ABSTRACT Demand for sustainable energy is growing, so generating energy from new, sustainable sources is becoming increasingly important. One approach is to utilize the thermoelectric (TE) effect, which involves sustainably generating electrical energy from waste heat. Single‐walled carbon nanotubes (SWCNTs) can be used as thermoelectric materials for this purpose. This study investigates how the solvents used to prepare self‐standing buckypapers affect the thermoelectric properties of two SWCNT types that differ in purity, structure, and thickness. Scanning electron microscopy (SEM), transmission electron microscopy (TEM), elemental analysis, and Raman investigations revealed clear differences between the two types of SWCNTs. The Raman spectra also show various interactions between SWCNTs and solvents. It is determined that the type of solvent exerts a significant change on the Seebeck coefficient (S) of Tuball SWCNT buckypapers between 14.34 and 60.11 µV/K, while only minor alterations in the S value are detected in Zeonano SWCNT (47.69–67.24 µV/K). The volume conductivities of the buckypapers are influenced by both the solvent type and the SWCNT. The power factor varied between 7.94 and 397.52 µW/(m∙K 2 ).
ABSTRACT The growing interest in artificial intelligence (AI) and edge computing has motivated the need for alternative computational frameworks such as neuromorphic computing, which can efficiently handle data‐intensive tasks. Reservoir computing (RC) is a neuro‐inspired approach for processing complex temporal data using a fixed network (the reservoir) and a simple trainable readout layer. To reduce the complexity of computation and energy consumption, physical reservoir computing (PRC) extends this concept to dynamical physical systems by exploiting their intrinsic nonlinearities and memory effects. Among the different substrates that have been proposed for PRC, microelectromechanical systems (MEMS) have shown promise due to their nonlinear mechanical behavior and direct interaction with physical stimuli. Also, being compact and scalable, MEMS devices are widely used as sensors and transducers in commercial applications, making them a suitable choice for low‐power in‐sensor computing and edge intelligence applications. This review discusses recent progress in MEMS‐based PRC with a focus on its underlying principles, device architectures, and the potential of MEMS in neuromorphic computing applications.
ABSTRACT Ultrafast measurement of broadband radio frequency (RF) spectrum is of great importance in wireless communication, radar, remote sensing, and radio astronomy research. Among them, the RF spectrum analyzer based on the Optical Short‐Time Fourier Transform has received extensive attention and research. It can quickly perform the short‐time Fourier transform (STFT) at over 10 MHz frame rate in the optical domain and intuitively analyze the signal under test (SUT) with a bandwidth over 40 GHz. In this article, we review recent methods for performance improvement of an ultrafast RF spectrum analyzer based on STFT. The measurement frame rate has increased to 10 MHz to improve performance and support applications. All these recent efforts could inspire more innovative applications of temporal imaging modalities in the future.
ABSTRACT Data Field Theory (DFT), introduced in, models learning as the evolution of a data field governed by a stochastic Ginzburg–Landau equation on a Riemannian manifold. The present work extends DFT with advanced mathematical structures: hyperbolic geometry, principal bundles, renormalization group methods, Lévy processes, noncommutative geometry, and higher category theory. Responding to reviewer comments on an earlier version, we make three specific contributions. First, we correct an erroneous claim about the ultraviolet beta function: the one‐loop UV divergence of the quartic coupling is curvature‐independent, and the modification of theory on hyperbolic space is therefore an infrared effect. We derive this rigorously via the Helgason–Fourier transform and prove that the correlation length is bounded by 1/ρ for every m 2 ≥ −ρ 2 , with equality precisely on the critical line m 2 = − ρ 2 . A full one‐loop heat‐kernel computation on a general smooth Riemannian background confirms that the 1/ε pole of the quartic coupling is curvature‐independent, with the curvature entering only through finite infrared contributions. Numerical computation of the integrated IR density of states confirms that hyperbolic phase space is suppressed below the spectral gap and enhanced above it. Second, we provide a systematic numerical test of the Lévy stochastic‐resonance prediction on five benchmark families: two metastable‐escape potentials, too many‐minima global‐optimization functions, and one gradient‐based learning task. On the bistable and periodic potentials, Cauchy noise escapes metastable states significantly faster than Gaussian noise of matched effective variance (p < 0.01 at every tested scale, up to a factor of 20). On the Rastrigin function and on a small MLP trained on two‐moons classification, the advantage vanishes; on Ackley the result is mixed. This establishes a clear regime of validity for the prediction: heavy‐tailed noise accelerates escape from a single dominant barrier but does not provide a general advantage on many‐minima landscapes or gradient‐based learning. A first empirical evaluation of the proposed gauge‐attention regularizer on a small transformer is also reported, confirming that the mapping is operational and compatible with standard training, and showing a + 5.6 percentage‐point mean improvement at large regularizer weight (not statistically significant at n = 5 seeds). Third, we correct all citation mismatches identified in review, provide a self‐contained definition of DFT, and restate Connes’ reconstruction theorem with its hypotheses correctly separated from its conclusion. All code and data are provided as Supplementary Material.
ABSTRACT Lightly Cu‐doped BiMn 7‐ x Cu x O 12 ( x = 0.05, 0.10, and 0.15) was investigated using high‐pressure synthesis, single‐crystal X‐ray diffraction, pair distribution function (PDF) analysis, Scanning Transmission Electron Microscope (STEM), magnetic measurements, and first‐principles calculations. All compositions retain an average monoclinic I 2/ m structure, while Cu substitution progressively suppresses the monoclinic distortion and drives the lattice toward a pseudo‐cubic metric symmetry. PDF analysis reveals increasing local structural disorder and reduced medium‐range coherence with increasing Cu concentration, despite preservation of the overall quadruple‐perovskite framework. Comparative single‐crystal refinements qualitatively favor Cu association with the octahedral Mn B sublattice over the square‐planar Mn A’ sites. However, due to the low Cu concentration and substantial systematic residual density, the Cu content was determined by EDS, and the local Cu distribution cannot be quantified independently of the X‐ray data. Magnetic measurements reveal an intrinsic transition near T M ∼ 50–60 K, accompanied by magnetic irreversibility, field‐dependent hysteresis, and unsaturated magnetization. An additional broad feature near T L ∼ 30K becomes weaker and less distinct with increasing Cu concentration. First‐principles calculations favor Cu occupation at the square‐planar sites, in contrast to the qualitative preference of the average B‐site refinement model.
