Job Opportunities at CQSL At the Computational Quantum Science Lab we typically have several openings yearly (at the PhD/ Postdoc level). Applications are reviewed twice a year, at the beginning of April and at the beginning o...

I have a new opening for several funded PHD positions in my group (the Computational Quantum Science Lab, at EPFL). If you are a talented, motivated student, please apply here www.epfl.ch/labs/cqsl/jo.... I am especially looking to hire in scientific ML applications/NQS; not in quantum computing.

congratulazioni!

Or you can go here neos-server.org/neos/solvers... and wait about 0.02 seconds for the problem to be solved exactly 😂

Job Opportunities at CQSL At the Computational Quantum Science Lab we typically have several openings yearly (at the PhD/ Postdoc level). Applications are reviewed twice a year, at the beginning of April and at the beginning o...

I am looking for a talented postdoc to join my group at EPFL, in Lausanne, Switzerland. Goal is to develop and apply state-of-the-art neural quantum states for electronic structure and related applications. Position to be filled soon, excellent conditions! Apply here: www.epfl.ch/labs/cqsl/jo...

Two-panel drake meme, titled "Morris' algorithm be like". Top panel: Drake looks away, disapprovingly, next to the legend "Counting to n in O(log n) space". Bottom panel: Drake points approvingly at the legend "Counting to n in O(loglog n) space.

I recently published the LaTeX notes I took in three amazing classes this semester:
- Computational quantum physics (Prof. @gppcarleo.bsky.social)
- Quantum information theory (Prof. @qzoeholmes.bsky.social)
- Sublinear algorithms for big data analysis (Prof. Michael Kapralov)
Link below👇

Grassmann Variational Monte Carlo with neural wave functions Excited states play a central role in determining the physical properties of quantum matter, yet their accurate computation in many-body systems remains a formidable challenge for numerical methods. W...

New work with Douglas Hendry and Alessandro Sinibaldi: Grassmann Variational Monte Calro with neural wave functions arxiv.org/abs/2507.10287
This is a formalization and an extension (e.g. natural gradient descent-wise) of the nice excited-state framework for VMC developed by @davidpfau.com et al.

Congrats Samuele for this nice work, it was a fun collaboration with IBM Zurich!

Nuclear responses with neural-network quantum states We introduce a variational Monte Carlo framework that combines neural-network quantum states with the Lorentz integral transform technique to compute the dynamical properties of self-bound quantum man...

Paper alert, this time on response functions obtained inverting the Lorentz Integral Transform, especially useful when there is an unbounded spectrum: "Nuclear responses with neural-network quantum states" arxiv.org/abs/2504.20195

Foundation Neural-Network Quantum States Foundation models are highly versatile neural-network architectures capable of processing different data types, such as text and images, and generalizing across various tasks like classification and g...

The training framework is certainly important (for example one needs to use the "correct version" of SR in this case, but also the neural networks should be good enough to generalize. For Ising it definitely works with transformers ! arxiv.org/abs/2502.09488

Universal neural wave functions for high-pressure hydrogen We leverage the power of neural quantum states to describe the ground state wave function of solid and liquid dense hydrogen, including both electronic and protonic degrees of freedom. For static prot...

A foundation NQS allows to reduce (by several orders of magnitude!!) the cost needed in QMC/Diffusion Monte Carlo calculations to span the same phase diagram (while achieving higher accuracy than DMC). We also include, at T=0, the full quantum wave functions for protons, beyond Born-Opp. (2/2)

Universal neural wave functions for high-pressure hydrogen We leverage the power of neural quantum states to describe the ground state wave function of solid and liquid dense hydrogen, including both electronic and protonic degrees of freedom. For static prot...

Ab-initio preprint: we study high-pressure hydrogen over a relatively large range of pressures and temperatures (in the Born-Opp. approx.). Crucially, we do it with a **single** wave function for all values of proton configurations. Effort led by my PHD David Linteau, arxiv.org/abs/2504.07062 (1/2)

Nice work!

Simulating full quantum mechanical ground- and excited state surfaces with deep quantum Monte Carlo by Zeno Schätzle, Bernat Szabo and Alice Cuzzocrea.

arxiv.org/abs/2503.19847

🧵⬇️

Precise Quantum Chemistry calculations with few Slater Determinants Slater determinants have underpinned quantum chemistry for nearly a century, yet their full potential has remained challenging to exploit. In this work, we show that a variational wavefunction compose...

Read the full preprint here: arxiv.org/abs/2503.14502 5/5​

Our approach surpasses all second-quantized NQS results for molecules published so far, despite being a much simpler ansatz conceptually. This suggests that for small to intermediate molecules, fully correlated wave functions might not be necessary. 4/5​

Using optimized contractions, our method scales computational cost with the fourth power of the number of basis functions. Benchmarking against exact full-configuration interaction results, we achieved lower variational energies than CCSD(T) for several molecules in the double-zeta basis. 3/5​

While this ansatz is as old as quantum chemistry, fully optimizing it has been challenging. Our innovation lies in efficiently optimizing the determinants by leveraging the quadratic dependence of energy on selected parameters, allowing for exact optimization. 2/5​

Energy differences with respect to CCSD(T) in the cc-pVDZ basis set. The shaded area indicates results within chemical accuracy (1 kcal/mol) from FCI/DMRG, and the hatched area from CCSD(T). The molecules are ordered w.r.t increasing number of electrons, grouping those with equal numbers together. For those marked with * we employed a particle-hole transformation.

​Excited to share our latest quantum chemistry preprint led by Clemens Giuliani. We employ a "simple" variational wavefunction composed of a few hundred optimized non-orthogonal Slater determinants, achieving energy accuracies comparable to state-of-the-art methods. arxiv.org/abs/2503.14502 1/5​

Computational Quantum Science Lab at the APS Global Physics Summit At this year's APS Global Physics Summit in Anaheim, the Computational Quantum Science Lab showcased several contributions, spanning quantum dynamics, neural-network methodologies, topological quantum...

At #APS2025, our lab presented 8 innovative studies spanning quantum-classical hybrid simulations, neural-network quantum states, quantum dynamics, and quantum chemistry. Read the highlights of our contributions here! actu.epfl.ch/news/computa... #APSsummit

right...

In the absence of a community consensus on what it really means to obtain quantum advantage over **all possible** classical methods, why keep stirring controversy instead of stating that advantage is over some X or Y classical numerical methods "only"?(2/2)

How Scientists, Publishers and Investors Create Quantum Hype D-Wave’s fresh claim that it has achieved “quantum advantage” has sparked criticism of the company—and of the scientific process itself

If it wasn't clear enough already that I have nothing against the results of Dwave, and others, read this: www.scientificamerican.com/article/are-... The problem is not the results, it's the way they are presented, and how they are perceived more broadly. (1/2)

I am insinuating nothing. the preprint is pretty clear we analyze the diamond geometry. we will provide more geometries soon, but the burden of proof that you can beat "all classical methods" is on your side, since you decided to make this questionable (cynical?!) claim in the first place.

we compared against ground truth computed with MPS at the largest scale you provided, and also against your own experiment at the scale where mps was not available... I'm not sure what your remark is about

If they are easy as you suggest, why you claimed they would take ~200 years on one of the largest supercomputers available ? I'm confused.

in our article and also in the interview with NS I have always stated there is a running time advantage by the experiment. that's not at all the point. the point is that it's not a million years advantage, it's much smaller.

I agree the new scientist article could have been written better.

We will add more cases and sizes, but this already tells you that those estimates of millions years or so do not hold, and should not be used to conclude about s****macy.

The dwave paper estimates classical computation time based on extrapolations of tensor network calculations on smaller systems. We have done a simulation they were estimating it would take 200 years or so on one of the largest clusters available.

Assassination chain meme with the following captions: * D-Wave claims quantum supremacy arXiv:2403.00910 * tensor networks arXiv:2503.05693 * variational Monte Carlo arXiv:2503.08247

Summary of the situation so far.