Neuroevolution, a field that draws inspiration from the evolution of brains in nature, harnesses evolutionary algorithms to construct artificial neural networks. It bears a number of intriguing capabilities that are typically inaccessible to gradient
-based approaches, including optimizing neural-network architectures, hyperparameters, and even learning the training rules. In this paper, we introduce a quantum neuroevolution algorithm that autonomously finds near-optimal quantum neural networks for different machine learning tasks. In particular, we establish a one-to-one mapping between quantum circuits and directed graphs, and reduce the problem of finding the appropriate gate sequences to a task of searching suitable paths in the corresponding graph as a Markovian process. We benchmark the effectiveness of the introduced algorithm through concrete examples including classifications of real-life images and symmetry-protected topological states. Our results showcase the vast potential of neuroevolution algorithms in quantum machine learning, which would boost the exploration towards quantum learning supremacy with noisy intermediate-scale quantum devices.
When the matter from a companion star is accreted towards the central compact accretor, i.e. a black hole (BH) or a neutron star (NS), an accretion disc and a jet outflow will form, providing bight X-ray and radio emission, which is known as X-ray bi
naries (XRBs). In the low/hard state, there exist disc-jet couplings in XRBs, but it remains uncertain whether the jet power comes from the disc or the central accretor. Moreover, BHXRBs have different properties compared with NSXRBs: quiescent BHXRBs are typically two to three orders of magnitude less luminous than NSXRBs in X-ray, whereas BHXRBs are more radio loud than NSXRBs. In observations, an empirical correlation has been established between radio and X-ray luminosity, $L_{rm R} propto L_{rm X}^b$, where $bsim 0.7$ for BHXRBs and $b sim 1.4$ for non-pulsating NSXRBs. However, there are some outliers of BHXRBs showing unusually steep correlation as NSXRBs at higher luminosities. In this work, under the assumption that the origin of jet power is related to the internal energy of the inner disc, we apply our magnetized, radiatively efficient thin disc model and the well-known radiatively inefficient accretion flow model to NSXRBs and BHXRBs. We find that the observed radio/X-ray correlations in XRBs can be well understood by the disc-jet couplings.
We study the spin transport through a 1D quantum Ising-XY-Ising spin link that emulates a topological superconducting-normal-superconducting structure via Jordan-Wigner (JW) transformation. We calculate, both analytically and numerically, the spectru
m of spin Andreev bound states and the resulting $mathbb{Z}_2$ fractional spin Josephson effect (JE) pertaining to the emerging Majorana JW fermions. Deep in the topological regime, we identify an effective time-reversal symmetry that leads to $mathbb{Z}_4$ fractional spin JE in the $textit{presence}$ of interactions within the junction. Moreover, we uncover a hidden inversion time-reversal symmetry that protects the $mathbb{Z}_4$ periodicity in chains with an odd number of spins, even in the $textit{absence}$ of interactions. We also analyze the entanglement between pairs of spins by evaluating the concurrence in the presence of spin current and highlight the effects of the JW Majorana states. We propose to use a microwave cavity setup for detecting the aforementioned JEs by dispersive readout methods and show that, surprisingly, the $mathbb{Z}_2$ periodicity is immune to $textit{any}$ local magnetic perturbations. Our results are relevant for a plethora of spin systems, such as trapped ions, photonic lattices, electron spins in quantum dots, or magnetic impurities on surfaces.
