Sign in

Jeongrak Son

@perp-waterfall.bsky.social
404 followers 296 following 32 posts

PhD student @ NTU Singapore Quantum Information and Thermodynamics jeongrak-son.github.io

PostsRepliesMedia
Reposted by Jeongrak Son
bihong.bsky.social @bihong.bsky.social · 15/01/2026
(2) quantum-journal.org/papers/q-202..., with coauthor Yudai Suzuki, @marekgluza.mathstodon.xyz.ap.brid.gy @perp-waterfall.bsky.social @nellynghy.bsky.social @qzoeholmes.bsky.social (meme creator). A new recipe for implementing polynomial transformation of a Hermitian matrix without post selection~~
quantum-journal.org
Double-bracket algorithm for quantum signal processing without post-selection
Yudai Suzuki, Bi Hong Tiang, Jeongrak Son, Nelly H. Y. Ng, Zoe Holmes, and Marek Gluza, Quantum 9, 1954 (2025). Quantum Signal Processing (QSP), a framework for implementing matrix-valued polynomials,...
141
Reposted by Jeongrak Son
bihong.bsky.social @bihong.bsky.social · 15/01/2026
Two double-bracket works just jumped out🧑‍🍳🧑‍🍳— (1) PRL journals.aps.org/prl/abstract... With coauthors @marekgluza.mathstodon.xyz.ap.brid.gy @perp-waterfall.bsky.social @nellynghy.bsky.social @qzoeholmes.bsky.social, Yudai Suzuki, René Zander, Raphael Seidel
131
Reposted by Jeongrak Son
Marek Gluza @marekgluza.mathstodon.xyz.ap.brid.gy · 12/01/2026
Double-bracket quantum imaginary-time evolution - a #quantum algorithm for implementing ground-states of local Hamiltonians is now published at PRL. journals.aps.org/prl/abstract/10.11… and the equivalent freely accessible preprint is here arxiv.org/abs/2412.04554 […]
mathstodon.xyz
Original post on mathstodon.xyz
022
Jeongrak Son @perp-waterfall.bsky.social · 10/09/2025
Thank you! It was great visiting you last year!
000
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
Same!! And the journey was great thanks to you and all my new friends! Hope to see you more often at conferences.
010
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
Thanks so much!! It was amazing collaborating with you!
020
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
Check all my papers on arXiv: arxiv.org/a/son_j_2.html The ones I covered in my thesis are: arxiv.org/abs/2209.15213, arxiv.org/abs/2303.13020, arxiv.org/abs/2403.09187, arxiv.org/abs/2412.04554, arxiv.org/abs/2412.06900
arxiv.org
010
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
My next step hasn't been decided yet, and I will be staying at NTU until (at most) Feb-Apr 2026 to wrap up several projects. You may soon hear about Gaussian phase-covariant operations or hybrid resource theories from me:)
110
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
My interest in catalysts grew to include other auxiliaries, particularly memories. We introduced several quantum algorithms that approximately solve certain non-linear differential equations and showed that using quantum memories without reading them can drastically improve these algorithms.
120
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
We then observed an important caveat (catalysis is fragile to noise) and characterised rather general classes of theories where robust catalysis is possible/impossible.
110
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
Since I first encountered catalysis in QI, I've been fascinated by how powerful it can be despite it returning to its exact original state. My PhD study started there: we identified a source of its power (memory effects!) and found that this alone can bridge different thermodynamic paradigms.
110
Jeongrak Son @perp-waterfall.bsky.social · 09/09/2025
Last week I defended my thesis "Quantum Qomrades: Catalysts in Resource Theories and Memories in Dynamic Programming". This journey was possible thanks to all my human comrades, especially my amazing adviser @nellynghy.bsky.social !
Jeongrak receives a cake and a hat for celebration
4221
Jeongrak Son @perp-waterfall.bsky.social · 12/08/2025
Yayyyy the title survived!!
010
Reposted by Jeongrak Son
Alex @itsmealeale.bsky.social · 11/08/2025
Our tutorial "A friendly guide to exorcising Maxwell's demon" (journals.aps.org/prxquantum/p...) is out! The big question after the tutorial is whether I’m finally done drawing demons?
151
Jeongrak Son @perp-waterfall.bsky.social · 23/07/2025
This is actually the first no-go result for robust catalysis (i.e. catalytic transformations that are catalytic even with small state preparation noise) outside completely resource non-generating operations. Next step: robust catalysis in LOCC or stabiliser operations?🥴🥴
040
Reposted by Jeongrak Son
Henrik Wilming @arrr.de · 23/07/2025
A fun project with Seok Hyung, Jeongrak ( @perp-waterfall.bsky.social ), @nellynghy.bsky.social , and Paul Boes. We dusted off some old notes because there seemed to be renewed interest in what differentiates thermal operations from Gibbs-preserving maps. arxiv.org/abs/2507.16637
arxiv.org
Thermal operations from informational equilibrium
Thermal operations are quantum channels that have taken a prominent role in deriving fundamental thermodynamic limitations in quantum systems. We show that these channels are uniquely characterized by...
0143
Reposted by Jeongrak Son
Nelly Ng @nellynghy.bsky.social · 23/07/2025
Years down the road, Jeongrak and I were trying to figure out whether robust catalytic advantage exists for thermal operations. We then realized that the conceptual key was hidden in those early, unannounced results by our friends all along! arxiv.org/abs/2412.06900
arxiv.org
Robust Catalysis and Resource Broadcasting: The Possible and the Impossible
In resource theories, catalysis refers to the possibility of enabling otherwise inaccessible quantum state transitions by providing the agent with an auxiliary system, under the condition that this au...
131
Reposted by Jeongrak Son
Nelly Ng @nellynghy.bsky.social · 23/07/2025
Whether or not you're a fan of thermal operations, there's something fundamentally special about them: by pinning down what it means to equilibrate, thermal operations uniquely emerge! With this, we also uncover nice hierarchy of unital channels, in contrast with the classical Birkhoff theorem.
2102
Reposted by Jeongrak Son
The Hilbert Space Post @hilbertspacepost.bsky.social · 22/07/2025
Grover's algorithm is reinterpreted as a product formula approximation of imaginary-time evolution in a new paper. @marekgluza.mathstodon.xyz.ap.brid.gy @perp-waterfall.bsky.social @nellynghy.bsky.social @qzoeholmes.bsky.social arxiv.org/abs/2507.15065
arxiv.org
Grover's algorithm is an approximation of imaginary-time evolution
We reveal the power of Grover's algorithm from thermodynamic and geometric perspectives by showing that it is a product formula approximation of imaginary-time evolution (ITE), a Riemannian gradient f...
151
Reposted by Jeongrak Son
Zoe Holmes @qzoeholmes.bsky.social · 22/07/2025
Finally, we show that quantum signal processing can be used to implement imaginary time evolution for unstructured search without post selection. And this enables us to design a new `fixed-point' quantum search algorithm i.e., a Grover type algorithm that never overshoots the solution
121
Reposted by Jeongrak Son
Zoe Holmes @qzoeholmes.bsky.social · 22/07/2025
Here's a new perspective on why Grover’s algorithm algorithm works: Unstructured search can be written as ground state problem. Then Grover's is just a product formula approximation of imaginary-time evolution or, equivalently, a Riemannian gradient flow on SU(d) to find this ground state.
1402
Reposted by Jeongrak Son
bihong.bsky.social @bihong.bsky.social · 22/07/2025
Check out this arXiv paper if you're interested! arxiv.org/abs/2507.15065
arxiv.org
Grover's algorithm is an approximation of imaginary-time evolution
We reveal the power of Grover's algorithm from thermodynamic and geometric perspectives by showing that it is a product formula approximation of imaginary-time evolution (ITE), a Riemannian gradient f...
031
Reposted by Jeongrak Son
bihong.bsky.social @bihong.bsky.social · 22/07/2025
Our Grover preprint just popped up! (with Yudai Suzuki, @qzoeholmes.bsky.social, @marekgluza.mathstodon.xyz.ap.brid.gy, @nellynghy.bsky.social, @perp-waterfall.bsky.social) Turns out… Grover's algorithm is secretly moonlighting as a first-order approximation to the imaginary time evolution!
1122
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
3) We have since introduced many algorithms that are quantum recursions (check out a series of papers on double-bracket quantum algorithms). In particular, we think that double-bracket quantum imaginary-time evolution (arxiv.org/abs/2412.04554) is a strong candidate for QDP application.
arxiv.org
Double-bracket quantum algorithms for quantum imaginary-time evolution
Efficiently preparing approximate ground-states of large, strongly correlated systems on quantum hardware is challenging and yet nature is innately adept at this. This has motivated the study of therm...
010
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
2) We also compare the circuit size (depth x width) with or without QDP. Sometimes, QDP reduces the circuit size. However, even when it does not, QDP allows us to have circuit depth-width tradeoff, which I believe is quite cool.
110
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
1) We now focus on pure quantum recursions (i.e. the recursive non-linear evolution of a pure state), for which we have a better performance guarantee: the circuit depth can be exponentially reduced by using exponentially many initial copies of the state.
110
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
For those who have only read the arXiv v1, here are some summaries of the changes:
110
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
It took us over a year to publish this, since we went through numerous revisions including significant strengthening of the results. The screenshot below is an excerpt from our reply to referees.
110
Jeongrak Son @perp-waterfall.bsky.social · 21/05/2025
Our paper (w/ Marek, @ryujitakagi.bsky.social, @nellynghy.bsky.social ) on Quantum Dynamic Programming has recently been published in PRL: journals.aps.org/prl/abstract... Xin Yi from CQT wrote a great highlight article for this paper: www.cqt.sg/highlight/20...
journals.aps.org
Quantum Dynamic Programming
We introduce a quantum extension of dynamic programming, a fundamental computational method that efficiently solves recursive problems using memory. Our innovation lies in showing how to coherently ge...
152
Jeongrak Son @perp-waterfall.bsky.social · 04/04/2025
Yess exactly! We are basically working in an effective qubit space spanned by \Psi and H \Psi, and the second unitary e^{i \theta \Psi} gives the relative phase difference.
010
Jeongrak Son @perp-waterfall.bsky.social · 04/04/2025
Thm.2 is derived from Eq. (16), which I am particularly fond of. This equation shows that any linear combination of a Hermitian matrix and the identity can be applied to a quantum state exactly and deterministically (given one can synthesise the exponential of a commutator).
170
Reposted by Jeongrak Son
Zoe Holmes @qzoeholmes.bsky.social · 03/04/2025
Today we posted a paper showing how a double-bracket quantum algorithm can implement quantum signal processing (DB-QSP), i.e., apply polynomial functions of operators to states. Crucially our approach doesn't need any post-selection - but this comes at the expense of increased circuit depths.
2333
Reposted by Jeongrak Son
The Hilbert Space Post @hilbertspacepost.bsky.social · 03/04/2025
A new paper introduces Double-Bracket QSP (DB-QSP), a novel quantum signal processing approach that eliminates the need for auxiliary qubits and post-selection by using approximate unitary synthesis. arxiv.org/abs/2504.01077
arxiv.org
Double-bracket algorithm for quantum signal processing without post-selection
Quantum signal processing (QSP), a framework for implementing matrix-valued polynomials, is a fundamental primitive in various quantum algorithms. Despite its versatility, a potentially underappreciat...
131
Reposted by Jeongrak Son
Mark M. Wilde @markwilde.bsky.social · 18/03/2025
Nelly Ng "Robust Catalysis and Resource Broadcasting: The Possible and the Impossible" #QuantumResources2025 I asked if there might be implications for Deutschian closed timelike curves. When I asked my question, I wasn't so serious, but thinking more afterwards, I think there is more to explore
3132
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
Joint work with amazing collaborators: Ray Ganardi, @sminagawa.bsky.social, Francesco Buscemi, Seok Hyung Lie, and @nellynghy.bsky.social
010
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
If you've met me this year, you must have seen how obsessed I was with this project. We have so many open questions that I'd love to discuss if you are interested (e.g. exactly when does a theory become non-extensive).
130
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
We also compiled the list of theories with/without robust catalysis.
Table summarising resource theories with/without robust catalysis.
110
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
4. For certain classes of resource theories, resource broadcasting is determined by the max-relative entropy of the resource.
100
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
3. We explicitly studied minimal and maximal ways of defining the set of free states for a multi-partite system, given the set of free states in each subsystem. Depending on this choice of composition, the possibility of robust catalysis and resource broadcasting is determined.
100
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
2. Such robust catalysis is intricately related to resource broadcasting: you can broadcast the resource from one system to another if and only if robust catalysis exists.
110
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
Some highlights of the paper: 1. It doesn't matter how much error you tolerate. If catalysis is robust under the smallest error, it must be robust for any amount of errors.
100
Jeongrak Son @perp-waterfall.bsky.social · 11/12/2024
Another paper of the week! scirate.com/arxiv/2412.0... In resource theories, catalysis provides many amazing advantages. However, it turns out that whenever there is system preparation error, most of them stop being catalytic. Fret not! We introduced a new class that is robust under such errors.
catalytic advantages go away as soon as state preparation noise comes
191
Reposted by Jeongrak Son
Zoe Holmes @qzoeholmes.bsky.social · 09/12/2024
Nature cools things easily but getting a quantum computer to do it is hard! Give us an approx ground state, we present an algorithm that approximates imaginary time evolution to: - Cool that state by an amount proportional to its energy fluctuations - Increase its fidelity with the ground state
0353
Jeongrak Son @perp-waterfall.bsky.social · 09/12/2024
Fourth paper of Marek's double-bracket quantum algorithms (DBQA) series (and the second one I'm participating in)! A joint work with amazing collaborators: Marek Gluza, Bi Hong Tiang, Yudai Suzuki, @qzoeholmes.bsky.social , and @nellynghy.bsky.social
List of DBQA series on arXiv
020
Jeongrak Son @perp-waterfall.bsky.social · 09/12/2024
2. It is known that double-bracket flows are steepest-decent flows. Hence, our algorithm coherently implements steepest-decent minimisation of the average energy without resorting to variational methods.
110
Jeongrak Son @perp-waterfall.bsky.social · 09/12/2024
Other highlights of the paper: 1. The compiled algorithm retains all the good properties of QITE. The average energy of the state reduces proportionally to the variance of the energy (we call this fluctuation-refrigeration relation😉). The ground state infidelity reduces exponentially.
130
Jeongrak Son @perp-waterfall.bsky.social · 09/12/2024
Good news: our compilation is algorithmic, and we don't need variational methods, block encodings, etc.
110
Jeongrak Son @perp-waterfall.bsky.social · 09/12/2024
New preprint on arXiv! scirate.com/arxiv/2412.0... Quantum imaginary-time evolution (QITE) is amazing, but its compilation is not straightforward. We found an iterative way of doing this, using the equivalence of QITE and double-bracket flows.
scirate.com
Double-bracket quantum algorithms for quantum imaginary-time evolution
Efficiently preparing approximate ground-states of large, strongly correlated systems on quantum hardware is challenging and yet nature is innately adept at this. This has motivated the study of therm...
1110
Reposted by Jeongrak Son
arXiv bot (quant-ph) @krxiv-quant-ph.bsky.social · 09/12/2024
Double-bracket quantum algorithms for quantum imaginary-time evolution arxiv.org/pdf/2412.04554 Marek Gluza, Jeongrak Son, Bi Hong Tiang, Yudai Suzuki, Zoë Holmes, Nelly H. Y. Ng
arxiv.org
https://arxiv.org/abs/2412.04554
arXiv abstract link
051