arXiv highlights

Posting has been less frequent as I've been busy finishing revisions on some submitted manuscripts. Quite a few noteworthy preprints were posted this week:

 

Topogivity: A Machine-Learned Chemical Rule for Discovering Topological Materials

The authors propose a machine-learning approach for discovering new classes of topological materials. Looks interesting, but hard to say more because frustratingly all the details are delegated to supplementary materials which are not included...

Estimating the Euclidean Quantum Propagator with Deep Generative Modelling of Feynman paths

In the path integral formulation of quantum mechanics the probability amplitude of a particle transiting from state A to state B is given by summing over all possible trajectories between A and B, weighted by the action of each trajectory. While this is a theoretically elegant picture, the summation is extremely difficult to evaluate in practice due to the enormous number of possible trajectories. Here the authors show how machine learning techniques can be used to efficiently sample from those trajectories that contribute significantly to the transition amplitude.

 

Quantifying information scrambling via Classical Shadow Tomography on Programmable Quantum Simulators

By mapping evolution operators to density matrices in a higher-dimensional Hilbert space one can use shadow tomography to efficiently characterise quantum channels. Related works: arXiv:2110.02965 and arXiv:2110.03629.

Observation of wave-packet branching through an engineered conical intersection

When a Hamiltonian with two near-degenerate energy levels varies slowly in time, non-adiabatic Landau-Zener transitions between the energy eigenstates occur. A wavepacket initially in one eigenstate will end up in a superposition of the two eigenstates. A similar phenomenon plays a crucial role in certain chemical reaction dynamics and can be understood qualitatively in terms of the potential energy surfaces during the reaction. Quantum simulation of these complicated reaction dynamics is one potential near-term application of quantum processors, demonstrated here at a small scale using circuit QED.

Experimental observation of thermalisation with noncommuting charges

Macroscopic thermal states are described by conserved quantities such as the total energy or particle number. Curiously, the conserved quantities characterising certain thermal quantum states known as non-Abelian thermal states do not commute with each other and thus cannot simultaneously have well-defined values. This work reports the observation of a non-Abelian thermal state using a trapped ion quantum simulator.

 

To See a World in a Grain of Sand -- The Scientific Life of Shoucheng Zhang

Qi and Zhang's Reviews of Modern Physics article on topological insulators and superconductors was one of my first introductions to topological phases and I was shocked to hear of his passing in 2018. I enjoyed learning more about his life by reading this article.