arXiv preprints from January 1, 2026 through September 23, 2026 — 10:01:38 EST
We investigate the simultaneous estimation of two optical phases in a three-mode interferometer assisted by optical parametric amplification (OPA). By employing the normally ordered characteristic-function formalism, we analytically obtain all photon-number moments of the output quantum state, enabling an explicit evaluation of the...
In this work, we present an analytical description of the energy distributions of primary and secondary cosmogenic particles on Earth in terms of parameters having clear physical meaning. A modified power law is assumed for energy distributions, incorporating terms such as energy loss/decay, which are effective at low energies, and a...
This paper presents a conceptual and operational framework for developing and operating safe and trustworthy AI agents based on a Three-Pillar Model grounded in transparency, accountability, and trustworthiness. Building on prior work in Human-in-the-Loop systems, reinforcement learning, and collaborative AI, the framework defines an...
Multimodal learning aims to enhance perceptual and decision-making capabilities by integrating information from diverse sources. However, classical deep learning approaches face a critical trade-off between the high accuracy of black-box feature-level fusion and the interpretability of less outstanding decision-level fusion, alongside...
Malicious image manipulation threatens public safety and requires efficient localization methods. Existing approaches depend on costly pixel-level annotations which make training expensive. Existing weakly supervised methods rely only on image-level binary labels and focus on global classification, often overlooking local edge cues...
We introduce Parallel Coordinated Reasoning (PaCoRe), a training-and-inference framework designed to overcome a central limitation of contemporary language models: their inability to scale test-time compute (TTC) far beyond sequential reasoning under a fixed context window. PaCoRe departs from the traditional sequential paradigm by...
Recently, there has been a lot of work on combinatorial inequalities related to hook-lengths in $t$-regular partitions. In this short note, we give a proof using generating functions for a result proved by Singh and Barman (2026) using combinatorial methods. In addition, we give an alternate proof of another result of Singh \& Barman...
Solar flares and coronal mass ejections (CMEs) are among the most energetic phenomena in the solar system, often impacting space weather and terrestrial technologies. In this study, we utilize SunPy, an open-source Python library for solar physics, to analyze solar active regions and their correlation with flare and CME events...
In web search and recommendation systems, user clicks are widely used to train ranking models. However, click data is heavily biased, i.e., users tend to click higher-ranked items (position bias), choose only what was shown to them (selection bias), and trust top results more (trust bias). Without explicitly modeling these biases, the...
Coronal mass ejections (CMEs) often exhibit a three-part structure consisting of a bright inner core, an outer leading edge, and an intervening dark cavity. While the core has traditionally been attributed to prominence material, an alternative interpretation suggests it may arise from the projection effects of a twisted flux rope. We...
Physical reservoir computing is a computational framework that offers an energy- and computation-efficient alternative to conventional training of neural networks. In reservoir computing, input signals are mapped into the high-dimensional dynamics of a nonlinear system, and only a simple readout layer is trained. In most physical...
We investigate the Yb valence instabilities in the strongly correlated YbPd compound using resonant X-ray emission spectroscopy as a function of pressure across the charge-order (CO) transition. At a low temperature (T = 30 K) in the CO phase, the Yb $4f$ valence remains nearly constant up to a pressure P$_L$ = 1.5 $\pm {0.2}$ GPa,...
We investigate the impact of domain-specific self-supervised pre-training on agricultural disease classification using hierarchical vision transformers. Our key finding is that SimCLR pre-training on just 3,000 unlabeled agricultural images provides a +4.57% accuracy improvement--exceeding the +3.70% gain from hierarchical...
We propose unified data structures and algorithms for free-surface fluid simulations based on partial optimal transport, such as the Power Particles method or Gallouët-Mérigot's scheme. Such methods previously relied on a discretization of the cells by leveraging a classical convex cell clipping algorithm. However, this results in a...
In active matter and living matter, such as clusters of migrating cells, collective dynamics emerges from the underlying interactions. A common assumption of theoretical descriptions of collective cell migration is that these interactions are pairwise additive. It remains unclear, however, if the dynamics of groups of cells is solely...
In this paper, based on the two-step discretization scheme proposed by Dahlquist, Liniger and Nevanlinna (DLN), we develop a semi-implicit Galerkin finite element method for solving the coupled generalized Ginzburg-Landau equations. By virtue of a novel analytical technique, the boundedness of the numerical solution in the infinity...
Noncollinear spin structures have attracted tremendous attention because they offer a versatile platform for spin control and manipulation, essential in spintronics. Realizing noncollinearity in ferrimagnetic insulators is of particular interest as they can be potentially utilized in low-damping spintronics with tunable magnetic...
The proliferation of generative AI systems creates unprecedented opportunities for content creation while raising critical concerns about controllability, copyright infringement, and content provenance. Current generative models operate as "black boxes" with limited user control and lack built-in mechanisms to protect intellectual...
We narrow the gap between the family of graphs that do and the family of graphs that do not satisfy the fat minor conjecture by obtaining much simpler counterexamples than were previously known, including $K_t, t \geq 6$ and $K_{s,t}, s,t \geq 4$ and $K_{2,2,2}$. This is achieved by establishing a `coarse self-similarity' property...
The majority of galaxies are known to have supermassive black holes (SMBHs) at their core, which have a tremendous gravitational pull on the objects around them. When embedded within extended matter distributions such as prolate, shell-like halos, they give rise to complex gravitational fields that often drive nearby particles into...
Accurate P-wave detection is critical for earthquake early warning, yet strong-motion records pose challenges due to high noise levels, limited labeled data, and complex waveform characteristics. This study reframes P-wave arrival detection as a self-supervised anomaly detection task to evaluate how architectural variations regulate...
Ice nucleation is greatly important in areas as diverse as climate change, cryobiology, geology or food industry. Predicting the ability of a substrate to induce the nucleation of ice from supercooled water is a difficult problem. Here, we use molecular simulations to analyse how the ice nucleating ability is affected by the substrate...
Methanol (CH3OH) is thought to form on interstellar ice dust via successive hydrogenation reactions. The reaction between CH3 and OH radicals could also conceivably generate methanol at temperatures above approximately 20 K, at which temperature hydrogen atoms will not adhere to the ice surface. However, this process has not been...
Incentivizing flexible consumption of end-users is key to maximizing the value of local exchanges within Renewable Energy Communities. If centralized coordination for flexible resources planning raises concerns regarding data privacy and fair benefits distribution, state-of-the-art approaches (e.g., bi-level, ADMM) often face...
LLM agents operating in open environments face escalating risks from indirect prompt injection, particularly within the tool stream where manipulated metadata and runtime feedback hijack execution flow. Existing defenses encounter a critical dilemma as advanced models prioritize injected rules due to strict alignment while static...