Featured
-
-
World View |
Artificial intelligence needs a scientific method-driven reset
AI needs to develop more solid assumptions, falsifiable hypotheses, and rigorous experimentation.
- Luís A. Nunes Amaral
-
Article |
Self-organized intracellular twisters
Cytoplasmic flows in the fruit fly oocyte can reorganize cellular components. These structured vortical flows arise through self-organizing dynamics of microtubules, molecular motors and cytoplasm.
- Sayantan Dutta
- , Reza Farhadifar
- & Michael J. Shelley
-
News & Views |
Intrinsic simplicity of complex systems
Predicting the large-scale behaviour of complex systems is challenging because of their underlying nonlinear dynamics. Theoretical evidence now verifies that many complex systems can be simplified and still provide an insightful description of the phenomena of interest.
- Jianxi Gao
-
Article |
The low-rank hypothesis of complex systems
Although using low-rank matrices is the go-to approach to model the dynamics of complex systems, its validity remains formally unconfirmed. An analysis of random networks and real-world data now sheds light on this low-rank hypothesis and its implications.
- Vincent Thibeault
- , Antoine Allard
- & Patrick Desrosiers
-
Article |
Quantum-inspired classical algorithms for molecular vibronic spectra
It has been suggested that Gaussian boson sampling may provide a quantum computational advantage for calculating the vibronic spectra of molecules. Now, an equally efficient classical algorithm has been identified.
- Changhun Oh
- , Youngrong Lim
- & Liang Jiang
-
-
Article |
Interactive cryptographic proofs of quantumness using mid-circuit measurements
Being able to perform qubit measurements within a quantum circuit and adapt to their outcome broadens the power of quantum computers. These mid-circuit measurements have now been used to implement a cryptographic proof of non-classical behaviour.
- Daiwei Zhu
- , Gregory D. Kahanamoku-Meyer
- & Christopher Monroe
-
Article
| Open AccessScaling and intermittency in turbulent flows of elastoviscoplastic fluids
Elastoviscoplastic fluids combine solid- and liquid-like behaviour depending on applied stress. Simulations of elastoviscoplastic fluids at high Reynolds number now show that plasticity plays a key role in the turbulent flows seen in these systems, leading for example to intermittency.
- Mohamed S. Abdelgawad
- , Ianto Cannon
- & Marco E. Rosti
-
Article |
A quantum complexity lower bound from differential geometry
Quantum operations can be considered as points in a high-dimensional space in which distance reflects the similarity of two operations. Applying differential-geometric methods in this picture gives insights into the complexity of quantum systems.
- Adam R. Brown
-
Research Briefing |
Understanding the formation of gas bubbles at liquid–liquid interfaces
The formation of bubbles at liquid–liquid interfaces is challenging to explain because gas pockets cannot be stabilized by cracks on solid impurities. Experiments show that a difference in the gas solubilities of two immiscible liquids provides a gas reservoir, which allows gas to accumulate at the interface, leading to bubble formation.
-
Comment |
The flow from simulation to reality
Fluid simulations today are remarkably realistic. In this Comment I discuss some of the most striking results from the past 20 years of computer graphics research that made this happen.
- Károly Zsolnai-Fehér
-
News & Views |
A cool quantum simulator
Experiments with ultracold atoms can be used to create nearly ideal quantum simulations of theoretical models. A realization of a model of exotic magnetism has tested the limits of what can be studied numerically on a classical computer.
- Evgeny Kozik
-
Article
| Open AccessTransition from sub-Rayleigh anticrack to supershear crack propagation in snow avalanches
Avalanches can occur when a porous snow layer lies beneath a dense cohesive snow slab. Field experiments and simulations now reveal different crack-propagation regimes in slab avalanches, similar to rupture propagation following an earthquake.
- Bertil Trottet
- , Ron Simenhois
- & Johan Gaume
-
Article |
The temporal rich club phenomenon
Uncovering structures in temporal networks requires different tools than in their static counterparts. A metric now quantifies whether the nodes with a large number of connections also tend to stay simultaneously connected for longer times.
- Nicola Pedreschi
- , Demian Battaglia
- & Alain Barrat
-
-
Article |
Measuring the capabilities of quantum computers
Evaluations of quantum computers across architectures need reliable benchmarks. A class of benchmarks that can directly reflect the structure of any algorithm shows that different quantum computers have considerable variations in performance.
- Timothy Proctor
- , Kenneth Rudinger
- & Robin Blume-Kohout
-
Article |
A rigorous and robust quantum speed-up in supervised machine learning
Many quantum machine learning algorithms have been proposed, but it is typically unknown whether they would outperform classical methods on practical devices. A specially constructed algorithm shows that a formal quantum advantage is possible.
- Yunchao Liu
- , Srinivasan Arunachalam
- & Kristan Temme
-
Letter |
Quantum advantage for computations with limited space
In general, it isn’t known when a quantum computer will have an advantage over a classical device. Now it’s proven that computers with limited working memory are more powerful if they are quantum.
- Dmitri Maslov
- , Jin-Sung Kim
- & Sarah Sheldon
-
Article |
Learning models of quantum systems from experiments
Quantum systems make it challenging to determine candidate Hamiltonians from experimental data. An automated protocol is presented and its capabilities to infer the correct Hamiltonian are demonstrated in a nitrogen-vacancy centre set-up.
- Antonio A. Gentile
- , Brian Flynn
- & Anthony Laing
-
Article |
Topological limits to the parallel processing capability of network architectures
The ability to perform multiple tasks simultaneously is a key characteristic of parallel architectures. Using methods from statistical physics, this study provides analytical results that quantify the limitations of processing capacity for different types of tasks in neural networks.
- Giovanni Petri
- , Sebastian Musslick
- & Jonathan D. Cohen
-
News & Views |
Work with what you’ve got
Quantum computing combines great promise with daunting challenges — the road to devices that solve real-world problems is still long. Now, an implementation of a quantum algorithm maps the problems we want to solve to the devices we already have.
- Boaz Barak
-
Article |
Quantum approximate optimization of non-planar graph problems on a planar superconducting processor
It is hoped that quantum computers may be faster than classical ones at solving optimization problems. Here the authors implement a quantum optimization algorithm over 23 qubits but find more limited performance when an optimization problem structure does not match the underlying hardware.
- Matthew P. Harrigan
- , Kevin J. Sung
- & Ryan Babbush
-
Measure for Measure |
Virtually a measurement
Simulations are as much a part of science as hypothesis and experiment. But can their outcomes be considered observations? Wendy S. Parker investigates.
- Wendy S. Parker
-
Comment |
Fixed-time descriptive statistics underestimate extremes of epidemic curve ensembles
The uncertainty associated with epidemic forecasts is often simulated with ensembles of epidemic trajectories based on combinations of parameters. We show that the standard approach for summarizing such ensembles systematically suppresses critical epidemiological information.
- Jonas L. Juul
- , Kaare Græsbøll
- & Sune Lehmann
-
Article |
Machine learning the thermodynamic arrow of time
The phrase ‘arrow of time’ refers to the asymmetry in the flow of events. A machine learning algorithm trained to infer its direction identifies entropy production as the relevant underlying physical principle in the decision-making process.
- Alireza Seif
- , Mohammad Hafezi
- & Christopher Jarzynski
-
News & Views |
Toward noise-robust quantum advantage
Near-term quantum computations are susceptible to noise that — left uncorrected — can destroy the correlations responsible for quantum computational speedups. New work develops tools for bolstering the noise resilience of these speedups.
- Bill Fefferman
-
Comment |
Understanding deep learning is also a job for physicists
Automated learning from data by means of deep neural networks is finding use in an ever-increasing number of applications, yet key theoretical questions about how it works remain unanswered. A physics-based approach may help to bridge this gap.
- Lenka Zdeborová
-
-
Article |
Quantum convolutional neural networks
A quantum circuit-based algorithm inspired by convolutional neural networks is shown to successfully perform quantum phase recognition and devise quantum error correcting codes when applied to arbitrary input quantum states.
- Iris Cong
- , Soonwon Choi
- & Mikhail D. Lukin
-
News & Views |
Nearly perfect quark–gluon fluid
A statistical analysis of data from ultra-relativistic heavy-ion collisions has uncovered the specific viscosities of the quark–gluon plasma — suggesting that the hottest matter in the current Universe behaves like a near-perfect fluid.
- Kari J. Eskola
-
Letter |
Bayesian estimation of the specific shear and bulk viscosity of quark–gluon plasma
As the quark–gluon plasma is a short-lived state of matter, its properties cannot be measured directly. A Bayesian parameter estimation method now provides a reliable estimation of the temperature-dependent specific shear and bulk viscosities.
- Jonah E. Bernhard
- , J. Scott Moreland
- & Steffen A. Bass
-
Article |
Low-dimensional dynamics of two coupled biological oscillators
Modelling and microscopy of thousands of cells together reveal the coupling through which the cell cycle influences the circadian clock. This coupling may explain why mammalian tissues growing at different rates have shifted circadian rhythms.
- Colas Droin
- , Eric R. Paquet
- & Felix Naef
-
Letter |
Classifying snapshots of the doped Hubbard model with machine learning
Quantum gas microscopes provide high-resolution real-space snapshots of quantum many-body systems. Now machine-learning techniques are used in choosing theoretical descriptions according to the consistency of their predictions with these snapshots.
- Annabelle Bohrdt
- , Christie S. Chiu
- & Michael Knap
-
Measure for Measure |
Unbridled mental power
Artificial intelligence is set to rival the human mind, just as the engine did the horse. José Hernández-Orallo looks at how we compare cognitive performance.
- José Hernández-Orallo
-
Article |
On the complexity and verification of quantum random circuit sampling
Evidence is provided that quantum random circuit sampling, a near-term quantum computational task, is classically hard but verifiable, making it a leading proposal for achieving quantum supremacy.
- Adam Bouland
- , Bill Fefferman
- & Umesh Vazirani
-
Perspective |
Beyond CMOS computing with spin and polarization
- Sasikanth Manipatruni
- , Dmitri E. Nikonov
- & Ian A. Young
-
Article |
Mutual information, neural networks and the renormalization group
Finding the relevant degrees of freedom of a system is a key step in any renormalization group procedure. But this can be difficult, particularly in strongly interacting systems. A machine-learning algorithm proves adept at identifying them for us.
- Maciej Koch-Janusz
- & Zohar Ringel
-
Letter |
Network structure from rich but noisy data
A technique allows optimal inference of the structure of a network when the available observed data are rich but noisy, incomplete or otherwise unreliable.
- M. E. J. Newman
-
Letter |
Neural-network quantum state tomography
Unsupervised machine learning techniques can efficiently perform quantum state tomography of large, highly entangled states with high accuracy, and allow the reconstruction of many-body quantities from simple experimentally accessible measurements.
- Giacomo Torlai
- , Guglielmo Mazzola
- & Giuseppe Carleo
-
News & Views |
Quantum advantage deferred
A type of optics experiment called a boson sampler could be among the easiest routes to demonstrating the power of quantum computers. But recent work shows that super-classical boson sampling may be a long way off.
- Andrew M. Childs
-
Letter |
Classical boson sampling algorithms with superior performance to near-term experiments
A classical algorithm solves the boson sampling problem for 30 bosons with standard computing hardware, suggesting that a much larger experimental effort will be needed to reach a regime where quantum hardware outperforms classical methods.
- Alex Neville
- , Chris Sparrow
- & Anthony Laing
-
Letter |
Numerical test of the Edwards conjecture shows that all packings are equally probable at jamming
A decades-old proposal that all distinct packings are equally probable in granular media has gone unproven due to the sheer number of packings involved. Numerical simulation now demonstrates that it holds — precisely at the jamming threshold.
- Stefano Martiniani
- , K. Julian Schrenk
- & Daan Frenkel
-
Letter |
Learning phase transitions by confusion
A neural-network technique can exploit the power of machine learning to mine the exponentially large data sets characterizing the state space of condensed-matter systems. Topological transitions and many-body localization are first on the list.
- Evert P. L. van Nieuwenburg
- , Ye-Hua Liu
- & Sebastian D. Huber
-
Letter |
The effect of a prudent adaptive behaviour on disease transmission
The common policy of replacing infected individuals with healthy substitutes can have the effect of accelerating disease transmission. A dynamic network model suggests that standard modelling approaches underplay the effect of network structure.
- Samuel V. Scarpino
- , Antoine Allard
- & Laurent Hébert-Dufresne
-
News & Views |
Machines learn to recognize glasses
The dynamics of a viscous liquid undergo a dramatic slowdown when it is cooled to form a solid glass. Recognizing the structural changes across such a transition remains a major challenge. Machine-learning methods, similar to those Facebook uses to recognize groups of friends, have now been applied to this problem.
- Michele Ceriotti
- & Vincenzo Vitelli
-
Letter |
General relativity and cosmic structure formation
When general relativity is included in large-scale simulations of the cosmic structure of the Universe, relativistic effects turn out to be small but measurable, thus providing tests for models of dark matter and dark energy.
- Julian Adamek
- , David Daverio
- & Martin Kunz
-
Letter |
A structural approach to relaxation in glassy liquids
The relation between structure and dynamics in glasses is not fully understood. A new approach based on machine learning now reveals a correlation between softness—a structural property—and glassy dynamics.
- S. S. Schoenholz
- , E. D. Cubuk
- & A. J. Liu
-
-
Article |
Secondary reconnection sites in reconnection-generated flux ropes and reconnection fronts
New three-dimensional simulations of magnetic reconnection suggest the existence of secondary reconnection sites that could be observed by the new NASA Magnetospheric MultiScale Mission.
- Giovanni Lapenta
- , Stefano Markidis
- & David L. Newman