Accelerator Physics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 8 September
We address the challenge of scalable uncertainty quantification in large-scale scientific applications, where complex state-of-the-art machine learning methods are often computationally infeasible. Our primary contribution is a simple yet effective empirical Bayes method for automatically tuning the hyperparameters of a flexible, heteroscedastic Spectral-normalized Neural Gaussian Process. This approach retains the expressiveness and uncertainty-awareness of semi-Bayesian neural models while significantly reducing the computational burden by integrating hyperparameter learning directly into the training loop. We demonstrate the practical impact of our method on the task of estimating the dynamic aperture in circular particle accelerators, a fundamental problem in high-energy physics colliders and storage rings, using simulation data from the case of the Large Hadron Collider at CERN. Traditional approaches to DA estimation require extensive particle-tracking simulations, which are prohibitively time-consuming and resource-intensive. Our results show that the proposed method achieves competitive predictive performance and well-calibrated uncertainty estimates at much lower computational cost than state-of-the-art approaches. We stress that, beyond this application, the proposed empirical Bayes framework offers a general solution for training heteroscedastic neural models in situations where manual hyperparameter tuning is impractical. Accordingly, we anticipate that this framework can be applied to other domains that encounter comparable computational limitations.
Dynamical Systems
Phase oscillator networks with multiple and state-dependent delays: A framework for exploring white matter plasticity in neurodynamics
Network science is increasingly focused on how node dynamics influence emergent phenomena such as oscillations, waves, chimeras, and turbulence. In oscillatory systems modeled by networks of coupled ordinary differential equations (ODEs), a common approach is to reduce the system to phase variables. When understanding how network delays shape emergent properties, the delays are often absorbed in the reduced description as a phase shift. The reduced system is a set of ODEs that loses some information about the full delay differential equation (DDE) system. We adopt a less restrictive approach and consider limit cycle oscillator networks with multiple delays that can be reduced to a DDE system. This captures the effects of delayed couplings, including coexistence and multistability of states. We analyze relative equilibria in the form of phase-locked states with tools previously developed for more general DDE settings. This allows us to explore patterning in networks with space-dependent delays, including in neuroscience, using symmetric bifurcation theory, linear stability analysis, and numerical simulations and continuation. In brain dynamics, time delays are determined by the speed of a communicating signal (action potential) along a fiber (axon). Importantly, these are now known to be state-dependent since the myelin (white matter) that insulates axons is plastic and can change in response to neuronal activity. A simple phenomenological model of this process (in the phase reduced description) is introduced and analyzed by extending techniques developed for fixed delays. Our analysis suggests that white matter plasticity can drive networks to more coherent behavior.
Read the full abstract
Machine Learning
Target-Independent Micro-Interventions for Predicting Training Response Across Language-Model Families
Benchmark scores describe what a checkpoint can do now, but they do not determine how it will respond to the next training episode. We measure this missing state by branching four short, standardized, target-independent micro-interventions from the same checkpoint and recording their effects in a common capability space. Together with current capability, these responses form L-State; its pulse block supports a flexible direct readout and a structure-preserving operator readout. Under smooth local dynamics, the operator construction admits an end-to-end cross-family bound with explicit source- and target-family coordinate heterogeneity. In three-family leave-one-family-out development, both pulse readouts reduce source-standardized MSE by 39.4% relative to capability alone, while separating the best response and direction estimates. On sealed GLM-4-9B, the direct and operator readouts reduce MSE by 71.8% and 78.3%, respectively, and the operator readout raises sign balanced accuracy from 0.366 to 0.754. On sealed Granite-3.1-8B, the direct readout reaches RMSE 0.544 and a development-fitted action-wise selector reaches 0.554, compared with 1.172 for capability alone. A five-family audit finds that the operator coordinate varies by action and family, and that modeling these deviations improves retrospective held-trajectory prediction. Target-independent interventions therefore expose training-response information that current capability misses, with direct and structured readouts covering complementary transfer regimes.
Read the full abstract
Symplectic Geometry
Shifted Contact Structures on Exact Symplectic Fibrations
Within the framework of classical contact geometry, the first author introduced the notion of contact symplectic structure –the data of an exact symplectic fibration with a contact base– and proved, under mild conditions, that the total space of such a fibration inherits a contact structure compatible with the fibration map, thereby establishing the contact Thurston theorem. This paper provides a derived contact version of that result by incorporating our prior work on the derived symplectic Thurston theorem. In this paper, we prove, under certain conditions, that if a morphism $\pi: X \rightarrow S$ of derived stacks has a shifted exact symplectic fibration structure and the target stack $S$ admits a shifted contact structure, then one can construct a shifted contact structure on the source stack $X$, compatible with $\pi$ in a sense similar to the smooth case. Our framework relies on the theory of relative shifted structures; hence our result, called the derived contact Thurston theorem, in fact establishes a relative-to-absolute type construction. As an application, we present examples of our relative-to-absolute construction formalism in the derived contact setting, including conormal stacks, quotient mapping stacks, and affine exact symplectic fibrations.
Read the full abstract
Representation Theory
Essential algebra of the shifted Burnside biset functor with abelian shift
Let $T$ be a finite group and let $kB_T$ denote the shifted Burnside biset functor over a commutative ring $k$. We study the essential algebra $\widehat{kB_T}(G)$ of $kB_T$ at an arbitrary finite group $G$. We first establish criteria determining when a covering subgroup of $G\times G\times T$ factors, with respect to the star product, through a group of order strictly smaller than $|G|$; the first criterion places no hypothesis on $T$. As an application, we describe $\widehat{kB_T}(G)$ completely whenever no nontrivial quotient of $G$ is isomorphic to a section of $T$, extending the coprime case treated by Romero. When $T$ is abelian and $|T|$ is invertible in $k$, we prove that $\widehat{kB_T}(G)$ decomposes as a direct sum of matrix algebras over the group algebras of the groups $Out_T(A)$, where $A$ runs over the linkage classes of reduced subgroups of $G\times T$, and we parametrize the simple modules of $\widehat{kB_T}(G)$.
Read the full abstract
Machine Learning
Why shared attention vectors fail: a case for outcome-indexed tuning
Dimensional attention in learning is often implemented as a globally shared attention vector, where each stimulus dimension corresponds to a single scalar. These scalars are learned by models through gradient-descent on error, where predictive features acquire more salience. We show that under multi-outcome learning, where models predict more than one outcome, this shared vector becomes unstable; it collapses to its bounds and prevents the models from learning meaningful attentional tunings for learning and generalization. We address this by introducing an outcome-indexed attentional matrix that converts globally shared attentional tuning into an outcome-indexed representation. We present an analysis of the unstable shared vectors and derive the conditions under which it holds. Empirically, three synthetic experiments benchmark the proposed attention matrices and show that they converge to meaningful representations, something shared attention vectors fail to do. These results suggest that outcome-indexed attentional matrices are a general fix for gradient-based attentional processes, which improves models of learning under multi-outcome conditions.
Read the full abstract
Optimization and Control
McCormick relaxations for PDE-constrained optimization on multi-dimensional domains
We are interested in computing approximate dual bounds for nonconvex optimal control problems using McCormick inequalities. To handle the numerical burden concerned with the tightening of these inequalities, we locally average the bilinear term in the PDE constraint. We extend recent theory in this area to elliptic PDEs on multidimensional domains that involve bilinear terms. As an additional result, we also generalize the necessary assumptions to allow for more flexibility in the choice of the discretization for the total variation penalty term from the underlying control problem. Furthermore, we prove the existence of valid bounds on the state variable, which are crucial to enable the optimization-based bound-tightening (OBBT) procedure to tighten the McCormick inequalities. Since theoretical bounds on the state variable are rather conservative, OBBT is both crucial and the main computational bottleneck for deriving a sharp approximate dual bound. We show how the sequence of LPs arising during OBBT may sensibly be warmstarted, allowing us to successfully scale the numerical experiments to finer discretizations than what was possible before.
Read the full abstract
Combinatorics
A Chvátal-type codegree condition for Hamiltonian cycles in $k$-uniform hypergraphs
We prove an asymptotic Chvátal-type codegree criterion for tight Hamiltonian cycles in $k$-uniform hypergraphs for every fixed $k\ge3$. The criterion allows small degrees to be compensated by large degrees in the link graphs of $(k-2)$-tuples. It strengthens the asymptotic Dirac-type theorem by Rödl, Ruciński and Szemerédi and implies a Pósa-type criterion conjectured by Schülke.
Read the full abstract
Metric Geometry
Around higher-order Godbersen conjectures
We consider two higher-order generalizations of the Godbersen conjecture for mixed volumes of convex bodies, which were first proposed by Schneider in 2000. First, we establish the conjectures in several special cases, in particular in low dimensions. Second, we prove that certain consequences of the conjectures hold. More precisely, we define a higher-order version of the unbalanced difference body and prove related weighted inequalities, generalizing previous results of Artstein-Avidan and Putterman. Finally, we introduce higher-order analogs of unbalanced joins of convex bodies considered by Artstein-Avidan, Einhorn, Florentin, and Ostrover, prove the corresponding volume bounds, and propose conjectures that interpolate between the higher-order Godbersen conjectures and a conjecture due to Fáry and Rédei.
Read the full abstract
Information Theory
Sequential Lossy Compression With Causal Conditional Perception
In this paper, we study sequential lossy compression under a causal conditional perception criterion comparing source and reconstruction distributions given the same reconstruction history. For first-order Markov sources, we formulate the finite-horizon nonanticipative rate-distortion-perception function (NRDPF) with stagewise constraints and establish one-shot lower and upper bounds on the minimum variable-length sum rate using a strengthened strong functional-representation lemma (SFRL) and common randomness. For time-varying scalar Gauss–Markov sources under pointwise mean-squared error (MSE) and conditional squared Wasserstein-$2$ fidelity, we prove Gaussian optimality, derive a log-variance characterization, and obtain a closed-form solution that recovers the classical Gaussian nonanticipative rate-distortion function (NRDF) when perception is unconstrained and the classical Gaussian RDPF when the source is stationary and memoryless.
Read the full abstract
Statistical Methodology
Cursive: The Trace from the Curse of Dimensionality
Modern data are increasingly high-dimensional or non-Euclidean. As dimension grows, new statistical patterns can emerge in the relations among observations, while a conventional statistical summary may fail to retain the signal they carry. This paper names and organizes a research program around this observation, calling it Cursive. Cursive asks which relational information is lost when the summary is formed and how the analysis should be redesigned around that information. The canonical example is the generalized edge-count test, which keeps the two within-sample edge counts whose opposing deviations can cancel in the classical between-sample count. This paper traces how the same design question has led to task-specific methods built from graphs, graph-based ranks, kernels, and dissimilarity profiles for testing, change-point detection, covariate-balance assessment, clustering, classification, and generative-model evaluation.
Read the full abstract
Computation and Language
Dynamics of Meaning: Towards the Evaluation of Diachronic Semantic Change in Sinhala
Tracking semantic change in low-resource languages across extensive historical timelines presents significant challenges due to data scarcity and the limitations of static embedding alignments. This study investigates the diachronic evolution of the Sinhala language from the 13th to the 20th century using a multi-stage computational framework. We first align century-specific Word2Vec and FastText embeddings using Similarity Matrix Based Alignment (SMA) and Orthogonal Procrustes (OP) techniques, finding that OP alignment provides more stable neighbourhood tracking for identifying temporal similarity dips. To move beyond aggregate measures, we introduce a Bidirectional Semantic Impact Pruning approach using contextualised embeddings from a fine-tuned Llama-3.1-8B. By applying Leave-One-Out (LOO) diagnostics, we attempt to isolate influential sentences to distinguish between systemic semantic shifts and transient polysemic expansion. Our results show that semantic drift in the fine-tuned Llama-3.1-8B is not evenly distributed across all usages. Instead, a significant part of the change is driven by a smaller set of high-impact contextual instances, rather than gradual and uniform change across all occurrences. This work provides a preliminary framework for low-resource Sinhala diachronic analysis, highlighting the trade-offs between model sensitivity and data availability.
Read the full abstract
Number Theory
Reduction Modulo Binary Polynomials with Logarithmic Feedback Depth
Polynomial modular reduction is central to binary finite-field arithmetic and repeated Frobenius powering. Top-down shift/XOR folding can use few operations when the nonleading support is sparse, but a tap near the leading term creates a long feedback chain. We formulate this recurrence as inversion of a nilpotent shift operator and factor its inverse by characteristic-two Frobenius powers. The resulting Frobenius-factorized reduction (FFR) applies to every monic binary modulus without materializing a reciprocal or dense reduction matrix, and its shifts can be generated online without a persistent modulus-specific schedule. For degree $m$ and nonempty nonleading support of size $s$, with nearest-tap distance $\Delta_{\min}$, FFR uses exactly $\lceil\log_2(m/\Delta_{\min})\rceil$ sequential feedback stages and has scheduled work $O(ms(1+\log(m/s)))$. A portable-C evaluation on 1,096 supports through degree $131072$ identifies distinct FFR, López–Dahab, and gf2x-backed Barrett regions. On four certified irreducible moduli, FFR makes complete Rabin irreducibility testing $1.35$–$8.04$ times faster than NTL and $1.60$–$6.81$ times faster than the matched Barrett implementation.
Read the full abstract
Computer Vision
GOLF: Global Observation with Local Focus for Calibration-Aware Stereo Interaction Field Estimation
We present GOLF, the first-place solution to the SHOW3D Interaction Field Estimation Challenge at HANDS@ECCV 2026. Given synchronized egocentric stereo views, the task is to predict a 3D vector from each of 21 hand joints to the closest point on the manipulated object. GOLF combines dense global context, locally sampled hand/object evidence, and common-frame Plücker-ray geometry. We adapt DINOv3 ViT-H+/16 with LoRA and trainable LayerNorm parameters, then jointly decode both interaction fields. Our primary model achieves an official score of 27.61 and a mean ADE of 27.96 mm on the hidden test set. An equal-weight ensemble with a complementary directly fine-tuned variant improves these results to an official score of 27.47 and a mean ADE of 27.82 mm, securing first place.
Read the full abstract
Atmospheric and Oceanic Physics
Neptune: An AI model for Global Ocean Subseasonal Prediction
Subseasonal-to-seasonal (S2S) forecasting is societally critical, supporting decision-making in sectors ranging from water and agricultural management to disaster risk reduction, energy planning, and insurance. Achieving reliable predictions at these timescales requires representing the ocean and its dynamics, but traditional physics-based Ocean General Circulation Models (OGCMs), are computationally expensive and difficult to develop and improve because of the code complexity. In this work, we propose Neptune, an end-to-end data-driven framework for global ocean and sea-ice components emulation tailored for S2S timescales, up to 60 days. Neptune combines Convolutional Neural Networks (CNNs) and Spherical Fourier Neural Operators (SFNOs) to effectively capture local features and global cross-scale interactions, thereby obtaining a coherent representation of the ocean state. Forced by prescribed daily atmospheric fields, Neptune emulates ocean state variables, from temperature and salinity, to zonal and meridional currents, from sea surface height to sea ice thickness and concentration, with daily outputs at the ocean surface and through the water column. Specifically, we propose two variants of Neptune, Neptune-1 and Neptune-025, capable of emulating the ocean state at 1{\deg} and 0.25{\deg} resolution, respectively. Evaluated against a suite of metrics, including statistics (RMSE, CRPS and ACC), physical coherency (Ocean Heat Content, Eddy Kinetic Energy and Ice Brier Score) and climate indices (ENSO and Z20 metric, IOD), Neptune successfully reproduces the spatio-temporal evolution of the oceanic fields up to 60 days, and is stable over long timescales. Neptune provides compelling evidence that end-to-end data-driven ocean emulators can become a powerful component of next-generation S2S forecasting systems, emulating ocean state at high spatio-temporal resolution.
Read the full abstract
Mathematical Finance
Numeraire Invariance of Entropy-Projected Martingale Measures
Let \(P\) be a fixed physical law and let \(Q\) be an equivalent martingale measure selected from the martingale-measure set associated with a chosen numeraire. A change of numeraire maps \(Q\) to \(T_LQ\), where \(d(T_LQ)=L\,dQ\) and \(L\) is the terminal likelihood ratio. The forward relative-entropy projection minimizing \(D_{\mathrm{KL}}(P\Vert Q)\) commutes with this transform because its objective changes only by the constant \(-E_P\log L\). The minimal entropy martingale measure (MEMM) orientation \(D_{\mathrm{KL}}(Q\Vert P)\) does not have this property, and a trinomial counterexample shows that independently recomputed MEMMs need not be likelihood compatible. We make two economic consequences explicit. First, the two entropy orientations are precisely the \(Q\)-dependent terms in the classical convex-dual objectives for logarithmic and exponential utility, respectively. Second, likelihood compatibility is equivalent to equality of the pricing functionals obtained in the two numeraires. Hence the forward selectors value every integrable claim consistently across numeraires, whereas the two MEMMs in the counterexample assign different prices to a nonreplicable digital claim. We also prove a finite-state class-level characterization: uniform invariance over the elementary one-period likelihood-ratio families forces a smooth convex \(f\)-divergence to be logarithmic, up to scaling and affine equivalence. Finally, in finite-state markets, the forward projection exists under the usual strictly positive feasible-point condition; its density \(dP/dQ^*\) is attainable log-optimal terminal wealth, and the minimum forward entropy equals maximal expected log growth.
Read the full abstract
Instrumentation and Methods for Astrophysics
A Missing Tool for Calculating Auto/Cross-correlation Function under Nonuniform Sampling Observations
Nonuniform sampling presents a long-standing challenge in astrophysical time-domain analysis, invalidating the standard autocorrelation and cross-correlation functions and forcing researchers to adopt ad-hoc methods like interpolation or binning, which introduce unquantified biases and lack rigorous error estimation. Here we introduce a new method for calculating the nonuniform autocorrelation function (NUACF) and nonuniform cross-correlation function (NUCCF) for irregularly sampled time series. Instead of relying on interpolation, it naturally evaluates the correlation function by incorporating time-interval weights and misalignment penalties. Monte Carlo simulations provide confidence bands for significance assessment and a complete error budget for the time delays that accounts for both flux uncertainties and sampling irregularity (essential but generally absent from existing methods). Through extensive simulations, we demonstrate our method outperforms traditional methods across various conditions, from strictly periodic to complex repeating variability patterns (e.g., intermittent but aperiodic). Its effectiveness is demonstrated via various real astrophysical data sets, revealing repetitive variability in stellar light curves, measuring time delays for multi-band disc reverberation in the AGN Fairall 9, and providing model-independent validation of time delays for the gravitationally lensed quasar HE 0435-1223. The method provides a rigorous and general solution to the ubiquitous problem of nonuniform sampling, positioning it as a useful tool for large-scale time-domain survey data analysis. The framework is also directly applicable to emerging time-domain phenomena such as fast radio bursts (FRBs), enabling, e.g., the study of correlations between persistent radio source luminosity and repeating FRB activity, or among the multi-parameter variability curves of FRB emission itself.
Read the full abstract
Functional Analysis
On the contractivity of the Riesz projection
This note presents a proof that the Riesz projection is contractive from $L^q$ to $L^{4(1-1/q)}$ for the case $1<q<\infty$. This fills in the previously unresolved parameter range of a conjecture of Brevig, Ortega-Cerdà, Seip, and Zhao.
Artificial Intelligence
CLAMP: Constrained Decoding for Vision-Language Embodied Planning
Embodied planning increasingly relies on vision-language models (VLMs) to translate instructions and visual observations into executable action sequences. However, fluent plans are not always executable. A VLM may refer to objects that are not visually observed, select actions whose required affordances are unavailable, or violate syntax and action constraints. We introduce CLAMP, a multimodal constraint-grounding framework that turns scene evidence into decoding-time constraints for a frozen VLM planner. CLAMP uses the initial observation to restrict object references to those supported by the scene, while a provided symbolic action model specifies state transitions and goals. During decoding, hard masks eliminate invalid next-token candidates, while a Hidden Markov Model (HMM)-based world-state lookahead module reweights the probabilities of the remaining feasible candidates based on action preconditions and goal reachability. This allows the planner to retain the VLM's language prior while preventing visually unsupported, unsafe, or infeasible candidates from entering the plan. For unseen tasks and environments, CLAMP adapts the HMM at test time using label-free continuations sampled from the frozen VLM. Experiments on VLABench, SafeAgentBench, and TaPA show that scene-grounded constraints improve object grounding and safety, while most remaining failures stem from perception errors or misaligned constraint specifications.
Read the full abstract
Materials Science
Polymorph Engineering of the Layered Rare-Earth Magnet GdAlGe
The recent advances in research on layered magnets have established them as invaluable materials for spintronic applications and precursors to 2D magnets. The problem is that the number of such materials available for experimentation is still low. To tackle the problem, one may employ phase engineering; in particular, the interplay between the phase stability and dimensionality provides great opportunities for design of layered magnets. Here, this interrelation is harnessed to produce epitaxial films of an overlooked layered polymorph of GdAlGe. The compound is formed by alternating triangular Gd and honeycomb AlGe lattices. It is stable in ultrathin films, up to 5 monolayers. In thicker films, an additional unwanted phase emerges. The structural quality of the ultrathin films is witnessed by electron and X-ray diffraction as well as electron microscopy. Magnetization measurements detect an anisotropic ferromagnetic state. The GdAlGe magnetism is corroborated by a study of electron transport revealing the anomalous Hall effect and negative magnetoresistance. The material is naturally integrated with Ge, a technological semiconductor, which should facilitate applications. The present study can serve as a blueprint for design of layered magnets via polymorph engineering.
Read the full abstract
Complex Variables
Sharp Bohr and Bohr-Rogosinski inequalities involving area measure for close-to-convex harmonic mappings
In this article, we investigate refined and generalized versions of the Bohr and Bohr–Rogosinski inequalities for a normalized subclass $\mathcal{P}_{\mathcal{H}}^{0}(\alpha)$ ($0 \le \alpha < 1$) of univalent close-to-convex harmonic mappings $f = h + \overline{g}$ defined on the open unit disk $\mathbb{D} \subset \mathbb{C}$. By incorporating non-negative monotone increasing functions associated with the planar area integral $S_r/\pi$ of the image domain $f(\mathbb{D}_r)$, we establish new sharp Bohr-type inequalities expressed in terms of the Euclidean distance $d(f(0), \partial f(\mathbb{D}))$. Furthermore, we establish sharp Bohr–Rogosinski-type inequalities involving powers of the modulus of the mapping $|f(z)|^p$ ($p \ge 1$). All associated radii are proven to be sharp, and extremal functions realizing the equality cases are explicitly identified. As applications, our results generalize and unify several well-known classical and recent theorems in geometric function theory.
Read the full abstract
Artificial Intelligence
Graph-Based Personalized Memory for LLM Agents: Representation, Evolution, Retrieval, and Evaluation
Large Language Model (LLM) agents are evolving from single-session tools toward long-term personal assistants that must adapt to individual users across tasks, contexts, and interactions. This shift makes memory a core requirement for personalization, since user preferences, goals, constraints, relationships, and past experiences are accumulated gradually and often change over time. Graph-based personalized memory provides a structured way to model such user information through explicit relations, temporal context, and evidence links. Such representations can model not only what an agent remembers about a user but also how memories are connected, revised, and retrieved to support personalized decisions. However, existing work remains fragmented across personalized agents and generic graph memory frameworks, making it difficult to understand the design space as a whole. This survey develops a lifecycle-oriented view of graph-based personalized memory for LLM agents. We organize existing studies around memory representation, memory evolution, memory retrieval, and memory evaluation. We further compare key design choices, discuss current evaluation practices, and open challenges in building reliable long-term personalized agents. This survey aims to clarify how graph-based memory can support adaptive, controllable, and user-centric LLM agents.
Read the full abstract
Quantum Physics
Fisher-Information Recovery in Superconducting-Qubit Magnetometry with Squeezed-Microwave Readout
The performance of superconducting-qubit magnetometers depends not only on magnetic-field encoding during Ramsey interrogation, but also on how efficiently the encoded information is recovered during readout. Squeezed microwaves provide a promising route to improve information extraction by suppressing readout noise, yet how such readout enhancement translates into accessible information for magnetic-field estimation remains unclear. Here we develop an effective detected-mode framework linking projected quadrature noise, state-assignment error, and the classical Fisher information accessible from binary readout outcomes. A finite mismatch between the squeezed quadrature and the discrimination axis produces an optimal squeezing strength through the competition between squeezed and anti-squeezed fluctuations. We further show that squeezed readout improves the readout-limited magnetic-field sensitivity by recovering information otherwise lost during state assignment, with the enhancement remaining robust against quadrature mismatch, transmission loss, and added noise. These results establish a quantitative connection between squeezed-microwave readout and measurement-stage information recovery in superconducting quantum sensing.
Read the full abstract
Materials Science
The Hidden Cost of Alloying: Disorder-Driven Transport Collapse in TMDs
Alloying in two dimensional semiconductors is widely used to tune bandgaps, yet its implications for charge and energy transport remain poorly understood. Here, we investigate MoS2xSe2(1-x) alloys as a model system to study the interplay between composition, thickness, and disorder. Optical transitions and valence band dispersions evolve continuously with both stoichiometry and number of layers, with negligible bandgap bowing and a composition dependent attenuation of thickness driven renormalization. In contrast, time resolved spatial mapping of photoexcited carriers reveals a pronounced and asymmetric collapse of carrier diffusivity at intermediate compositions, which cannot be accounted for by changes in effective mass or band alignment, and instead emerges from strong real space fluctuations in the local energetic landscape generated by random chalcogen substitution. Microscopic simulations reproduce the experimental trends and show that the character of disorder depends critically on the direction of alloying, producing either scattering barriers or deep trapping sites. Together, these results demonstrate that transport in TMD alloys is governed by disorder physics, overlooked by conventional optical and photoemission probes at equilibrium. Our findings establish transport as a stringent metric of electronic quality and high-light intrinsic limitations in the usage of TMD alloys for layered semiconductor devices.
Read the full abstract
Number Theory
A problem on the largest divisor $d$ of $N$ with $d\leq \sqrt{N}$
For a given number $N$, we consider the problem of computing two integers $1\leq r,f < N$ such that the set $$\mathcal{X}(N,r,f) = \{(a+b)-(f+\frac{Nr+1}{f}): ab=Nr\}$$ consists only of positive integers. Computing a solution to the problem is equivalent to finding a pair $(r,f)$ satisfying $l(Nr) < f \leq l(Nr+1)$, where $l(x)$ is the largest divisor of $x$ bounded by $\sqrt{x}$. This requires factoring both $Nr$ and $Nr+1$. We present a simple randomized algorithm that - avoiding factoring - computes pairs $(r,f)$. We give an exact formula for the total number of possible pairs $(r,f)$, and with the aid of empirical data we estimate that the ratio $$\frac{\phi(N)-2}{|\mathfrak{F}(N)|}$$ is approximately about $c*\log \log N$. Here, $\mathfrak{F}(N)$ is the set of unique $r$ appearing among all possible pairs $(r,f)$, $\phi(.)$ is the Euler's Totient function, and $c$ is a constant equal to 2 for prime $N$ and oscillates much for composite $N$. As a separate and independent case, we study the same problem of computing $(r,f)$ with $r>N$. We present a procedure to find such an $r$, which requires finding the least prime in an arithmetic progression.
Read the full abstract
Number Theory
Four explicit continued fractions for values of the Lerch transcendent and the Hurwitz zeta function
We prove four explicit continued fraction representations for the Lerch transcendent $\Phi(z, s, M+1)$, where $\Re M>0$ and $(z,s)\in\{(-1,1),(-1,2),(1,2),(1,3)\}$. All of the continued fractions have unit partial numerators, while partial denominators depend on the parameter $M$. The proofs combine equivalent transformations of continued fractions, generalized hypergeometric functions, three-term recurrences, and asymptotic analysis of minimal solutions. For $z=1$, the corresponding representations give continued fractions for the values of the Hurwitz zeta function $\zeta(2, M+1)$ and $\zeta(3, M+1)$.
Read the full abstract
Machine Learning
Structured Multi-Level-Set Inversion for 3-D Electromagnetic Inverse Scattering
Many dielectric targets encountered in 3-D electromagnetic inverse scattering exhibit piecewise-homogeneous material distributions, which provide useful structural information for inversion. To exploit this property, this paper proposes a structured multi-level-set inversion (SMLSI) method within a physics-driven neural network (PDNN) framework. Instead of directly parameterizing the contrast distribution, SMLSI represents it in terms of object support and material distribution. Multiple level-sets are mapped to soft-support functions and combined through a soft-union operation to construct the object support. The normalized contributions of these soft-support functions, together with trainable complex relative permittivities, determine the material distribution within the support. The level-set residuals are parameterized by a coordinate-based neural network, and the trainable parameters are optimized from scattered field data. Structure-aware regularization terms are introduced to promote binary object support and spatially coherent material regions while reducing material mixing and redundant activation of level-set supports. Adaptive regularization weighting is further employed to reduce the dependence on manual parameter tuning. Numerical experiments demonstrate that SMLSI improves reconstruction accuracy over direct contrast parameterization and enables the recovery of irregular 3-D geometries, closely spaced objects, and multiple material regions. Compared with CSI and $L_{2/3}$-FBE-WCIE, SMLSI achieves more accurate reconstructions on representative synthetic targets. Experimental validation using 3-D Fresnel data further demonstrates its applicability to measured scattering data.
Read the full abstract
Signal Processing
FedGenSC: Federated Generative Semantic Communication with Channel-Aware Adaptation
Integrating generative adversarial networks (GANs) into federated semantic communication (SemCom) is a natural progression, as generative priors can recover semantic fidelity under channel distortion that discriminative decoders cannot. However, naive GAN federation introduces three failure modes that prior work has, to the best of our knowledge, neither identified nor resolved: discriminator aggregation instability under non-independent and identically distributed (non-IID) data, semantic drift caused by divergent local embedding spaces, and channel-agnostic generation that cannot adapt to heterogeneous link conditions. We propose federated generative semantic communication (FedGenSC), which mitigates all three by employing a global generator with local-only discriminators, providing cross-client semantic information through a semantic prototype bank, and conditioning generation on the instantaneous signal-to-noise ratio (SNR). Experiments on the Europarl dataset over Rayleigh fading channels (K=10 clients, Dirichlet {\alpha}=0.5) show that FedGenSC under non-IID data outperforms the FedDeepSC baseline across the tested SNR range, achieving up to a 58.2% relative improvement in bilingual evaluation understudy (BLEU)-1 at 18 dB. Ablation studies confirm the independent contribution of each component.
Read the full abstract
Artificial Intelligence
A Three-Tier Persona Vector for Controllable User Simulation in Agentic Evaluation
Evaluating tool-augmented LLM agents requires diverse, realistic user inputs yet most evaluation frameworks use flat role descriptions ("you are an angry customer") that produce near-identical conversations regardless of the underlying scenario. In this paper, we propose a three-tier persona vector with 23 operationalized dimensions: 6 categorical demographics (jurisdiction, age, channel, device, language proficiency, time availability), 12 continuous behavioral traits (patience, assertiveness, digital literacy, etc.) sampled with Gaussian noise around curated profile base vectors, and 5 continuous emotional states (frustration, anxiety, trust, confidence, stress) that shift in response to scenario context. Orthogonal to the persona, a 4-level query-complexity overlay controls utterance phrasing from direct to deliberately vague. We evaluate the persona model inside a synthetic data generation pipeline across 64,698 multi-turn conversations spanning 8 named profiles and 3 production corpora. Key findings: (i) a 15.8 percentage-point spread in agent goal-achievement across personas confirms trait vectors produce measurably different user behavior; (ii) the same persona behaves differently across scenarios due to scenario-reactive emotional state shifts, validating the scenario-reactive design; (iii) domain-specific projects show persona sensitivity on booking-flow compliance (~15-20 percentage points gap between tier-aware and pressure-test personas), demonstrating the model faithfully reproduces real-world difficulty distributions; (iv) seven rule-described trait correlations produce auditable co-occurrence patterns without requiring learned covariance matrices. The persona model is fully specified for reproduction.
Read the full abstract
Quantum Physics
Multimode squeezed light generation and characterization
Nowadays, the realization of quantum computations and communications based on continuous variables has attracted significant attention due to a substantial expansion of the system dimensionality. The main progress in this area is attributed to the implementation of multimode systems based on squeezed states of light. One of the simplest ways to generate such states relies upon their producing in a single-pass optical parametric amplifier (OPA) under ultrafast pumping. However, for homodyne detection of such multimode states, the profile of the local oscillator (LO) must perfectly match the profile of the measured mode. Usually, this is not the case; therefore, a proper treatment of multimode squeezing is required. In this work, we study both theoretically and experimentally the multimode squeezed light generated in type-0 and type-II OPA. We characterize such sources using the Schmidt-mode representation and investigate the degree of squeezing depending on the LO spectral profile, employing a pulse shaping technique. In addition, we investigate cluster states that can be generated under the given experimental conditions. This work might have a significant impact on the realization of multimode quantum protocols.