623509 research outputs found
Sort by
Anisotropic gravastar as horizonless regular black hole spacetime and its images illuminated by thin accretion disk
No full textA connection between regular black holes and horizonless ultracompact objects was proposed in~\cite{Carballo-Rubio:2022nuj}. In this paper, we construct a model of a horizonless compact object, specifically an anisotropic gravastar with continuous pressure, that corresponds to regular black hole spacetime in the appropriate limit. The construction begins by modeling an equation of state that satisfies the anisotropic gravastar conditions and transitions to the de Sitter () upon horizon formation. The spacetime structure is similar to the {\it Quantum Horizonless Compact Object} (QHCO) described in~\cite{Chen:2024ibc}. Within this model, we also generate images of the corresponding objects surrounded by a thin accretion disk. The resulting images reveal that assuming that the emitting matter exists only outside the object, the inner light ring structure closely resembles that of the horizonless configuration of a regular black hole and the QHCO, yet it exhibits a distinct light ring structure compared to the thin-shell gravastar model. However, the opposite occurs when emitting matter is taken into account inside the object
Convergence of Nonmonotone Proximal Gradient Methods under the Kurdyka-Lojasiewicz Property without a Global Lipschitz Assumption
No full textWe consider the composite minimization problem with the objective function being the sum of a continuously differentiable and a merely lower semicontinuous and extended-valued function. The proximal gradient method is probably the most popular solver for this class of problems. Its convergence theory typically requires that either the gradient of the smooth part of the objective function is globally Lipschitz continuous or the (implicit or explicit) a priori assumption that the iterates generated by this method are bounded. Some recent results show that, without these assumptions, the proximal gradient method, combined with a monotone stepsize strategy, is still globally convergent with a suitable rate-of-convergence under the Kurdyka-Lojasiewicz property. For a nonmonotone stepsize strategy, there exist some attempts to verify similar convergence results, but, so far, they need stronger assumptions. This paper is the first which shows that nonmonotone proximal gradient methods for composite optimization problems share essentially the same nice global and rate-of-convergence properties as its monotone counterparts, still without assuming a global Lipschitz assumption and without an a priori knowledge of the boundedness of the iterates
Semi-Automatic Extraction of Formal Models from Object Oriented Code
No full textBehavioral models are incredibly useful for understanding and validating software. However, the automatic extraction of such models from actual industrial code remains a largely unsolved problem with current solutions often not scaling well with the complexity and size of industrial systems or having to rely on approximations. To enable the extraction of useful models from code, we provide a framework for transforming object-oriented code into processes from which, when paired with minimal user input, models can be automatically generated and composed. Paired with this, we introduce the novel SSTraGen (StateSpace Transformation & Generation) tool, which provides an implementation of this framework. Through case studies at Philips Image Guided Therapy Systems, we showcase the practical applicability and usefulness of this tool, including the transformation of a component with >1000 LOC
Chemical Evolution during Molecular Cloud Formation Triggered by an Interstellar Shock Wave: Dependence on Shock Parameters and Comparison with Molecular Absorption Lines
No full textWe investigate chemistry in the compression layer behind the interstellar shock waves, where molecular cloud formation starts. We perform three-dimensional magnetohydrodynamics simulations of converging flows of atomic gas with shock parameters of inclination between the interstellar magnetic field and the shock wave, pre-shock density, and shock velocity. Then we derive 1D mean-flow models, along which we calculate a detailed gas-grain chemical reaction network as a post process with various chemical parameters, i.e. cosmic-ray ionization rate, abundances of PAHs, and metals in the gas phase. While carbon chains reach their peak abundances when atomic carbon is dominant in the pseudo-time-dependent models of molecular clouds, such behavior is less significant in our models since the visual extinction of the compression layer is low ( mag) when atomic carbon is abundant. Carbon chains, CN, and HCN increase at mag, where the gas-phase C/O ratio increases due to water ice formation. Shock parameters affect the physical structure and the evolutional timescale of the compression layer, and thus molecular evolution. Carbon chains are more abundant in models with higher post-shock density and slower gas accumulation. We calculate molecular column densities in the compression layer and compare them with the observations of diffuse and translucent clouds, which show reasonable agreement for water ice, carbon chains, and HCO. The observed variation of their column densities could be due to the difference in shock parameters and chemical parameters. The column density of CN is overestimated, for which we discuss possible reasons.Accepted in MNRAS. 20 pages, 14 figure
Variational learning of integrated quantum photonic circuits
No full textIntegrated photonic circuits play a crucial role in implementing quantum information processing in the noisy intermediate-scale quantum (NISQ) era. Variational learning is a promising avenue that leverages classical optimization techniques to enhance quantum advantages on NISQ devices. However, most variational algorithms are circuit-model-based and encounter challenges when implemented on integrated photonic circuits, because they involve explicit decomposition of large quantum circuits into sequences of basic entangled gates, leading to an exponential decay of success probability due to the non-deterministic nature of photonic entangling gates. Here, we present a variational learning approach for designing quantum photonic circuits, which directly incorporates post-selection and elementary photonic elements into the training process. The complicated circuit is treated as a single nonlinear logical operator, and a unified design is discovered for it through variational learning. Engineering an integrated photonic chip with automated control, we adjust and optimize the internal parameters of the chip in real time for task-specific cost functions. We utilize a simple case of designing photonic circuits for a single ancilla CNOT gate with improved success rate to illustrate how our proposed approach works, and then apply the approach in the first demonstration of quantum stochastic simulation using integrated photonics
Towards Unifying Feature Interaction Models for Click-Through Rate Prediction
No full textModeling feature interactions plays a crucial role in accurately predicting click-through rates (CTR) in advertising systems. To capture the intricate patterns of interaction, many existing models employ matrix-factorization techniques to represent features as lower-dimensional embedding vectors, enabling the modeling of interactions as products between these embeddings. In this paper, we propose a general framework called IPA to systematically unify these models. Our framework comprises three key components: the Interaction Function, which facilitates feature interaction; the Layer Pooling, which constructs higher-level interaction layers; and the Layer Aggregator, which combines the outputs of all layers to serve as input for the subsequent classifier. We demonstrate that most existing models can be categorized within our framework by making specific choices for these three components. Through extensive experiments and a dimensional collapse analysis, we evaluate the performance of these choices. Furthermore, by leveraging the most powerful components within our framework, we introduce a novel model that achieves competitive results compared to state-of-the-art CTR models. PFL gets significant GMV lift during online A/B test in Tencent\u27s advertising platform and has been deployed as the production model in several primary scenarios
StrTune: Data Dependence-based Code Slicing for Binary Similarity Detection with Fine-tuned Representation
No full textBinary Code Similarity Detection (BCSD) is significant for software security as it can address binary tasks such as malicious code snippets identification and binary patch analysis by comparing code patterns. Recently, there has been a growing focus on artificial intelligence-based approaches in BCSD due to their scalability and generalization. Because binaries are compiled with different compilation configurations, existing approaches still face notable limitations when comparing binary similarity. First, BCSD requires analysis on code behavior, and existing work claims to extract semantic, but actually still makes analysis in terms of syntax. Second, directly extracting features from assembly sequences, existing work cannot address the issues of instruction reordering and different syntax expressions caused by various compilation configurations. In this paper, we propose StrTune, which slices binary code based on data dependence and perform slice-level fine-tuning. To address the first limitation, StrTune performs backward slicing based on data dependence to capture how a value is computed along the execution. Each slice reflects the collecting semantics of the code, which is stable across different compilation configurations. StrTune introduces flow types to emphasize the independence of computations between slices, forming a graph representation. To overcome the second limitation, based on slices corresponding to the same value computation but having different syntax representation, StrTune utilizes a Siamese Network to fine-tune such pairs, making their representations closer in the feature space
Dynamics of phagocytosis through interplay of forces
No full textPhagocytosis is the process by which cells, which are 5 to 10 times larger than the particle size, engulf particles, holding substantial importance in various biological contexts ranging from the nutrient uptake of unicellular organisms to immune system of humans, animals etc. While the previous studies focused primarily on the mechanism of phagocytosis, in this study we have a taken a different route by studying the dynamics of the phagocytes in a system consisting of many bacteria and a small number of phagocytes. We put forward a minimalist framework that models bacteria and phagocytes as active and passive circular disks, respectively. The interactions are governed by directional forces: phagocytes are attracted toward bacteria, while bacteria experience a repulsive force in proximity to phagocytes. Bacteria are capable of reproduction at a fixed rate, and the balance between bacterial reproduction and phagocytic engulfment is governed by the interplay of the two opposing forces. In attraction dominated regimes, bacterial populations decrease rapidly, while in repulsion dominated regimes, bacterial clusters grow and impede phagocytes, often resulting in phagocyte trapping. Conversely, in attraction-dominated scenarios, only a few bacteria remain at later times, rendering the motion of the phagocytes diffusive. Further, the transition between the two regimes occurs through a regime of bi-stability. Our study further describes the dynamics of both species using the tools of statistical analysis, offering insights into the internal dynamics of this system.9 pages, 6 figure
Near-Optimal Time-Sparsity Trade-Offs for Solving Noisy Linear Equations
No full textWe present a polynomial-time reduction from solving noisy linear equations over in dimension with a uniformly random coefficient matrix to noisy linear equations over in dimension where each row of the coefficient matrix has uniformly random support of size . This allows us to deduce the hardness of sparse problems from their dense counterparts. In particular, we derive hardness results in the following canonical settings. 1) Assuming the -dimensional (dense) LWE over a polynomial-size field takes time , -sparse LWE in dimension takes time 2) Assuming the -dimensional (dense) LPN over takes time , -sparse LPN in dimension takes time These running time lower bounds are nearly tight as both sparse problems can be solved in time given sufficiently many samples. We further give a reduction from -sparse LWE to noisy tensor completion. Concretely, composing the two reductions implies that order- rank- noisy tensor completion in takes time , assuming the exponential hardness of standard worst-case lattice problems.Abstract shortened to match arXiv requirement
Rethinking Top Probability from Multi-view for Distracted Driver Behaviour Localization
No full textNaturalistic driving action localization task aims to recognize and comprehend human behaviors and actions from video data captured during real-world driving scenarios. Previous studies have shown great action localization performance by applying a recognition model followed by probability-based post-processing. Nevertheless, the probabilities provided by the recognition model frequently contain confused information causing challenge for post-processing. In this work, we adopt an action recognition model based on self-supervise learning to detect distracted activities and give potential action probabilities. Subsequently, a constraint ensemble strategy takes advantages of multi-camera views to provide robust predictions. Finally, we introduce a conditional post-processing operation to locate distracted behaviours and action temporal boundaries precisely. Experimenting on test set A2, our method obtains the sixth position on the public leaderboard of track 3 of the 2024 AI City Challenge.Computer Vision and Pattern Recognition Workshop 202