t really a big thing. And he said well you know if you were doing research on something you know what would you want to make a textbook in and why? And I don’t know he just he’s just awesome to talk or at least even just like from that interaction I know that he’d be like a great person to like get dinner with. I’m gonna talk to you for a bit.

Alan: I have a little bit of a cop-out answer. Actually not many people think of like Einstein as like a quantum guy because he didn’t really believe in like a lot of the core tenets of quantum like quantum mechanics. But he actually was one of the people who developed this like quantum theory of light where like so he demonstrated the photoelectric effect. And so he actually was one of the founders of quantum like as like a study as like a huge field. Right. So obviously who wouldn’t love to like pick Einstein’s brain out. So that’s yeah that’s my answer.

Ishika: Is it me next? I don’t know. Okay. Okay this is pretty tough to answer but I know one piece of text that I’ve pretty much kept next to me since day one of quantum computing was a quantum engineer’s guide to superconducting qubits I think it’s called. Which is just this like pages and pages of just like fundamentals of quantum computing right from the start. And one of the authors of that is Will Oliver. And so I did get to see him at the MIT hackathon but I would really love to actually talk to him and understand you know his ideas about this whole field in depth. And maybe just another person Jens Koch who I think like created the Transmon qubit. I’ve seen his name pop up in so many papers and texts. It’s one of the most cited I think. That would be pretty cool as well.

Zirui: My answer is going to be a bit weird. I think the one I’m going to go with is Professor Peter Shor. I was lucky enough to speak with him briefly at the hackathon like during the kickoff. He seemed like a really cool guy to not just only speak to but I think again to answer my personal big question how they bring quantum to the general public. How they make quantum usable. I think he will be a key person in just being able to answer that for me. Just trying to bridge this gap between the quantum world and the real life application.

Yuval: Very good. So Alan, Jeff, Zirui, Ishika and Kevin thank you so much for spending some time with me today.

Group: Thank you for having us.

To subscribe to the audio podcast, please Spotify here

If you found this article to be informative, you can explore more current quantum news here, exclusives, interviews, and podcasts.


The Future of Materials Discovery: Reducing R&D Costs significantly with GenMat’s AI and Machine Learning Tools

When: July 13, 2023 at 11:30am

What: GenMat Webinar

Picture of Jake Vikoren

Jake Vikoren

Company Speaker

Picture of Deep Prasad

Deep Prasad

Company Speaker

Picture of Araceli Venegas

Araceli Venegas

Company Speaker

Yuval Boger

Share this article:


The Future of Materials Discovery: Reducing R&D Costs significantly with GenMat’s AI and Machine Learning Tools

When: July 13, 2023 at 11:30am

What: GenMat Webinar

Picture of Jake Vikoren

Jake Vikoren

Company Speaker

Picture of Deep Prasad

Deep Prasad

Company Speaker

Picture of Araceli Venegas

Araceli Venegas

Company Speaker

Keep track of everything going on in the Quantum Technology Market.

In one place.

Join Our Newsletter