Probability
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
We study to what extent the majorisation order between a probability measure and its $1$-Lipschitz image is preserved when both measures undergo Gaussian convolution. We show that the majorisation order is fully preserved in dimensions $n\leq 2$, and obtain dimension-dependent partial preservation in higher dimensions. The perspective taken is that majorisation between densities amounts to comparison through internal energies satisfying a certain pressure condition. Accordingly we introduce a notion of iterated nonnegativity of pressure, closely related to iterated pressures arising in optimal transport, placing the majorisation order within a graded hierarchy of internal energy comparisons. Using the observation that the volume of Euclidean neighbourhoods can be detected from internal energy measurements along the heat flow, we show that our results imply several principal known cases of the Kneser–Poulsen conjecture.
Astrophysics of Galaxies
Recovering the characteristic halo mass of rare populations from clustering: a benchmark using the Millennium Simulation
We measure how accurately the characteristic halo mass of a rare, massive population can be recovered from clustering alone, using dark matter haloes in the Millennium simulation, where the true masses are known. Massive samples of 25 to 10000 haloes are cross-correlated against field reference samples of 5000 to 50000 haloes at six redshifts over $0 \leq z \leq 5.29$. The bias is measured at $r = 8\,h^{-1}\,\mathrm{Mpc}$ and inverted to return a characteristic mass, which we compare with the mean logarithmic halo mass measured directly from the simulation. Across 192 configurations the characteristic mass is recovered with a median absolute offset of $0.103$ dex. We show that (i) accuracy depends only weakly on the number of rare objects while the uncertainty falls from $\pm 0.71$ dex at 25 objects to $\pm 0.08$ dex at 10000; (ii) enlarging the smallest samples gives the greatest return, with a doubling from 25 to 50 objects reducing the uncertainty by about a third, against a tenth from 5000 to 10000; and (iii) the reference population carries an error of its own, which propagates coherently into every cross-correlation measured against it and is not reduced by observing more rare objects. Working with haloes rather than galaxies isolates the clustering measurement from the galaxy–halo connection, providing a benchmark before population-specific complications are introduced. For the best configurations, the uncertainty contributed by the calibration of the bias–mass relation, at $\sim 0.07$ dex, already matches that of the measurement itself.
Read the full abstract
Statistical Mechanics
Irreversibility of many-body harmonic oscillators characterized by generalized Husimi's adiabaticity parameter
In a classical paper [K. Husimi, Prog. Theor. Phys. 9, 381 (1953)], Husimi showed that, for a single harmonic oscillator with a time-dependent angular frequency and initial conditions sampled from an equilibrium distribution, the averaged energy cannot decrease after a cyclic operation. This irreversibility is quantified by Husimi's adiabaticity parameter constituted with adiabatic invariants. In this work, we generalize Husimi's framework to a many-body system of one-dimensional harmonic oscillators coupled on an arbitrary connected network, with all spring constants sharing a common time dependence. For a system attached to a fixed wall, the dynamics can be decomposed into independent normal modes by transforming to mass-weighted coordinates and diagonalizing the resulting positive definite matrix, with each mode characterized by its own Husimi's adiabaticity parameter. By defining the generalized Husimi's adiabaticity parameter as their arithmetic mean, we derive the exact time evolution of the averaged energy and establish its non-decrease under cyclic operations. Numerical simulations confirm these results for both a uniform nearest-neighbor chain and a heterogeneous network.
Read the full abstract
Multimedia
AdoDAS: A Privacy-Preserving Multimodal Challenge for Adolescent Depression, Anxiety, and Stress Assessment
Adolescent depression, anxiety, and stress (D/A/S) call for scalable tools that complement, rather than replace, professional evaluation. Under a privacy-preserving policy, the AdoDAS Grand Challenge withholds minors' raw recordings and distributes anonymized audio-visual representations and ASR-derived text. Its 6,000 participants provide 24,000 segments across one scripted-reading and three open-response sessions. Two tracks assess multi-task binary D/A/S screening and ordinal prediction of 21 DASS-21 item responses. From 191 registrations, the final leaderboards included 95 eligible screening teams and 64 item-prediction teams. Audio-visual baselines achieved 0.4604 mean F1 and 0.2675 mean Quadratic Weighted Kappa; leading submissions reached 0.5921 and 0.2776. Representative systems emphasize cross-session modelling, temporal multimodal fusion, psychometric structure, and task-aware calibration.
Read the full abstract
Machine Learning
Disentangling Steering Vectors
Activation steering has emerged as a lightweight, inference-time approach to control the behavior of Large Language Models (LLMs). However, traditional steering vectors used to intervene in LLMs' activations, such as those derived from the difference-in-means method, tend to entangle multiple semantic and stylistic concepts into a single composite direction, leading to unpredictable steering effects. Our core objective is to disentangle this composite direction into its constituent concepts. To this end, we propose Steering Vector Dissection, a framework to explicitly isolate individual and semantically consistent features from these composite directions. Specifically, we pair positive and negative activations and take their differences to generate a set of instance-level steering vectors, and train a dedicated Sparse Autoencoder (SAE) directly on them. Quantitative evaluations across two datasets, two models, and two intervention depths show that our method yields a set of semantically consistent basis vectors whose steering effects are mutually distinguishable. Furthermore, we show that this disentanglement enables precise control over model behaviors.
Read the full abstract
Computation and Language
Aha-Flow Distillation: Flow Markers Matter in LLM Reasoning
We identify the Flow Moment, a reasoning pattern characterized by sustained, process-confirming verbalizations such as I'm doing, in contrast to the revision- and backtracking-oriented Aha Moment. We refer to their corresponding linguistic expressions as Flow Markers and Aha Markers, respectively. Based on this observation, we construct Flow-CoT by rewriting the discourse markers of original reasoning traces while preserving their underlying reasoning content, and use it as auxiliary supervision for on-policy self-distillation (OPSD). We further propose Aha-Flow Distillation (AFD), a dual-mode extension of OPSD that pairs different forms of privileged information with corresponding reasoning instructions. The Aha branch retains concise solution-based supervision, while the Flow branch introduces rewritten Flow-CoT under a direct and confident reasoning instruction. At inference time, the model uses only the standard reflective instruction, so Flow-style reasoning serves purely as a training signal. Experiments on AIME25 and HMMT25 show consistent improvements across Qwen3-8B and Qwen3-4B: AFD improves Avg@12 from 60.8 to 61.3 on Qwen3-8B and from 57.5 to 58.6 on Qwen3-4B over our reproduced OPSD baselines. Controlled ablations further show that, with the same Flow-CoT/Aha-CoT composition, dual-mode training improves Avg@12 from 59.5 to 60.1, indicating that the benefit comes not only from introducing heterogeneous reasoning supervision, but also from how it is organized during self-distillation. The code is available at https://github.com/Wang-Xiaodong1899/Aha-Flow-Distillation.
Read the full abstract
Quantum Physics
Capability-Gated Conformance Testing of Quantum Error-Correction Decoder Libraries
A quantum error correction decoder is a library other people's results depend on, judged in one dominant way. Sample errors, decode, and count wrong logical observables. We ask what else can be checked there. Our conformance contract needs no oracle. One check asks that a returned correction explain the syndrome in the caller's index space. The other hands a decoder one instance under two presentations differing only in bookkeeping, where two feasible corrections of different weight prove the heavier is not minimum-weight. Verdicts are gated on what each library declares, so a firing contradicts a published guarantee. Nine configurations from five public libraries give three results. Documentation answers 4 of 54 capability questions. Bounded-distance correctness, the property callers most depend on, has a direct declaration yield of 0.0%, though its hypotheses hold in 62.1% of cases. Presentation sensitivity is real but shallow. One solver moved to a 26% heavier correction under a different numbering, which reaches the logical class at most once in twenty thousand shots. Established evaluation misses corruptions that preserve logical parity, while one summation over the caller's weights catches every one we injected. All 639 certificates ship as bundles a standalone verifier re-derives from first principles.
Read the full abstract
Sound
Frequency Estimation Based on SNR-adaptive Frequency Estimator Under Wide SNR Range
Frequency estimation is the problem of estimating individual tone frequencies from noisy multi-tone sinusoidal signals. Existing frequency estimation methods have difficulty accurately estimating both the number of tone frequencies and the individual tone frequencies in low signal-to-noise ratio (SNR) environments, because weak tone frequency components are buried in noise. In addition, existing methods generally exhibit a trade-off between robustness at low SNR and frequency estimation precision at high SNR, making it difficult to achieve consistently superior frequency estimation performance over a wide SNR range. To overcome these limitations, this paper proposes an SNR-adaptive frequency estimator (SAFE). SAFE consists of a time-frequency image neural network (TFINet), which enhances weak tone frequency components at low SNR, and an SNR-based frequency selector (SFS), which selects an appropriate frequency estimator according to the SNR of the estimated tone frequencies. TFINet enhances tone frequency components even in the low-SNR range, while SFS estimates the SNR of each tone frequency and selects either a robust frequency estimator or a super-resolution frequency estimator according to the estimated SNR. This enables SAFE to achieve robustness at low SNR while preserving high precision at high SNR. Simulation results show that SAFE achieves an False Negative Rate (FNR) of 13.00% over the SNR range from -10 dB to 0 dB, corresponding to an 13.04% improvement over the state-of-the-art method. In addition, SAFE reduces the Nearest Neighbor-Root Mean Squared Error (NN-RMSE) by 56.67% compared with the state-of-the-art method, demonstrating that SAFE performs more accurate frequency estimation. Furthermore, experiments using real-world data demonstrate that SAFE provides robust frequency estimation performance even in practical environments with clutter.
Read the full abstract
Econometrics
Choosing the Dictionary and Penalty for IV-LASSO
Estimating an instrumental variables (IV) first stage with the least absolute shrinkage and selection operator (LASSO) requires choosing a dictionary and penalty. Existing rules target first-stage objectives rather than precision of the structural parameter estimate. We derive a feasible approximate mean squared error (AMSE) criterion for full-sample IV-LASSO with one endogenous regressor and homoskedastic Gaussian errors. For a prespecified list of dictionary-penalty pairs, the criterion trades off approximation of the optimal instrument against a complexity cost that depends on the first two moments of the LASSO effective dimension and on the degree of endogeneity. The selected pair attains the smallest candidate-specific AMSE component up to vanishing relative error. In simulations calibrated to Paravisini et al. (2015), the criterion improves substantially on plug-in rules at low endogeneity and on cross-validation at high endogeneity, with nearly identical results under non-Gaussian errors. A supplementary appendix formally extends analysis to a class of non-Gaussian errors.
Read the full abstract
Quantum Physics
A Single Fixed Shallow Circuit for Classical Shadows of Arbitrary n-Qubit States
Classical shadows extract quantum properties from reusable classical records, typically obtained through randomized measurement settings. Although individual settings may be shallow, switching among them introduces control, calibration, and reconfiguration costs beyond conventional metrics. Here we construct, for every $n$, a single fixed shallow quantum analyzer whose Born outcomes replace externally sampled settings as labels for reusable shadow snapshots. The analyzer combines a freshly prepared $n$-qubit fiducial register $A$ with parallel Bell readout of $A$ and the unknown system $S$, producing one $2n$-bit record per copy. The same circuit realizes a rank-one minimal informationally complete measurement: one fixed setting replaces the $3^n$ local-Pauli settings conventionally used for complete reconstruction while retaining the minimum $d^2$ outcomes with $d=2^n$. Its Pauli-diagonal frame admits an analytic inverse. For fixed Hermitian observables, the Haar-averaged conditional variance has a dimension-independent coefficient, whereas the state-uniform coefficient is dimension dependent. For $n\ge3$, worst-state Pauli variances remain bounded in the axial sector and scale as $\Theta(d)$ in the mixed sector. Fiducial preparation before system contact uses $n-1$ arbitrary two-qubit gates, no work qubits beyond $A$, and logarithmic depth under all-to-all connectivity. The unknown system undergoes one parallel system-ancilla entangling layer followed by local Hadamards and readout. Measurement-assisted preparation achieves $O(1)$ adaptive quantum depth using $n+1$ extra qubits in the heralded route, or $O(n\log n)$ auxiliaries for deterministic completion. These tradeoffs show that part of the measurement-setting randomness and control complexity normally supplied shot by shot can instead be compiled into a fixed reusable analyzer.
Read the full abstract
Machine Learning
Efficient Learning and Symmetry Discovery under Exact Invariances
Learning with group invariances is central to many scientific and geometric learning problems, yet its computational foundations remain poorly understood. Even for classical supervised regression settings, it has been unclear whether one can efficiently compute a regression function that is exactly invariant to a given group action. Recent work showed that exact invariance can be enforced in polynomial time when the underlying group is finite and known, but left open the cases of infinite groups and unknown symmetries. In this paper, we resolve both challenges. First, we present the first polynomial-time algorithm for learning with exact group invariances that applies uniformly to finite and infinite groups. The runtime is polynomial in the data dimension and sample size, and independent of the group, while achieving strong generalization guarantees. This provides a computational explanation for the empirical success of invariant and equivariant methods in geometric machine learning and partially answers a recent open question in the literature. Second, we study learning in the symmetry discovery setting, where the invariance group is unknown. Focusing on the subgroup lattice of a finite group, we show that exact symmetries can be identified from data and exploited for learning in polynomial time. For regression over finite-dimensional feature spaces, our algorithm provably recovers the underlying symmetry, matches the minimax-optimal sample complexity of the known-symmetry setting, and runs in time polynomial in the data dimension and sample size. Our analysis relies on tools from random Cayley graphs and expander theory, which may be of independent interest.
Read the full abstract
Metric Geometry
Hopf quotients of the infinite-dimensional Gaussian pyramid
We study the infinite-dimensional Gaussian pyramid and its quotients by the global sign flip and the $U(1)$-Hopf action. We resolve affirmatively a long-standing problem posed by Tomohiro Fukaya around 2014: these three limiting geometries are pairwise non-similar, meaning that no positive rescaling makes any two of them coincide.
Read the full abstract
Computer Vision
LoGAN: Multilingual Font Localization with Generative Agents
Localizing a font into new languages is a highly intricate task requiring precise design adaptation of glyphs, color/texture, and spacing/kerning, from source to target languages. Most existing methods focus on single glyph generation with limited capability in handling multilingual font rendering. In this work, we propose LoGAN, a VLM-based agentic framework for few-shot multilingual font localization, which takes in a small number of individual glyphs from a font or letters from a logo and uses them to generate complete character sets in other languages. LoGAN breaks down this task into multiple components: a glyph-level diffusion model, a style finetuning module, a spacing and kerning transfer algorithm, and a texture expansion model, with a VLM agent coordinator. LoGAN achieves broad language coverage for font localization with various styles, including Chinese/Korean/Japanese (CJK). We evaluate our approach on both font and real-world logo datasets spanning more than 27 languages and compare it against both specialized font generation and state-of-the-art image editing models with strong text rendering capabilities (e.g., FLUX, Nano-Banana). Our approach yields higher glyph fidelity while maintaining better style, texture, and kerning consistency according to both quantitative and qualitative evaluations.
Read the full abstract
Statistics Theory
The joint density of top k sample eigenvectors and principal subspace inference
In this paper, the exact joint density of the top $k$ sample eigenvectors is derived for any $p\times p$ population covariance matrix $\varSigma$, sample size $n > p -1$, and $1 \le k \le p$. Using symmetric functions such as zonal polynomials and a dual summation identity by I. G. Macdonald, the result is expressed as a series in terms of determinants of differential operators. A matrix Kummer transformation converts the resulting local alternating expansion into a globally absolutely convergent series over the entire positive definite matrix space. For $k=2$, the ordered eigenvalue integrals reduce to the Gauss hypergeometric function ${}_2F_1$ with a simple closed form when $n=p+1$. Previously, explicit formulas were only available for a scalar matrix $\varSigma = \sigma^2 I_p$ or a general matrix $\varSigma$ with $p=2$ or $k=1$, obtained by T. W. Anderson and T. Sugiyama, respectively. The frame law also induces an exact Grassmann density that provides a finite-sample benchmark for principal subspace inference.
Read the full abstract
Nuclear Theory
Machine Learning Assisted Parametrisation and Prediction of Bare to Neutral $\beta^-$ Decay Rate Ratios of Fully Ionised Atoms
The \beta^- decay scenario of a nucleus differs significantly from its terrestrial characteristics when the atom is highly ionised or fully stripped of its electrons. Under such conditions, in addition to the conventional \beta^- decay to atomic continuum, the emitted electron may occupy a vacant atomic orbital of the daughter atom, giving rise to bound state \beta^- decay. Such environments occur naturally in stellar interiors and can also be produced in storage ring or plasma trap experiments. The relative contributions of the continuum and bound state decay channels depend on several nuclear and atomic properties, such as the decay Q value, the mass and proton numbers of the daughter nucleus. Consequently, predicting decay rate enhancements becomes a complex problem that generally requires detailed theoretical calculations. In this work, we explore the use of machine learning (ML) techniques to investigate the systematics of \beta^- decay in fully ionised atoms. ML based regression models are developed using theoretically calculated decay rate data for a set of astrophysically relevant allowed \beta^- transitions. A global parametric expression is proposed to estimate the bare-to-neutral decay rate ratio, and its predictive capability is examined using independent validation data. In addition, Random Forest and Artificial Neural Network models are trained to predict the bare-to-neutral decay rate ratio directly from a set of nuclear and atomic parameters. The results show that ML methods can successfully capture the underlying trends governing decay rate enhancement and provide rapid estimates without computationally demanding calculations. The developed models offer a practical tool for estimating the maximum possible \beta^- decay rates and will be useful for applications in nuclear astrophysics, future storage ring or plasma trap experiments, and nucleosynthesis modeling.
Read the full abstract
Probability
Locality of Bernoulli Site Percolation on Transitive Graphs
We prove locality of the critical probability for Bernoulli site percolation on infinite, connected, locally finite, vertex-transitive graphs, under the usual assumption that the critical probabilities stay uniformly below one. The proof develops site-percolation forms of the two-ghost inequality, sharp-threshold and snowballing estimates, and the nonunimodular height method. We also establish the plentiful-tube geometry needed for the multiscale argument from quantitative structure and random-walk estimates.
Read the full abstract
Astrophysics of Galaxies
Between Little Red Dots and Star-Forming Galaxies: A Sequence in the Optical Continuum Slope
The "Little Red Dots" (LRDs) uncovered by JWST are typically analyzed as a distinct class of objects; however, whether they are separated from ordinary star-forming galaxies by a genuine boundary has not been tested at the population level. Using a sample of $\sim3,500$ galaxies at $3 \lesssim z \lesssim 8$ with NIRSpec/PRISM spectra from the DAWN JWST Archive, we modeled their rest-frame UV-to-optical continuum emission with power laws and located the sources in the $\beta_{\rm{UV}}$-$\beta_{\rm{opt}}$ plane. Alongside a dense locus of star-forming galaxies and a diagonal sequence which can be explained by dust reddening and/or quiescence, we identify a vertical sequence extending from the locus of star-forming galaxies into the LRD region. This sequence is smoothly and continuously populated, with no clear boundary between the two populations. Its middle section, $-1 \lesssim \beta_{\rm{opt}} \lesssim 0$, hosts an intermediate population of $\sim50$ sources that falls outside the LRD selection boundary and has been largely overlooked; NIRCam imaging shows their rest-frame UV and optical median stacked images lie between point-like LRDs and extended star-forming galaxies. We modeled the vertical sequence as linear mixtures of stacked star-forming spectra and a black hole star (BH*) template derived from "The Cliff", one of the most compelling BH* candidates, and we infer black hole-to-stellar mass ratios $\eta = M_{\rm{BH}}/M_* \approx 0.006-0.1$ at the LRD end. We propose $-1 \lesssim \beta_{\rm{opt}} \lesssim 0$ as a practical selection window for possible transition candidates and argue that LRDs may be the compact, BH*-dominated extreme of a continuous population in which the relative contributions of the host and BH* vary smoothly.
Read the full abstract
Number Theory
$p$-adic properties of division polynomials and algebraic sigma functions
Let $p \geq 5$ be a prime, let $K$ be a finite extension of $\mathbb{Q}_p$, and let $E/K$ be an elliptic curve with good reduction. Let $F_n$ denote the $n$-division polynomial of $E$. Silverman proved that if the reduction is ordinary, then for every $P \in E(K) \setminus \hat{E}(K)$ and a suitable power $q$ of $p$, the sequences $(F_{mq^k}(P))_{k \geq 0}$ converge $p$-adically to limits that are algebraic over the field of definition of $E$. In both the ordinary and supersingular cases, we show that these sequences converge, and determine these limits explicitly in terms of the values of Mumford's algebraic sigma function attached to the Teichmüller lift of the prime-to-$p$ torsion component of the reduction of $P$. In particular, the limits are algebraic in the supersingular case as well. As an application, we obtain explicit $p$-adic limit formulas for nonsingular elliptic divisibility sequences.
Read the full abstract
Instrumentation and Methods for Astrophysics
An adaptive parameter optimization method for astronomical image alignment using Bayesian optimization. I. A hierarchical search strategy for FWHM and SNR
The alignment and stacking of astronomical images are fundamental steps for detecting faint objects and performing high?precision astrometry. In traditional alignment workflows, the extraction of source lists is critically dependent on key parameters such as the Full Width at Half Maximum (FWHM) and the Signal-to-Noise Ratio (SNR) threshold. These parameters are often selected manually through an inefficient trial-error process that lacks objectivity and does not guarantee optimal results. We present an adaptive method for optimizing astronomical image alignment parameters based on Bayesian Optimization (BO). We frame the parameter search as an optimization problem, with an objective function designed to maximize the number of successfully matched source pairs. By employing a hierarchical search strategy, we perform an efficient global search for FWHM and SNR to automatically determine the optimal combination for a given observational dataset. Experimental results demonstrate that our method effectively handles image data with varying seeing conditions and back?ground noise levels. It rapidly converges to a robust set of alignment parameters, achieving sub-pixel accuracy and significantly improving the automation level and success rate of the alignment process. This work may provide a useful basis for developing large-scale, automated astronomical data processing pipelines
Read the full abstract
Cryptography and Security
CIPHER: Benchmarking Cross-record Inference over Privacy-Hardened Evidence Records
Reasoning over privacy-constrained records requires combining structured attributes with evidence from free-text narratives. We introduce CIPHER (Cross-record Inference over Privacy-Hardened Evidence Records), a benchmark of expert-validated questions from consumer-finance, clinical, and law-enforcement records. The questions cover common tabular operations and include executable SQL supervision. We evaluate retrieval, prompting, table-specialist, and hybrid symbolic-neural systems under native redaction and surrogate-based evidence restoration. All system families exhibit substantial failures even when supporting records are provided. Most errors arise from incorrect record selection and predicate interpretation rather than arithmetic execution. Privacy transformations have non-uniform effects, sometimes obscuring necessary evidence and sometimes reducing distraction. CIPHER provides a reproducible testbed for diagnosing these failures and assessing how transformations of sensitive text affect reasoning over hybrid records.
Read the full abstract
Instrumentation and Methods for Astrophysics
Filter and baffle design for the short-wavelength camera on SOLAR-C EUVST
SOLAR-C, which is scheduled to launch in the late 2020s with the Extreme Ultraviolet High-Throughput Spectroscopic Telescope (EUVST), will deliver EUV spectroscopy of the solar atmosphere with unprecedented temperature coverage and resolution. We present the optical design and mechanical impacts for the short-wavelength camera, and how our design choices affect scientific performance using complete forward modelling. We discuss the design and optical impact of the thin-film aluminium filter and supporting mesh, including diffraction and shadowing effects in spectral and spatial dimensions. We analyse this using our own Fourier optics method and analytical modelling, and quantify the scientific impact using forward modelling driven by a simulated solar atmosphere. We address the vulnerability of thin-film filters during launch depressurisation using vacuum chamber testing. We describe the stray-light baffle design and low-scatter coatings which suppress visible stray light.
Read the full abstract
Networking
ASTRA: Low-Overhead Runtime Architecture for STReam Adaptation in Video Analytics
Real-time video analytics is crucial for smart city applications and cloud-connected vehicle control. To improve analytics accuracy, it is desirable to process the video at the highest resolution and frame rate. However, due to limited resources, streaming and processing video at the highest resolution and frame rate from all cameras is not feasible and adversely affects the analytics latency. Intelligent adaptation of cameras' resolutions and frame rates based on network conditions and the video content is crucial in order to optimize the performance. In this paper, we present ASTRA, a low-overhead runtime architecture for online adaptation of live camera analytics at the edge. ASTRA can execute various online algorithms as a black box. We deployed ASTRA in the realistic NSF COSMOS testbed and uniquely assessed its real-time performance using COSMOS' street-level cameras. We further evaluated ASTRA with up to eight emulated cameras by streaming a comprehensive video dataset under real-world network conditions. We used ASTRA's architecture to evaluate the practical performance of several classes of adaptation algorithms, including theoretical and empirical methods. The results indicate that ASTRA can provide system reliability (i.e., the probability of meeting accuracy and latency requirements) of more than 90% while maintaining performance within a deviation of less than 10% from optimal offline performance with average GPU utilization overhead of around 2% per camera.
Read the full abstract
Physics and Society
Phase spaces of phase spaces: A method of constructing generalizable and topologically informed bases from time series demonstrated by reconstruction of the geometry of the macroeconomy
Laws from physics to social science rest on a small, enumerable number of factors explaining variation in the systems they describe. While the orthodoxy of economics, physics and chemistry relies on a small number of analytically interpretable laws, large data models pursuing universality throw in all available data. Both fail in their extremes through the same flaw, an incorrect basis: too few orthogonal variables to describe system variation, or an explosion of dimensionality around fundamentally low dimensional but unknown dynamics. We propose and validate a framework that extracts an interpretable, mathematically sound, low dimensional and dynamically informed basis directly from arbitrary timeseries. Combining Takens' Delay Embedding Theorem, dimensionality reduction and optimal transport, we produce a phase space of phase spaces (PSoPS) on which all interrelated dynamical series exist: barycenters formed from many phase spaces, together with the families of transport plans between individual phase spaces and the barycenter, yield a basis describing the relational geometry of each timeseries and the topology of economic relationships. The macroeconomic PSoPS built from 281,536 series is a single connected object of effective dimension 7.3 +/- 0.2, into which the canonical laws (Phillips, Okun, Solow) embed as roughly 2-dimensional looped sub-attractors. The coupling between concept barycenters is directed, with the price of money and productivity as net sources and credit and output as sinks, recovering monetary transmission from geometry alone. IAAFT phase-randomized surrogates destroy 36-43% of the dimension across the three laws and 38% on the full corpus, so the basis is irreducibly nonlinear. We close with proposed extensions to fields whose dynamics share attractor geometries and to systems that are intrinsically low dimensional yet hard to navigate in raw coordinates.
Read the full abstract
Representation Theory
Affine and cyclotomic Brauer categorification via $\imath$-Kac–Moody $2$-categories: the half-integral type $\operatorname{AIII}$ case
Cyclotomic Brauer or cyclotomic Nazarov–Wenzl algebras arise in higher Schur–Weyl dualities involving parabolic categories $\mathcal O$ for Lie algebras of types $B,C$ and $D$. Their connection with Kac–Moody-type categorification is substantially less developed than the corresponding type $A$ theory for cyclotomic Hecke algebras. We construct a categorical bridge between affine Brauer-type representation theory and the half-integral quasi-split type $\operatorname{AIII}$ $\imath$-Kac–Moody $2$-category of Bao–Shan–Wang–Webster. More precisely, an action of the affine Brauer category on a locally Schurian category, with dot spectrum exactly $\frac{1}{2}+\mathbb Z$, determines a generalized nilpotent $2$-representation of the even component $\mathfrak U^{\imath}_{+}$. Conversely, every nilpotent $2$-subrepresentation of an ambient locally Schurian $2$-representation of $\mathfrak U^{\imath}_{+}$ carries a compatible affine Brauer action whose dot spectrum is contained in $\frac{1}{2}+\mathbb Z$. Applying these constructions to cyclotomic quotients, we prove that the locally unital algebra attached to a $\mathbf u$-admissible cyclotomic Brauer category is isomorphic to the locally unital algebra attached to the corresponding cyclotomic quotient of the principal $2$-representation of $\mathfrak U^{\imath}_{+}$. Consequently, the associated cyclotomic Brauer (or cyclotomic Nazarov–Wenzl) algebras acquire natural $\mathbb Z$-gradings. To our knowledge, this is the first such categorical realization of $\mathbf u$-admissible half-integral cyclotomic Brauer algebras by means of an $\imath$-Kac–Moody $2$-category. It provides the categorical and graded framework toward a Brauer-type extension of the Brundan–Kleshchev–Ariki theory in which coideal algebras and $\imath$-canonical bases are expected to replace ordinary quantum groups and canonical bases.
Read the full abstract
Machine Learning
HyperTransfer: Understanding the Equivalence between Base Optimizer and Hyperball
Hyperball optimizers constrain parameter norms and update only their directions, establishing a distinct paradigm for neural network optimization. Although this geometry appears fundamentally different from that of conventional Base Optimizers, which update both parameter norms and directions, we show that the two paradigms are dynamically equivalent for scale-invariant networks. Building on this equivalence, we propose HyperTransfer, which constructs a Hyperball optimizer that reproduces the dynamics of a target Base Optimizer using only its initialization and learning-rate schedule, without running the target optimizer itself. We further derive the inverse mapping and extend the framework to non-scale-invariant networks. Experiments show that both HyperTransfer and the inverse mapping produce loss trajectories nearly identical to those of their targets, suggesting that Hyperball dynamics are governed primarily by the induced effective learning-rate schedule and optimizer state.
Read the full abstract
Classical Physics
Formation of grain boundaries in ductile single crystals under plane-strain simple shear: a block-coordinate finite element method
Large plastic deformation can drive an initially uniform single crystal to spontaneously subdivide into misoriented grains separated by thin dislocation walls – a pattern-forming instability rooted in the loss of convexity of the crystal's elastic energy at large strain. We study this phenomenon for a ductile crystal in plane-strain simple shear within continuum dislocation theory, using a polyconvex (Ciarlet–Geymonat) elastic energy that guarantees existence of minimizers for the coupled deformation–slip problem. Minimizing over the plastic slip yields a condensed energy of double-well form whose non-quasiconvexity favours a lamellar microstructure; the gradient of the geometrically necessary dislocation density regularizes it, giving the grain boundaries a finite thickness and energy as functions of the misorientation angle. A block-coordinate finite element scheme – alternating a convex non-smooth solve for the slip with a Levenberg-regularized Newton solve for the deformation – resolves this microstructure numerically and detects its spontaneous onset, reproducing the lamellar grain structure in agreement with the closed-form analysis.
Read the full abstract
Populations and Evolution
Synergistic Effects of Behavioral Feedback and Seasonality Generate Chaos in Cooperative Multi-Pathogen Systems
Infectious diseases may interact by competing for the same hosts or by facilitating subsequent infections. Understanding the dynamics of such multi-pathogen systems, particularly those subject to endemic seasonality and mitigation, is essential for designing robust public health interventions. We propose a three-stage modeling framework to disentangle the interplay between seasonality and behavioral feedback as a way of mitigation. First, we analyze a coupled susceptible-infectious-recovered-susceptible (SIRS) system without external forcing and show that the abrupt transition between the disease-free and endemic equilibria arises from a backward bifurcation-induced first-order phase transition. Second, we independently examine seasonality and behavioral feedback, characterizing where and when oscillatory behavior is induced near critical tipping points. Third, we demonstrate that their combination generates complex multi-annual wave patterns, with high-incidence cycles driven by seasonality and low-incidence intervals driven by behavioral feedback. By mapping stability as a function of seasonal forcing, mitigation strength, and cooperativity, we identify distinct period-doubling cascades with chaotic signatures arising from different mechanisms: the interplay between seasonality and behavior, and inter-pathogen cooperativity with the backward bifurcation it induces. We then analyze how these mechanisms interact across parameter ranges. Altogether, we show that cooperation fundamentally expands the spectrum of possible epidemic patterns, highlighting the importance of considering multi-pathogen interactions in epidemic modeling and control strategies.
Read the full abstract
Combinatorics
Bounds for the Vertex Chromatic Number of Connected Triangle-Free Graphs
It was recently shown that every connected graph of order $n \geq 5$ and size $m$ satisfies $\chi(G) \leq \left\lceil \frac{m}{\sqrt{n}} \right\rceil$, and it was asked whether the stronger inequality $\chi(G) \leq \left\lceil \frac{m}{2\sqrt{n}} \right\rceil + 1$ holds for every connected triangle-free graph. In this paper, we answer this question in the affirmative. In fact, we prove that every connected triangle-free graph $G$ with $G \not\cong C_5$ satisfies $\chi(G) \leq \left\lceil \frac{m}{\sqrt{5.5n}} \right\rceil + 1$, where the constant $\sqrt{5.5}$ cannot be replaced by any constant greater than or equal to $\sqrt{6}$, and the equality holds for the Grötzsch graph and for every odd cycle of length between $7$ and $21$.
Read the full abstract
Computer Vision
ARNAI: Artifact Removal Network based on Autoencoding and Inpainting for Robust Spinal Image Segmentation and Measurement
Purpose: This study aims to develop an AI framework applicable for postoperative imaging for automated measurement of spinopelvic parameters on radiographs with robustness to the presence of spinal implants. Materials and Methods: We retrospectively reviewed lateral lumbar spine radiographs from two institutions (Internal: January 2017–December 2024; External: October 2021–September 2025). We developed the Restore, Segment, and Measure (RSM) framework, incorporating a novel Artifact Removal Network based on Autoencoding and Inpainting (ARNAI) to mitigate implant-related artifacts in postoperative radiographs. Segmentation and spinopelvic parameter (PT, LL, SS, SCA) measurement performance were assessed using Wilcoxon signed-rank tests and intraclass correlation coefficients. Results: When ARNAI was added to a recent Transformer-based segmentation model, FCBFormer, the mean DSC increased to 0.870 from 0.814, with marked gains at L3–L5 and smaller improvements at L1–L2. On 91 radiographs with implants, the mean L4–L5 segmental Cobb angle error decreased to 4.7 {\deg} from 15.6–16.2 {\deg}, an average error reduction of 70%. The ICC for L4–L5 segmental Cobb angle improved to 0.54 (Rater 1) and 0.59 (Rater 2) from 0.18, and ICCs for pelvic tilt, lumbar lordosis, and sacral slope all exceeded 0.70. The improvement in L4–L5 segmental Cobb angle error was statistically significant in the internal implant-containing cohort after correction for multiple comparisons. Conclusion: The proposed RSM framework improved automated spinopelvic parameter measurement in implant-containing postoperative radiographs. By mitigating implant-related artifacts, ARNAI improved segmentation and downstream measurement accuracy, with the greatest benefit observed for L4–L5 segmental Cobb angle estimation, where the mean error was reduced by approximately 70%.
Read the full abstract
Image and Video Processing
AstraMoE-SR: Trajectory-Guided Diffusion for Blind Satellite Jitter Deblurring and Super-Resolution
Pushbroom satellite imaging couples limited spatial resolution with platform attitude instability. Platform jitter produces spatially varying motion blur because each scan line is acquired under a different instantaneous attitude, while perspective geometry causes the same perturbation to induce different pixel displacements across the field of view. Existing blind restoration methods that assume a spatially invariant kernel and satellite jitter correction methods that rely on auxiliary observations are therefore not directly applicable. We present AstraMoE-SR, a single-image framework that jointly restores motion blur and spatial resolution without auxiliary measurements. Rather than estimating a blur kernel, we infer how the camera moved by reparameterizing degradation as a local exposure trajectory under pushbroom geometry. A conditional diffusion model estimates the trajectory distribution, mitigating the over-smoothing of high-frequency jitter by deterministic point estimation. The predicted trajectory conditions a pretrained latent diffusion backbone through trajectory-guided geometric alignment and spatially adaptive reconstruction. We further show that the remaining point-wise trajectory error is consistent with intrinsic jitter-phase ambiguity that is not resolved by increasing estimator capacity. On all 1,411 DOTA-v1.0 images degraded using our physically motivated forward model, AstraMoE-SR is the only evaluated method to outperform the no-restoration baseline across every fidelity metric, improving on StableSR by 0.64 dB PSNR, 15.2% LPIPS, and 0.091 DINO feature similarity. Reconstructions conditioned on predicted trajectories differ negligibly from those using ground-truth trajectories, indicating that the estimates retain the degradation information required for effective restoration.