arXiv subjects

Zhenhua Zhao

Publications and source records attributed to Zhenhua Zhao.

7 recordsLinked to original sources

arXiv2026

Millikelvin magnetometry up to 30~T using a multi-cross micro-Hall sensor

We report a multi-cross micro-Hall sensor fabricated from an MBE-grown GaAs/AlGaAs heterostructure for high-sensitivity magnetometry. The device enables dc magnetization measurements at temperatures down to $\sim$15~mK and in dc magnetic fields up to 30 T. Bulk samples with dimensions ranging from 100~$μ$m to 1~mm can be measured with a resolution better than 1$~μ$T with a 1~$μ$A excitation at 13.77 Hz. Test-of-performance measurements were carried out by probing the field-dependent magnetization plateaus of the triangular-lattice quantum magnet K$_{2}$Co$_{2}$(SeO$_{3}$)$_{3}$ at 20~mK, the antiferromagnetic transition of K${_3}$GdTe${_2}$O${_9}$ near 220 mK via temperature-dependent magnetization, and the quantum oscillations in a ruthenium single crystal. A simple and reliable sample-mounting setup that ensures reproducible magnetic measurements is provided.

cond-mat.mtrl-sci↗

arXiv2026

Deep Stochastic Residual Methods for Fully Nonlinear Parabolic Equations

We develop and analyze neural stochastic residual methods for fully nonlinear parabolic terminal problems under exact diffusion sampling. A quadratic stochastic residual supports global training and backward regression with scalar networks and losses involving only spatial derivatives. For global training, the correction removes a Hessian penalty from the limiting first-order loss. Under a uniform output regularity budget, we bound the continuous residual norm by the square roots of the global loss or normalized backward-regression excess loss, an $O(h^{1/2})$ discretization term for time step $h$, and terminal error. Residual stability then gives value, gradient and Hessian error bounds. Under suitable solution and terminal assumptions, we construct admissible networks with $O(h^{1/2})$ bounds on these approximation errors and the square roots of both loss criteria. In ten-dimensional backward-regression tests using a shared schedule tuned for the quadratic residual, mean relative Hessian errors are $5$--$6\%$ with the quadratic residual and about $15\%$ with the first-order residual. Tuned backward regression achieves Hessian accuracy comparable to physics-informed neural networks across the tested nonlinear coefficients. Translation tests show that the ratio of Hessian error to residual can grow sharply under a fixed sampling law.

math.NA↗

arXiv2026

Optimal Weighted $L^2$ Hessian Estimates for Parabolic Equations under General Diffusion Marginals

We study weighted $L^2$ Hessian estimates for parabolic terminal equations under the marginals of a prescribed diffusion, allowing singular initial laws and different principal coefficients in the equation and the diffusion. Under coefficient regularity, uniform ellipticity and bounded reference drift, we characterize the limiting optimal constant uniformly over shrinking subintervals with zero terminal data. Its finiteness is equivalent to linear well-posedness in the weighted parabolic Sobolev space on the full interval. For initial laws satisfying weighted doubling and coercivity conditions, this constant is the maximum of contributions from heat evolutions of rescaled initial measures, constant weights and exponential mixtures. The phase contributions are bounded above in terms of the coefficients at time zero outside the initial support and at spatial infinity. Under radial reference structure, we obtain exact coefficient formulas invariant under bounded changes of reference drift. In the matching case, with reference covariance $a$, principal matrix $a/2$ and normalized Hessian $a^{1/2}D^2u\,a^{1/2}$, the constant is $2$ for every initial law with time weight $t^α$, $α\ge1/2$; point starts attain $2\sqrt2$ at $α=0$. Whenever the limiting constant is finite, increasing damping makes the full-interval Hessian constant converge to it and the value and gradient gains tend to zero. As applications, we obtain nonlinear solvability and residual stability for small Hessian perturbations. For solutions and candidates with a common regularity bound, secant estimates give linear stability under a relative spectral gap, while covariance interpolation extends residual stability beyond it.

math.PR↗

arXiv2025

FPT-Noise: Dynamic Scene-Aware Counterattack for Test-Time Adversarial Defense in Vision-Language Models

Vision-Language Models (VLMs), such as CLIP, have demonstrated remarkable zero-shot generalizability across diverse downstream tasks. However, recent studies have revealed that VLMs, including CLIP, are highly vulnerable to adversarial attacks, particularly on their visual modality. Traditional methods for improving adversarial robustness, such as adversarial training, involve extensive retraining and can be computationally expensive. In this paper, we propose a new Test-Time defense: Feature Perception Threshold Counterattack Noise (FPT-Noise), which enhances the adversarial robustness of CLIP without costly fine-tuning. Our core contributions are threefold: First, we introduce a Dynamic Feature Modulator that dynamically generate an image-specific and attack-adaptive noise intensity parameter. Second, We reanalyzed the image features of CLIP. When images are exposed to different levels of noise, clean images and adversarial images exhibit distinct rates of feature change. We established a feature perception threshold to distinguish clean images from attacked ones. Finally, we integrate a Scene-Aware Regulation guided by a stability threshold and leverage Test-Time Transformation Ensembling (TTE) to further mitigate the impact of residual noise and enhance robustness.Extensive experimentation has demonstrated that FPT-Noise significantly outperforms existing Test-Time defense methods, boosting average robust accuracy from 0.07% to 56.86% under AutoAttack while maintaining high performance on clean images (-1.1%). The code will be made public following the publication of the study. The code will be made public following the publication of the study.

cs.CR↗

arXiv2025

Effect of Attachment Surface on Biofilm Response and the Dissipation of PAHs and nitrogen in Submerged Plant System under the application of Bactericide and Algaecide

The biofilms response on active attachment surfaces (submerged plant leaves) and inert attachment surfaces (biomimetic plant glass attachment surfaces) and their effects on PAHs and nitrogen transformation under the application conditions of bactericide, algaecide and PAHs were investigated by lab simulated hydroponics and high-throughput molecular biology methods in Vallisneria natans (VN), Hydrilla verticillata (HV) and biomimetic plants (BP) systems. Results showed that the introduction of bactericide, algaecide and PAHs changed the microorganism composition in biofilm, the presence or absence of submerged plants was the primary factor causing differences in microbial communities. Moreover, the microbial diversity and endemic species of biomimetic plant biofilms were higher than those of submerged plant systems, indicating that submerged plants have selectively induced the reconstruction of biofilm-leaves. The introduction of bactericide, algaecide and PAHs leads to abnormal accumulation of TP and NH3-N in overlying water, as well as NO3-N and TP content in sediment. However, submerged plants can weaken and alleviate the stress effects of these factors on nitrogen and phosphorus conversion. Compared to the inert biomimetic plant glass surface, the presence of active surfaces in submerged plants results in a much higher abundance of PAHs degrading bacteria (Sphingomonas and Novosphingobium) and nitrogen converting bacteria (e.g. denitrifying bacteria of Methylophilus, Methylotenera, Flavobacterium, Hydrogenophaga, Aquabacterium, and nitrogen-fixing bacteria of Rhizobium, Azoarcus, Rhizobacter, Azonexus) in biofilm-leaves of submerged plants after the addition of bactericide, algaecide and PAHs. Coupled degradation of PAHs and nitrogen occurs, which is beneficial for the treatment of combined pollution caused by PAHs and nitrogen.

q-bio.CB↗

arXiv2025

Face processing emerges from object-trained convolutional neural networks

Whether face processing depends on unique, domain-specific neurocognitive mechanisms or domain-general object recognition mechanisms has long been debated. Directly testing these competing hypotheses in humans has proven challenging due to extensive exposure to both faces and objects. Here, we systematically test these hypotheses by capitalizing on recent progress in convolutional neural networks (CNNs) that can be trained without face exposure (i.e., pre-trained weights). Domain-general mechanism accounts posit that face processing can emerge from a neural network without specialized pre-training on faces. Consequently, we trained CNNs solely on objects and tested their ability to recognize and represent faces as well as objects that look like faces (face pareidolia stimuli).... Due to the character limits, for more details see in attached pdf

cs.CV↗

arXiv2005

Topological quantization of self-dual Chern-Simons vortices on Riemann Surfaces

The self-duality equations of Chern-Simons Higgs theory in a background curved spacetime are studied by making use of the U(1) gauge potential decomposition theory and $ϕ$-mapping method. The special form of the gauge potential decomposition is obtained directly from the first of the self-duality equations. Using this decomposition, a rigorous proof of magnetic flux quantization in background curved spacetime is given and the unit magnetic flux in curved spacetime is also found . Furthermore, the precise self-dual vortex equation with topological term is obtained, in which the topological term has always been ignored.

hep-th↗