Parametric amplification of attosecond coherent pulses around 100 eV at the single-atom level is demonstrated for the first time by using the 3D time-dependent Schrödinger equation in high-harmonic generation processes from excited states of He+. We present the attosecond dynamics of the amplification process far from the ionization threshold and resolve the physics behind it. The amplification of a particular central photon energy requires the seed XUV pulses to be perfectly synchronized in time with the driving laser field for stimulated recombination to the He+ ground state and is only produced in a few specific laser cycles in agreement with the experimental measurements. Our simulations show that the amplified photon energy region can be controlled by varying the peak intensity of the laser field. Our results pave the way to the realization of compact attosecond pulse intense XUV lasers with broad applications.Optimizing the shape of metasurface unit cells can lead to tremendous performance gains in several critically important areas. This paper presents a method of generating and optimizing freeform shapes to improve efficiency and achieve multiple metasurface functionalities (e.g., different polarization responses). The designs are generated using a three-dimensional surface contour method, which can produce an extensive range of nearly arbitrary shapes using only a few variables. Unlike gradient-based topology optimization, the proposed method is compatible with existing global optimization techniques that have been shown to significantly outperform local optimization algorithms, especially in complex and multimodal design spaces.Different techniques exist for determining chlorophyll-a concentration as a proxy of phytoplankton abundance. In this study, a novel method based on the spectral particulate beam-attenuation coefficient (cp) was developed to estimate chlorophyll-a concentrations in oceanic waters. A multi-layer perceptron deep neural network was trained to exploit the spectral features present in cp around the chlorophyll-a absorption peak in the red spectral region. Results show that the model was successful at accurately retrieving chlorophyll-a concentrations using cp in three red spectral bands, irrespective of time or location and over a wide range of chlorophyll-a concentrations.We describe a high-speed interferometric method, using multiple angles of incidence and multiple wavelengths, to measure the absolute thickness, tilt, the local angle between the surfaces, and the refractive index of a fluctuating transparent wedge. The method is well suited for biological, fluid and industrial applications.By computational optimization of air-void cavities in metallic substrates, we show that the local density of states (LDOS) can reach within a factor of ≈10 of recent theoretical upper limits and within a factor ≈4 for the single-polarization LDOS, demonstrating that the theoretical limits are nearly attainable. Optimizing the total LDOS results in a spontaneous symmetry breaking where it is preferable to couple to a specific polarization. Moreover, simple shapes such as optimized cylinders attain nearly the performance of complicated many-parameter optima, suggesting that only one or two key parameters matter in order to approach the theoretical LDOS bounds for metallic resonators.Ultra-thin metallic nanodisks, supporting localized plasmon (LP) modes, are used as a platform to facilitate high entanglement between distant quantum emitters (QEs). High Purcell factors, with values above 103, are probed for a QE placed near to an ultra-thin metallic nanodisk, composed of the noble metals Au, Ag, Al, and Cu. The disk supports two sets of localized plasmon modes, which can be excited by QEs with different transition dipole moment orientations. The two QEs are placed on opposite sides of the nanodisk, and their concurrence is used as a measure of the entanglement. We observe that the pair of QEs remains entangled for a duration that surpasses the relaxation time of the individual QE interacting with the metallic disk. Simultaneously, the QEs reach the entangled steady state faster than in the case where the QEs are in free space. Our results reveal a high concurrence value for a QES separation distance of 60 nm, and a transition energy of 0.8 eV (λ = 1550 nm). The robustness exhibited by this system under study paves the way for future quantum applications.Deep learning (DL) has been applied extensively in many computational imaging problems, often leading to superior performance over traditional iterative approaches. However, two important questions remain largely unanswered first, how well can the trained neural network generalize to objects very different from the ones in training? This is particularly important in practice, since large-scale annotated examples similar to those of interest are often not available during training. Second, has the trained neural network learnt the underlying (inverse) physics model, or has it merely done something trivial, such as memorizing the examples or point-wise pattern matching? This pertains to the interpretability of machine-learning based algorithms. In this work, we use the Phase Extraction Neural Network (PhENN) [Optica 4, 1117-1125 (2017)], a deep neural network (DNN) for quantitative phase retrieval in a lensless phase imaging system as the standard platform and show that the two questions are related and share a common crux the choice of the training examples. Moreover, we connect the strength of the regularization effect imposed by a training set to the training process with the Shannon entropy of images in the dataset. That is, the higher the entropy of the training images, the weaker the regularization effect can be imposed. P50515 We also discover that weaker regularization effect leads to better learning of the underlying propagation model, i.e. the weak object transfer function, applicable for weakly scattering objects under the weak object approximation. Finally, simulation and experimental results show that better cross-domain generalization performance can be achieved if DNN is trained on a higher-entropy database, e.g. the ImageNet, than if the same DNN is trained on a lower-entropy database, e.g. MNIST, as the former allows the underlying physics model be learned better than the latter.