共找到 20 条结果
Does gravity care about electric charge? Precision tests of the weak equivalence principle achieve remarkable sensitivity but deliberately minimize electric charge on test masses, leaving this fundamental question experimentally open. We present a minimalist framework coupling electromagnetism to linearized gravity through conservation of a complex charge-mass current, predicting charge-dependent violations $Δa/g = κ(q/m)$. Remarkably, this prediction occupies unexplored experimental territory precisely because precision gravity tests avoid charge variation. We identify this as a significant gap and propose a modified torsion balance experiment where $q/m$ is treated as a controlled variable. Such an experiment could test whether gravitational acceleration depends on electric charge, probing physics in genuinely new parameter space. This work exemplifies how theoretical minimalism can reveal overlooked opportunities in fundamental physics.
Recent audio-video generative systems suggest that coupling modalities benefits not only audio-video synchrony but also the video modality itself. We pose a fundamental question: Does audio-video joint denoising training improve video generation, even when we only care about video quality? To study this, we introduce a parameter-efficient Audio-Video Full DiT (AVFullDiT) architecture that leverages pre-trained text-to-video (T2V) and text-to-audio (T2A) modules for joint denoising. We train (i) a T2AV model with AVFullDiT and (ii) a T2V-only counterpart under identical settings. Our results provide the first systematic evidence that audio-video joint denoising can deliver more than synchrony. We observe consistent improvements on challenging subsets featuring large and object contact motions. We hypothesize that predicting audio acts as a privileged signal, encouraging the model to internalize causal relationships between visual events and their acoustic consequences (e.g., collision $\times$ impact sound), which in turn regularizes video dynamics. Our findings suggest that cross-modal co-training is a promising approach to developing stronger, more physically grounded world models
The recent synthesis of Goldene, a 2D atomic monolayer of gold, has opened new avenues in exploring novel materials. However, the question of when multilayer Goldene transitions into bulk gold remains unresolved. This study used density functional theory calculations to address this fundamental question. Our findings reveal that multilayer Goldene retains an AA-like stacking configuration of up to six layers, with no observation of Bernal-like stacking as seen in graphene. Goldene spontaneously transitions to a bulk-like gold structure at seven layers, adopting a rhombohedral (ABC-like) stacking characteristic of bulk face-centered cubic (FCC) gold. The atomic arrangement converges entirely to the bulk gold lattice for more than ten layers. Quantum confinement significantly impacts the electronic properties, with monolayer and bulk Goldene exhibiting a single Dirac cone at the X-point of the Brillouin zone. In contrast, multilayer Goldene shows two Dirac cones at the X- and Y-points. Additionally, monolayer Goldene exhibits anisotropic optical absorption, which is absent in bulk gold. This study provides a deeper understanding of multilayer Goldene's structural and electronic prope
Goal recognition is an important problem in many application domains (e.g., pervasive computing, intrusion detection, computer games, etc.). In many application scenarios, it is important that goal recognition algorithms can recognize goals of an observed agent as fast as possible. However, many early approaches in the area of Plan Recognition As Planning, require quite large amounts of computation time to calculate a solution. Mainly to address this issue, recently, Pereira et al. developed an approach that is based on planning landmarks and is much more computationally efficient than previous approaches. However, the approach, as proposed by Pereira et al., also uses trivial landmarks (i.e., facts that are part of the initial state and goal description are landmarks by definition). In this paper, we show that it does not provide any benefit to use landmarks that are part of the initial state in a planning landmark based goal recognition approach. The empirical results show that omitting initial state landmarks for goal recognition improves goal recognition performance.
We work on the family of topologies for the Minkowski manifold M. We partially order this family by inclusion to form the lattice Σ(M), and focus on the sublattice Z of topologies that induce the Euclidean metric space on every time axis and every space axis. We analyze the bounds of Z in the lattice Σ(M), in search for its supremum. Our conclusion --that such a supremum does not belong in Z-- is compared with constructive proofs of existence of the fine topology, defined as the maximum of Z and conceived to play an essential role in contemporary physical theories. Essential mathematical and physical questions arise.
The extensively cited work of Barzilai, J. (1997): Deriving weights from pairwise comparison matrices, published in Journal of the Operational Research Society, 48(12), 1226-1232., derives the geometric mean method from two simple axioms. This note reveals that the central result of the paper called the Fundamental Theorem by the author does not hold as there exists at least one further method satisfying both requirements.
We show that Kelley-Morse set theory does not prove the class Fodor principle, the assertion that every regressive class function $F:S\to\text{Ord}$ defined on a stationary class $S$ is constant on a stationary subclass. Indeed, it is relatively consistent with KM for any infinite $λ$ with $ω\leqλ\leq\text{Ord}$ that there is a class function $F:\text{Ord}\toλ$ that is not constant on any stationary class. Strikingly, it is consistent with KM that there is a class $A\subseteqω\times\text{Ord}$, such that each section $A_n=\{α\mid (n,α)\in A\}$ contains a class club, but $\bigcap_n A_n$ is empty. Consequently, it is relatively consistent with KM that the class club filter is not $σ$-closed.
The brain’s memory center may begin life more like a crowded web than an empty canvas。 Researchers discovered that early neural networks in the hippocampus are dense and seemingly random, then become more organized by shedding connections over time。 This pruning process creates a faster, more efficient system for linking experiences and forming mem
In this work, we study the thermodynamic functions of quantum gases confined to spaces of various shapes, namely, a sphere, a cylinder, and an ellipsoid. We start with the simplest situation, namely, a spinless gas treated within the canonical ensemble framework. As a next step, we consider \textit{noninteracting} gases (fermions and bosons) with the usage of the grand canonical ensemble description. For this case, the calculations are performed numerically. We also observe that our results may possibly be applied to \textit{Bose-Einstein condensate} and to \textit{helium dimer}. Moreover, the bosonic sector, independently of the geometry, acquires entropy and internal energy greater than for the fermionic case. Finally, we also devise a model allowing us to perform analytically the calculations in the case of \textit{interacting} quantum gases, and, afterwards, we apply it to a cubical box.
Retrieval-Augmented Generation (RAG) systems depend critically on the quality of document preprocessing, yet no prior study has evaluated PDF processing frameworks by their impact on downstream question-answering accuracy. We address this gap through a systematic comparison of four open-source PDF-to-Markdown conversion frameworks, Docling, MinerU, Marker, and DeepSeek OCR, across 19 pipeline configurations for extracting text and other contents from PDFs, varying the conversion tool, cleaning transformations, splitting strategy, and metadata enrichment. Evaluation was performed using a manually curated 50-question benchmark over a corpus of 36 Portuguese administrative documents (1,706 pages, ~492K words), with LLM-as-judge scoring averaged over 10 runs. Two baselines bounded the results: naïve PDFLoader (86.9%) and manually curated Markdown (97.1%). Docling with hierarchical splitting and image descriptions achieved the highest automated accuracy (94.1%). Metadata enrichment and hierarchy-aware chunking contributed more to accuracy than the conversion framework choice alone. Font-based hierarchy rebuilding consistently outperformed LLM-based approaches. An exploratory GraphRAG im
In this article, we give a new proof of a result due to J. Kim, which states that the Ricci tensor of a gradient Ricci soliton with dimension $n \geq 4$ and harmonic Weyl tensor has at most three distinct eigenvalues. This result constitutes an essential step in the classification of such manifolds, originally established by J. Kim in dimension $4$ and subsequently extended to dimensions $n\geq5$. Our proof offers two notable advantages: it is shorter and does not require the use of any specialized moving frame.
By introducing arbitrary parameters in the usual Fronsdal model, we find a region in the parameters space away from the ``Fronsdal point'' where we still have an irreducible description of massless particles of arbitrary integer spin $s\ge 3$. The higher spin gauge symmetry is further constrained by a vanishing double divergence condition on the traceless gauge parameter: $\partial\cdot\partial\cdot\barΛ=0$. Remarkably, it does not introduce extra propagating gauge invariants. We demonstrate that we only have spin-$s$ helicity states as propagating modes for arbitrary integer $s\ge 3$. For the simplest $s=3$ case we present a gauge invariant proof while for $s\ge 4$ we use a light-cone gauge. The reduction in the gauge symmetry allows for more general source couplings when compared to the Fronsdal model.
Given a projective variety $X$ over an algebraically closed field $k$, M. V. Nori introduced in 1976 a group scheme $π(X)$ which accounts for principal bundles $P\to X$ with finite structure, obtaining in this way an amplification the etale fundamental group. One drawback of this theory is that it is quite difficult to arrive at an explicit description of $π(X)$, whenever it does not vanish altogether. To wit, there are no known non-trivial examples in the literature where $π(X)$ is local, or local of some given height, etc. In this paper we obtain a description of $π(X)$ through amalgamated products of certain non-commutative local group schemes - we called them infinitesimal non-commutative Witt group schemes - in the case where $X$ is a non-normal variety obtained by pinching a simply connected one.
In-context learning (ICL) enables Large Language Models (LLMs) to learn tasks from demonstration examples without parameter updates. Although it has been extensively studied in LLMs, its effectiveness in Vision-Language Models (VLMs) remains underexplored. In this work, we present a systematic study of ICL in VLMs, evaluating seven models spanning four architectures on three image captioning benchmarks. We analyze how prompt design, architectural choices, and training strategies influence multimodal ICL. To our knowledge, we are the first to analyze how attention patterns in VLMs vary with an increasing number of in-context demonstrations. Our results reveal that training on imag-text interleaved data enhances ICL performance but does not imply effective integration of visual and textual information from demonstration examples. In contrast, instruction tuning improves instruction-following but can reduce reliance on in-context demonstrations, suggesting a trade-off between instruction alignment and in-context adaptation. Attention analyses further show that current VLMs primarily focus on textual cues and fail to leverage visual information, suggesting a limited capacity for multim
We obtain improved regularity results for solutions to a nonlocal dead-core problem at branching points. Our approach, which does not rely on the maximum principle, introduces a new strategy for analyzing two-phase problems within the local framework, an area that remains largely unexplored.
We study the spectral properties and local topology of the Kane-Mele-Rashba model on a Sierpinski Carpet (SC) fractal, constructed from a rectangular flake with an underlying honeycomb arrangement and open boundary conditions. When the system parameters correspond to a topologically trivial phase, the energy spectrum is characterized solely by bulk states that are not significantly modified by the system's fractality. For parameters corresponding to the quantum spin Hall insulator (QSHI) phase, in addition to bulk states, the energy spectrum exhibits in-gap topological states that are strongly influenced by the fractal geometry. As the fractal generation increases, the in-gap topological states acquire a staircase profile, which translates into sharp peaks in the density of states. We also show that both the QSHI and the trivial phase exhibit a large gap in the valence-projected spin spectrum, allowing the use of the local spin Chern marker (LSCM) to index the local topology of the system. Fractality does not affect this gap, allowing the application of LSCM to higher fractal generations. Our results explore the LSCM versatility, showing its potential to access local topology in co
This article presents an innovative extension of the Smagorinsky model incorporating dynamic boundary conditions and advanced regularity methods. We formulate the modified Navier-Stokes equations with the Smagorinsky term to model dissipation in turbulence and prove theorems concerning the existence, uniqueness, and asymptotic behavior of solutions. The first theorem establishes the existence and uniqueness of solutions in higher Sobolev spaces, considering the effect of the nonlinear Smagorinsky term and dynamic boundary conditions. The proof employs the Galerkin method and energy estimates, culminating in the application of Grönwall's theorem. The second theorem investigates the asymptotic behavior of solutions, focusing on anomalous dissipation in high-turbulence regimes. We demonstrate that the dissipated energy does not decrease with vanishing viscosity, indicating the occurrence of anomalous dissipation. The third theorem explores advanced regularity in higher Sobolev spaces, allowing for more rigorous control of nonlinear terms and ensuring improved stability conditions. The proof utilizes the energy method combined with Sobolev estimates and Grönwall's inequality. These mat
Eclipse mapping uses the shape of the eclipse of an exoplanet to measure its two-dimensional structure. Light curves are mostly composed of longitudinal information, with the latitudinal information only contained in the brief ingress and egress of the eclipse. This imbalance can lead to a spuriously confident map, where the longitudinal structure is constrained by out-of-eclipse data and the latitudinal structure is wrongly determined by the priors on the map. We present a new method to address this issue. The method tests for the presence of an eclipse mapping signal by using k-fold cross-validation to compare the performance of a simple mapping model to the null hypothesis of a uniform disk. If a signal is found, the method fits a map with more degrees of freedom, optimising its information content. The information content is varied by penalising the model likelihood by a factor proportional to the spatial entropy of the map, optimised by cross-validation. We demonstrate this method for simulated datasets then apply it to three observational datasets. The method identifies an eclipse mapping signal for JWST MIRI/LRS observations of WASP-43b but does not identify a signal for JWS
After the COVID-19 pandemic, we saw an increase in demand for epidemiological mathematical models. The goal of this work is to study the optimal control for an age-structured model as a strategy of quarantine of infected people, which is done via Pontryagin's maximum principle. Since quarantine campaigns are not just a matter of public health, also posing economic challenges, the optimal control problem does not simply minimize the number of infected individuals. Instead, it jointly minimizes this number and the economic costs associated to the campaigns, providing data that can help authorities make decisions when dealing with epidemics. The controls are the quarantine entrance parameters, which are numerically calculated for different lengths of isolation. The best strategies gives a calendar that indicates when the isolation measures can be relaxed, and the consequences of a delay in the start of the quarantine are analyzed by presenting the reduction in the number of deaths for the strategy with optimal control compared to a no-quarantine landscape.
The robust valorization of carbon dioxide (CO2) stays at the center of sustainable development. Since CO2 represents a low-energy compound, its transformation into commercially coveted products is cumbersome. In the present work, we report a revolutionary method to obtain dimethyl carbonate (DMC) out of methanol (CH3OH) and CO2 catalyzed by sodium chloride (NaCl) and similar inorganic salts. The computational exploration revealed a mechanism of favorable catalysis, which was subsequently confirmed experimentally. Unlike all competitive syntheses of DMC, the new one does not produce water and, therefore, the hydrolysis of a carbonate does not occur. No dehydrating agents are necessary. The employed catalyst is cheap and permanently exists in the same phase with the reactants and products. The action of NaCl was compared to those of other alkali metal salts, LiI, LiCl, and KI, and competitive performances were recorded. The experimentally obtained result outperforms most competing technologies according to the DMC yield, 19% with molecular sieves and 17% without molecular sieves. All existing competitors are excelled by the simplicity and cleanness of the synthesis. The reported adva