In a collaboration between Duke-NUS and the Center for Quantum Technologies of Singapore, we have multiple openings for GenAI for #drugdiscovery.
luongo.pro/openings/#gene…
Classical NN trainers willing to work with quantum people are welcome!
[Also, there is a little bit of quantum computing in this project, so math skills or previous quantum exposure is welcome]
Previous experience on drug-discovery is a big plus. We are available to start in the next months. Please spread the news! :)
In Duke-NUS medical school we have budget for 1 or 2 positions for researchers in (geometrical) deep learning. We are flexible on the length (6months/1 year) and the contract (research assistants, phd internship, postdocs). However, we are looking for candidates that know how to train NN.
We published this result last summer, but I never passed the energy threshold to make a post on twitter, and I thought it would be nice to contribute to this social with some useful content.
arxiv.org/abs/2407.20167
modular addition, modular addition by a constant, controlled modular addition, and controlled modular addition by a constant (for different architectures of modular adder!!)
Afterwards, we applied our MBU Lemma to all these results, improving each of them.
This very simple trick went unformalised(!) and unexploited (!!) for years. Our great intern Adithya Sireesh and now PhD student at Uni of Edinburgh, made a great job at formalizing into a THEOREM+PROOF all the architecture for ...
All modular adders fall in this circuit architecture. They compute a carry in an ancilla, and based on the value of the ancilla, they subtract or not the modulus.
MBU is very simple. When you have an ancilla qubit to uncompute, you just do a Hadamard, and measure. If you measure 0, you uncomputed the qubit, and if you measure 1 you have a phase kickback to perform (i.e. call the unitary that computed the ancilla qubit).
Do you want to optimize by 10 to 25% the #depth and the Toffoli #size of a #quantum modular adder? You can use MBU (Measurement Based Uncomputation). A short and easy thread. arxiv.org/abs/2407.20167
Came here looking for the quantum crowd. Super excited to see the lively discussions on not only #quantumphysics but also #scicomm, #womeninSTEM, and many more. Here is my team @qcrew-sg.bsky.social and I, pleased to be here! 😇
I did the same thing, without believing everyone lived up to 100. 😂
Is anyone around #Melbourne today for #QTML? I am happy to have a coffee (and do some slides?) together! #quantum people hit me up!
Weekly reminder that CQT is looking for postdocs in the CS department, working on quantum algorithms. :)
🖤☕🖤
I am hiring one postdoc for 1 year, starting March 2024, to work on quantum algorithms on graph problems, with applications to chemistry.
Who is at QTML tomorrow? Looking forward to meet friends and colleagues at CERN this week!
Still more qubits than Gidney’s :p
Really? I thought this result was using more qubits!
Indeed! But so what are good results in finance with non QMC advantage?
Hello quantum software developers! The Unitary Fund is running their yearly survey on quantum software.
www.surveymonkey.com/r/QuantumSur...
Take a look!
I’d also like to see more non-QMC based speed ups
Here is the thing… I haven’t found any convincing one. What do you think about “the problem with Grover Rudolph” paper?
A list of working use cases 😅
Heyyy! We have a new quantum friend on Bluesky! :)
So cool!
“Once you pick a side, if you're not loyal to that side, then who the hell are you?” - Harvey Specter.
Today I’ll learn what is a post-variational QNN!
