We establish a quantitative relation between local spin polarization and quantum entanglement in two-qubit systems by mathematically proving a closed-form upper bound on the concurrence at fixed local polarization magnitudes. The bound shows how the maximal concurrence is jointly constrained by the two local polarizations. We further demonstrate that this bound is saturated by pure states in certain cases with identical polarizations. As a concrete physical application, we consider the parity-violating process e+e−→Z0→qq¯, which generates final-state spin polarization. We show that the maximal concurrence is attained in specific kinematic regions and is significantly reduced relative to the unpolarized case. These results establish a general, process-independent framework connecting local polarization, maximal entanglement, and pure states.
Constant-rate depressurization, with control of the depressurization rate, has been proven to efficiently enhance hydrate dissociation in simulated confined natural gas hydrate (NGH) reservoirs. However, for unconfined reservoir simulation systems, the effects of depressurization rate on hydrate dissociation and gas production remain unknown. In this study, a stratified NGH reservoir environment was built to investigate CH4 hydrate dissociation in unconfined systems with varied depressurization rates. The results showed that the depressurization rate not only influenced heat and fluid transport within the original pore fluids, but also altered the flow pathways of the overlying water intruding into the reservoir. These changes affected the utilization efficiency of the sensible heat and the reformation characteristics of CH4 hydrates in the hydrate reservoir, resulting in significant spatial heterogeneity in the temperature and flow fields. Depressurization rate significantly influenced both CH4 recovery from NGH reservoirs and CH4 leakage into the overlying environment. By optimizing the depressurization rate to a moderate 0.10 MPa/min to balance factors such as overburden sediment collapse, overlying water intrusion, and hydrate dissociation, the CH4 recovery ratio increased by up to 17.17%, reaching 77.00%, while the CH4 leakage ratio decreased by as much as 11.12%, to 6.50%.
We study the size of minimum feedback vertex sets in digraphs under degree constraints. Our focus is on oriented graphs and arbitrary digraphs of bounded maximum degree. A digraph D is an oriented graph if D does not have a pair of opposite arcs. The degree of a vertex v of D is the sum of the in-degree and out-degree of v. Let fvs(D) be the minimum number of vertices whose deletion from D makes it acyclic. Let D be a digraph with n vertices and maximum degree Δ. We prove the following bounds. If D is an oriented graph, then fvs(D)≤3n7 when Δ≤4 and fvs(D)≤n2 when Δ≤5. If D is a connected digraph, Δ≤4 and D is not obtained from an odd undirected cycle by replacing every edge with a pair of opposite arcs with the same endvertices, then fvs(D)≤n2. If D is an arbitrary digraph with Δ≤5 then fvs(D)≤2n3. The above bounds are all sharp. The obtained bounds refine and extend earlier results on feedback vertex sets in sparse digraphs, and contribute to a more detailed understanding of how local degree restrictions influence global acyclicity properties.
With the rapid advancement of generative models, AI-generated images have become highly photorealistic, rendering authenticity detection challenging. Existing detection methods handle all image patches uniformly, overlooking that low-texture regions exhibit more pronounced artifact signals than complex-textured areas, where high-frequency information of complex texture inherent to the complex content interferes with artifact detection. To address this limitation, we proposed a single exposed patch (SEP) detection framework that strategically identifies and mines one patch per image to capture distinctive cues. Specifically, we developed an exposed patch selector (EPS) that precisely identifies the most discriminative patch region by analyzing pixel-level texture distributions, ensuring that the selected patch is minimally affected by high-frequency noise in the image. To better exploit artifact cues from the selected patch and mitigate the interference from irrelevant high-frequency information, we proposed the spatial artifact perception module (SAPM). The module preserves low-level features and spatial structure and introduces a guided bidirectional cross-spatial attention (GBCSA) mechanism guided by high-frequency cues, reducing reliance on high-frequency inputs and enhancing sensitivity to minor artifacts in flat regions. Building on this, we designed a frequency perception module (FPM) to detect the periodic artifact patterns generated in the frequency domain by generative models (e.g., GANs), thereby enhancing the ability of the model to detect artifacts across different generative mechanisms. Experimental results demonstrate the superior cross-generator generalization capabilities of SEP. We conducted comparisons on two major public datasets (GenImage and AIGCBenchmark). The million-scale GenImage dataset comprises eight different generative models, and the AIGCBenchmark encompasses a broader range of 16 diverse generative models. SEP achieved 97.05% accuracy on GenImage, outperforming recent state-of-the-art methods published in 2025. On AIGCBenchmark, SEP also delivered the best overall performance under the current evaluation protocol. The code is available at https://github.com/1112348/SEP-for-AIGCDetection
Clean and efficient utilization of coal is a critical research focus in energy transition. However, agglomeration issues were encountered during the utilization of a pulverized coal under fluidized reductive conditions. In this study, the primary organic structure of this coal was elucidated, and the mechanism of fluidized reductive agglomeration was revealed. The results indicated that agglomeration was predominantly governed by organic components and exhibited strong temperature dependence, with an onset temperature of 700 degrees C. This coal was characterized by an abundance of phenolic hydroxyls, substituted aromatic rings, naphthalene-like precursors, and polycyclic aromatic hydrocarbon skeletons connected by methylene bridges. During pyrolysis, these structures were transformed into organics featuring naphthalene skeletons and abundant aliphatic side chains, which functioned as critical binding media to drive agglomeration. With temperature increased, these substances underwent decomposition and restructuring. A portion of products was converted into acenaphthylene via dehydrogenation and aromatization, enriching high-temperature tar, where the maximum relative content of acenaphthylene reached 9.51%. Conversely, the residual aliphatic structures generated from decomposition were solidified within agglomerates, resulting in an increase in the ratio of aliphatic to aromatic hydrogen to 2.95. This study provides a reference for investigations on fluidized reductive agglomeration of pulverized coal and development of regulation technologies.
