共找到 20 条结果
Conclusions by Sears et al. (arXiv: 2601.04303) about dispersions of magnons in rutile FeF_2 are unsound because compulsory lattice vibrations are omitted in the analysis of their neutron scattering spectra. Hybridized magnon-phonon modes in magnetically ordered FeF_2 were observed and successfully analysed half a century ago.
Vision Transformer (ViT) architectures traditionally employ a grid-based approach to tokenization independent of the semantic content of an image. We propose a modular superpixel tokenization strategy which decouples tokenization and feature extraction; a shift from contemporary approaches where these are treated as an undifferentiated whole. Using on-line content-aware tokenization and scale- and shape-invariant positional embeddings, we perform experiments and ablations that contrast our approach with patch-based tokenization and randomized partitions as baselines. We show that our method significantly improves the faithfulness of attributions, gives pixel-level granularity on zero-shot unsupervised dense prediction tasks, while maintaining predictive performance in classification tasks. Our approach provides a modular tokenization framework commensurable with standard architectures, extending the space of ViTs to a larger class of semantically-rich models.
We present the Spectral Image Typer (SPIT), a convolutional neural network (CNN) built to classify spectral images. In contrast to traditional, rules-based algorithms which rely on meta data provided with the image (e.g. header cards), SPIT is trained solely on the image data. We have trained SPIT on 2,004 human-classified images taken with the Kast spectrometer at Lick Observatory with types of Bias, Arc, Flat, Science and Standard. We include several pre-processing steps (scaling, trimming) motivated by human practice and also expanded the training set to balance between image type and increase diversity. The algorithm achieved an accuracy of 98.7% on the held-out validation set and an accuracy of 98.7% on the test set of images. We then adopt a slightly modified classification scheme to improve robustness at a modestly reduced cost in accuracy (98.2%). The majority of mis-classifications are Science frames with very faint sources confused with Arc images (e.g. faint emission-line galaxies) or Science frames with very bright sources confused with Standard stars. These are errors that even a well-trained human is prone to make. Future work will increase the training set from Kast,
IP networks became the most dominant type of information networks nowadays. It provides a number of services and makes it easy for users to be connected. IP networks provide an efficient way with a large number of services compared to other ways of voice communication. This leads to the migration to make voice calls via IP networks. Despite the wide range of IP networks services, availability, and its capabilities, there still a large number of security threats that affect IP networks and for sure affecting other services based on it and voice is one of them. This paper discusses reasons of migration from making voice calls via IP networks and leaving legacy networks, requirements to be available in IP networks to support voice transport, and concentrating on SPIT attack and its detection methods. Experiments took place to compare the different approaches used to detect spam over VoIP networks.
This paper presents a formal framework for identifying and filtering SPIT calls (SPam in Internet Telephony) in an outbound scenario with provable optimal performance. In so doing, our work is largely different from related previous work: our goal is to rigorously formalize the problem in terms of mathematical decision theory, find the optimal solution to the problem, and derive concrete bounds for its expected loss (number of mistakes the SPIT filter will make in the worst case). This goal is achieved by considering an abstracted scenario amenable to theoretical analysis, namely SPIT detection in an outbound scenario with pure sources. Our methodology is to first define the cost of making an error (false positive and false negative), apply Wald's sequential probability ratio test to the individual sources, and then determine analytically error probabilities such that the resulting expected loss is minimized. The benefits of our approach are: (1) the method is optimal (in a sense defined in the paper); (2) the method does not rely on manual tuning and tweaking of parameters but is completely self-contained and mathematically justified; (3) the method is computationally simple and s
We search for stable bound states of non-extremal rotating three-charge black holes in five dimensions (Cvetic-Youm black holes) and supertubes. We do this by studying the potential of supertube probes in the non-extremal black hole background and find that generically the marginally bound state of the supersymmetric limit becomes metastable and disappears with non-extremality (higher temperature). However near extremality there is a range of parameters allowing for stable bound states, which have lower energy than the supertube-black hole merger. Angular momentum is crucial for this effect. We use this setup in the D1-D5 decoupling limit to map a thermodynamic instability of the CFT (a new phase which is entropically dominant over the black hole phase) to a tunneling instability of the black hole towards the supertube-black hole bound state. This generalizes the results of ArXiv:1108.0411 [hep-th], which mapped an entropy enigma in the bulk to the dual CFT in a supersymmetric setup.
A black hole observed during a dramatic 2023 eruption did not simply devour gas from its nearby companion star。 It also expelled large amounts of material through powerful jets and winds, even after the outburst had nearly faded。 The results suggest black holes may continue reshaping their surroundings long after their brightest fireworks end
This paper presents a framework for assigning intrinsic geometric structures to topological groups using only the data provided by their topological and algebraic structure. The geometrisation spits into small-scale and large-scale components, formalised respectively through local Lipschitz and quasimetric categories that, in turn, are definable from the canonical left uniform and left coarse structures of the group. For Polish groups, the paper characterises metrisability of the left coarse structure in terms of local boundedness, countable coverings by bounded sets, and the existence of compatible coarsely proper left-invariant metrics. It then introduces minimal metrics, which determine local Lipschitz structure, and maximal metrics, which determine quasimetric structure, and provides intrinsic characterisations of both. When both structures exist, they combine into a single canonical Lipschitz structure. Our framework is subsequently applied to specific examples such as homeomorphism groups, non-Archimedean Polish groups and automorphism groups of Fraïssé limits.
We study the instability in RuO$_2$ using the Hartree-Fock approximation followed by the random phase approximation. We employ a three-orbital Hubbard model without spin-orbit coupling. An analysis of the eigenvalues and eigenvectors of the static susceptibility in the non-magnetic phase for various local interaction parameters $U$, $J_H$, and hole doping $n$ shows that the spin susceptibility is the dominant response channel. In the stoichiometric system without spin-orbit coupling, commensurate altermagnetic order is identified as the leading instability at sufficiently low temperatures, whereas at higher temperatures or finite hole doping, incommensurate wave vectors emerge. To elucidate the origin of the magnetic instability, we analyze the band spitting by the staggered Weiss field and discuss the qualitative difference between altermagnets and antiferromagnets.
In this paper, we explore the relationship between quasisymmetric Schur $Q$-functions and peak Young quasisymmetric Schur functions. We introduce a bijection on $\mathsf{SPIT}(α)$ such that $\{\mathrm{w}_{\rm c}(T) \mid T \in \mathsf{SPIT}(α)\}$ and $\{\mathrm{w}_{\rm r}(T) \mid T \in \mathsf{SPIT}(α)\}$ share identical descent distributions. Here, $\mathsf{SPIT}(α)$ is the set of standard peak immaculate tableaux of shape $α$, and $\mathrm{w}_{\rm c}$ and $\mathrm{w}_{\rm r}$ denote column reading and row reading, respectively. By combining this equidistribution with the algorithm developed by Allen, Hallam, and Mason, we demonstrate that the transition matrix from the basis of quasisymmetric Schur $Q$-functions to the basis of peak Young quasisymmetric Schur functions is upper triangular, with entries being non-negative integers. Furthermore, we provide explicit descriptions of the expansion of peak Young quasisymmetric Schur functions in specific cases, in terms of quasisymmetric Schur $Q$-functions. We also investigate the combinatorial properties of standard peak immaculate tableaux, standard Young composition tableaux, and standard peak Young composition tableaux. We provide
We apply our idea, which previously we used in the analysis of the pure power NLS, consisting in spitting the virial inequality method into a large energy inequality combined with Kato smoothing, to the case of generalized Korteweg--De Vries pure power equations. We assume that a solution remains for all positive times very close to a soliton and then we prove an asymptotic stability result for $t\to +\infty$.
We investigate the influence of quark anomalous magnetic moments (AMMs) on the mass spectra of neutral pseudoscalar mesons ($π$, $K$, $η$, $η^{'}$) under external magnetic fields, finite temperatures, and quark chemical potentials using the three-flavor Nambu-Jona-Lasinio model. By incorporating AMMs at the quark level, we reveal that AMMs significantly alter the magnetic field dependence of constituent quark masses, inducing first-order phase transitions for light quarks at critical fields, while strange quarks exhibit nonmonotonic mass behavior. The inclusion of AMMs reshapes the QCD phase diagram, suppressing chiral transition temperatures, and the strong magnetic field shifts critical endpoints toward lower $μ$ and higher $T$ without AMMs. The crossover phase transition without AMMs is replaced by a first-order transition with AMMs under strong fields. Moreover, the inverse magnetic catalysis (IMC) induced by the introduction of AMMs qualitatively aligns with the predictions of lattice QCD (LQCD) for the dependence of phase transition temperature on the magnetic field. For mesons, a larger AMM triggers abrupt mass collapses and enhances flavor spitting at zero $μ$ and $T$ and a
We develop a unified framework that reconciles a barrier based geometric model of periodic sphere packings with a provably convergent discrete time dynamics. First, we introduce a C2 interior barrier U_nu that is compatible with a strict feasibility safeguard and has a Lipschitz gradient on the iterates domain. Second, we correct and formalize the discrete update and give explicit step size and damping rules. Third, we prove barrier to KKT consistency and state an interior variant clarifying the role of the quadratic term. Fourth, we show that strict prestress stability implies periodic infinitesimal rigidity of the contact framework. Fifth, we establish a Lyapunov energy descent principle, an energy nonexpansive feasibility projection (including a joint x and lattice basis B variant), and local linear convergence for the Spectral Projected Interior Trajectory (Spit) method. We also provide practical Hessian vector formulas to estimate smoothness and curvature, minimal schematic illustrations, and a short reproducibility stub. The emphasis is on rigorous assumptions and proofs; empirical evaluation is deferred.
It is well known that single real scalar field does not allow gauge coupling to the Abelian vector field. Using the complex scalar model as a starting point, we construct the Abelian gauge model with two real scalars. The gauge transformations for the scalars look different (albeit equivalent) from the conventional sQED. Spitting the masses of the scalars, or the scalar self-couplings, or the nonminimal parameters of scalar-curvature interaction, we arrive at a qualitatively new way of gauge symmetry breaking. Using the Schwinger-DeWitt technique, we explore the one-loop renormalization of this new model in curved spacetime.
Exploring combination of antiferromagnetic (AFM) spintronics and anomalous valley Hall effect (AVHE) is one of the most important questions for valleytronic applications. The key to address this issue is to achieve spin splitting around the valleys in AFM systems. Here, we propose a possible way for achieving AVHE in hexagonal AFM monolayer, which involves the isovalent alloying. This can break the combined symmetry ($PT$ symmetry) of spatial inversion ($P$) and time reversal ($T$), giving rise to spin splitting. More specifically, the large spin splitting around the Fermi energy level owes to $d$ orbital mismatch among these different transition metal ions. Based on first-principles calculations, the proposed way can be verified in out-of-plane AFM $\mathrm{CrMoC_2S_6}$ monolayer, which possesses spontaneous valley polarization and spitting splitting, providing possibility to realize AVHE. It is also proved that tensile strain can strengthen the valley splitting and maintain the out-of-plane AFM ordering. Our works provide an experimentally feasible way for developing AFM valleytronic devices.
The r-fold-n-point-splitting operation is an important operation in Graph Theory defined by Slater [15]. Later, Ghafari [6] extended 3-fold-n-point-splitting operation in binary matroids and obtained the result for Eulerian matroids whose 3-fold is Eulerian. In this paper, we give another approach to extend 3-fold-3-point-splitting in binary matroids in terms of splitting and haracterize binary gammoid whose 3-fold-3-point spitting is binary gammoid.
In this paper, we introduce a black-box prompt optimization method that uses an attacker LLM agent to uncover higher levels of memorization in a victim agent, compared to what is revealed by prompting the target model with the training data directly, which is the dominant approach of quantifying memorization in LLMs. We use an iterative rejection-sampling optimization process to find instruction-based prompts with two main characteristics: (1) minimal overlap with the training data to avoid presenting the solution directly to the model, and (2) maximal overlap between the victim model's output and the training data, aiming to induce the victim to spit out training data. We observe that our instruction-based prompts generate outputs with 23.7% higher overlap with training data compared to the baseline prefix-suffix measurements. Our findings show that (1) instruction-tuned models can expose pre-training data as much as their base-models, if not more so, (2) contexts other than the original training data can lead to leakage, and (3) using instructions proposed by other LLMs can open a new avenue of automated attacks that we should further study and explore. The code can be found at h
In this article, we establish a class of new projected type iteration methods based on matrix spitting for solving the linear complementarity problem. Also, we provide a sufficient condition for the convergence analysis when the system matrix is an $H_+$-matrix. We show the efficiency of the proposed method by using two numerical examples for different parameters. Keywords. Iterative method, Linear complementarity problem, $H_{+}$-matrix, $P$-matrix, Matrix splitting, Convergence.
Single spin state evolution induced by the Landau-Zener-Stückelberg-Majorana (LZSM) interference in a Zeeman-spit four level system in a periodically driven double quantum dot is studied theoretically by the Floquet stroboscopic method. An interplay between spin-conserving and spin-flip tunneling processes with the Electric Dipole Spin Resonance (EDSR) that is induced in an individual dot and enhanced by the LZSM multiple level crossings with the neighboring quantum dot is investigated as a function of the microwave (MW) frequency, driving amplitude, interdot detuning, and magnetic field. A number of special points in the parameter space are identified, out of which where all the three features are merged. Under this triple-crossing resonance condition the interdot tunneling is combined with a fast spin evolution in each dot at the EDSR frequency. Harmonics of the EDSR are revealed in the spin-dependent tunneling maps versus variable magnetic field and MW frequency. The results are applicable for both electron and hole systems with strong spin-orbit interaction and may be useful for developing new time-efficient schemes of the spin control and readout in qubit devices.
To date, Versatile Video Coding (VVC) has a more magnificent overall performance than High Efficiency Video Coding (HEVC). The Quadtree with Nested Multi-Type Tree (QTMT) coding block structure can substantially enhance video coding quality in VVC. However, the coding gain also leads to a greater coding complexity. Therefore, this letter proposes a Fast Decision Scheme Based on Structural Similarity Index Metric Variation (FDS-SSIMV) to solve this problem. Firstly, the Structural Similarity Index Metric Variation (SSIMV) characteristic among the sub coding units of the spit mode is illustrated. Next, to evaluate the SSIMV value, SSIMV measure strategies are designed for different split modes in this letter. Then, the desired split modes are selected by the SSIMV values. Experimental results show that the proposed method achieves 64.74\% average encoding Time Saving (TS) with a 2.79\% Bj$\varnothing$ntegaard Delta Bit Rate (BDBR), outperforming the benchmarks.