Peer Revenue Play credits. Real papers.

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 8 September

Artificial Intelligence

Harness or Model? Isolating the Harness Effect in Agentic Coding with a Contamination-Controlled Private Suite

An agentic coding system couples a language model to a harness: the tools, prompts and control flow that turn a chat model into an autonomous software engineer. Vendors ship harnesses tuned to their own models, and practitioners assume the vendor-native pairing solves more tasks. We measure that assumption with paired same-model contrasts on a private, contamination-controlled suite of 256 repository and post-cutoff contest tasks. The same 80 tasks ran under claude-agent-sdk and under deepagents on claude-opus-4-8, and under the openai-codex SDK and deepagents on gpt-5.5, with gemini-3.5-flash and deepseek-v3.2 as side cells. 792 of 800 planned runs were graded by an isolated oracle. Neither contrast resolves an average advantage for either harness: -1.25 pp for Opus 4.8 (48.8% vs 50.0%, task-bootstrap 95% CI [-10.0, +7.5]) and +1.25 pp for GPT-5.5 (55.6% vs 54.4%, CI [-4.4, +6.9]). The Opus average combines opposite strata: the native harness trails by 9.0 pp on the 61 repository tasks and leads by 23.7 pp on the 19 contest tasks (label-permutation p = 0.003). The partition was chosen after seeing the data and needs a designed replication. Correctness and completion also separate: 22 of 81 runs cancelled at the wall-clock ceiling had produced a passing patch. Re-priced from raw per-turn usage at frozen list prices, the neutral harness cost 1.3 to 1.6 times as much per solved task on Opus 4.8 and 1.2 times on GPT-5.5. These are observed-usage estimates. On the Anthropic account 58 runs left no usage record, and allocating that spend to either cell would move the Opus ratio between 0.7 and 2.3, so the billed ordering is unresolved. This revision corrects an August 2026 manuscript whose cost figures rested on a usage-semantics defect in our own telemetry (Section 5.1). We release the orchestrator, grading oracle, reanalysis code and derived aggregates. The tasks stay private.
Read the full abstract

Sound

Unlabeled Echoes: Pseudo-Labels and Genus-Aware Smoothing for Bat Call Recognition

Passive acoustic monitoring produces far more bat recordings than experts can label. We show that simple model-generated pseudo-labels turn this surplus into effective supervision. We compare pseudo-labeling with other semi-supervised learning methods on an 18-species European corpus using only 10% of its training labels, then transfer the strongest approaches to South African field audio containing nine bat taxa and a nuisance class. Pseudo-labeling outperforms the other semi-supervised learning methods on every European measure, recovering up to 61.5% of the gap to full supervision. It transfers to field audio with gains of 10.69 points in species accuracy and 4.96 points in species macro-F1. We also introduce genus-aware smoothing, which directs uncertain target mass toward congeneric species. Combined with uniform smoothing, it reaches 79.16 species macro-F1, 4.73 points above hard targets. Simple pseudo-labels are therefore highly effective at this ecological data scale, while genus-aware targets inject useful biological structure at no annotation cost. https://code4conservation.github.io/UnlabeledEchoes/
Read the full abstract

Mesoscale and Nanoscale Physics

Multiple Majorana zero modes realization based on superconducting topological crystalline metal ZrRuAs

The symmetry-protected multiple Majorana zero modes (MZMs) can be manipulated under external fields and have emerged as a promising pathway toward realizing topological quantum computing. While the suitable materials hosting multiple MZMs are still scarce, we propose a feasible candidate platform named superconducting topological crystalline metals (STCMs) that simultaneously possess symmetry-protected topological bands and intrinsic superconductivity. Model analyses demonstrate that the interplay among s-wave superconductivity, mirror symmetry-protected multiple surface Dirac cones, and the introduced spin splitting leads to high BdG Chern numbers of $\mathcal{N} = \pm C_M$, where $C_M$ is mirror Chern number of the STCM. First-principles calculations identify the experimentally synthesized superconductor ZrRuAs as a promising candidate with $C_M=2$, hosting two symmetry-protected surface Dirac cones. When integrated into a heterostructure with the ferromagnetic insulator (FMI) such as GdI$_{2}$, a topological superconducting phase with $\mathcal{N} = -2$ can be realized, giving rise to two branches of MZMs. This new scheme offers advantages of structural simplicity and tunability, making the FMI/STCM heterostructure an ideal platform for investigating multipole MZMs and novel topological qubit.
Read the full abstract

Combinatorics

Threshold Graphs Allow Few Distinct Eigenvalues: A New Approach

For any graph $G$, we associate a family of real symmetric matrices, $S(G)$, where for any $A \in S(G)$, the location of the nonzero off-diagonal entries of $A$ are governed by the adjacency structure of $G$. Let $q(G)$ represent the minimum number of distinct eigenvalues over all matrices in $S(G)$. In this work, we provide an alternative technique to establish that $q(G) \leq 4$ for any threshold graph $G$ as presented in [L. Emilio Allem, C. Hoppen, J. Lazzarin, L. Siviero Sibemberg, F. Colman Tura, The minimum number of distinct eigenvalues of a threshold graph is at most 4, Linear Algebra and its Applications, 726 (2025) 32 to 53]. In addition, we show that all connected threshold graphs admit a matrix having any four distinct eigenvalues. Further
Read the full abstract

Solar and Stellar Astrophysics

On modelling the 2017 galactic cosmic ray depression

In the second half of 2017, as Solar Cycle 24 approached its minimum, a pronounced and rigidity-dependent depression in the galactic cosmic ray (GCR) proton flux was observed using AMS-02. This event was driven by a sequence of strong coronal mass ejections (CMEs). In this work, we model the modulation of GCR protons during this period using a steady-state Parker transport framework with parameters constrained by AMS-02 observations. Treating the event as a purely long-term modulation interval yields excellent agreement with the observed proton intensities, but implies an inferred rigidity dependence of particle mean free paths that is inconsistent with basic turbulence theory, corresponding to an effective inversion of the scattering regime. We argue that this behaviour arises from the ability of highly flexible transport parameterizations to reproduce integrated modulation signatures through correlated adjustments of diffusion coefficients during transient events. To address this inconsistency, we relax the commonly adopted assumption of a time-independent ratio of perpendicular to parallel diffusion coefficients and introduce a physically motivated reduction of this ratio during the CME-dominated interval. This modification preserves the quality of the spectral fits whilst restoring the expected rigidity ordering of mean free paths. Our results indicate that allowing the perpendicular-to-parallel diffusion ratio to vary in time provides an effective means of constraining transport solutions during the 2017 depression and may help to resolve similar inconsistencies reported in other GCR modulation studies.
Read the full abstract

All authors O.P.M. Aslam, D. MacTaggart, M.S. Potgieter, M.D. Ngobeni, I.I. Ramokgaba

High Energy Physics (Theory)

Real-Time Instanton Dynamics in a Triple-Well to Double-Well Homotopy

We show that the singularities which the analytically continued double-well instanton develops in real (Minkowski) time are not an inherent feature of real-time tunneling solutions. Rather, they can be viewed as arising from a singular limit of a one-parameter family of regular real-time solutions. Using a linear homotopy between a triple-well and a double-well potential, we construct a family of potentials, parametrized by $p\in(0,1)$, whose real-time ($\alpha=0$) instanton solutions are everywhere regular and bounded. For any Wick rotation angle $\alpha\in(0,\pi/2)$, we prove that regularity is generic up to a measure-zero set. There exists a countably infinite, closed-form family of potential parameters $\{p^{\mathrm{singular}}_{k}(\alpha)\}_{k\geq0}$ at which the instanton becomes singular at exactly two points of the rotated time axis, never more. In the strict double-well limit, $p\to 1$ taken at $\alpha=0$, the entire family collapses onto the single point $(p,\alpha) = (1,0)$ at which the singular pairs of all $k$ occur simultaneously. This phenomenon produces the infinite singular comb found by Cherman and Ünsal. By solving the complexified equations of motion in terms of Weierstrass elliptic functions, we trace this comb to classical imaginary turning points receding to $\pm\mathrm{i}\infty$, giving a unified geometric and algebraic account of when, and exactly how often, real-time instantons become singular. We further show that the action of the continued instanton is independent of $\alpha$ only in its real part: its imaginary part jumps by one real-time period action at each singular angle, the $\alpha$-independence found by Cherman and Ünsal being the degenerate case of the double well, in which no such jump occurs.
Read the full abstract

General Physics

Bound states in the continuum of gravitational waves

Bound states in the continuum (BICs) are ubiquitous wave phenomena, but have not yet been demonstrated for gravitational waves (GWs). Here, in-plane periodic perturbations, exponentially localized at the plane $z = 0$, are shown to lead to distributional surface energy tensors at this plane and to be regular vacuum solutions ($T_{\mu \nu} = 0$) of the linearized Einstein field equations outside of it. These are achieved by explicitly calculating the Ricci tensor components and the Ricci scalar from the metric perturbation tensor. To fulfill each vacuum solution ($R_{{\sigma \nu}_{(+, \times)}} = 0$ and $R_{{}_{(+, \times)}} = 0$), different surface polariton-like dispersions are required. These bound perturbations decay exponentially to a flat metric ($h_{{BIC}_{(+,\times)}} \propto e^{- k_z |z|}$), and each localized metric has a correspondence to a different planar GW polarization ($+,\times$). The Lorenz gauge-fulfilling solutions exist at the $\Gamma$ point in momentum space, dwelling within the continuum of wavevectors of propagating GWs. The strains in the unit cell of the periodic perturbations have opposite parities relative to the corresponding planar GWs under a $C_2$ in-plane rotation, making them incompatible by symmetry with their propagating counterpart, indicating localization via symmetry protection.
Read the full abstract

General Relativity and Quantum Cosmology

Emergence of Gravity's Dynamical and Topological Sectors from Pre-geometry

We identify the complete set of fundamental building blocks for a 4D pre-geometric theory of gravity. Based on a gauge theory of $SO(1,4)$ or $SO(2,3)$ coupled to a Higgs-like field under the rigid constraint of general covariance in the unbroken phase, these building blocks - $\mathcal{L}_\text{MM}$, $\mathcal{L}_\text{W}$, $\mathcal{L}_{J}$, $\mathcal{L}_\theta$ and $\mathcal{L}_\vartheta$ - constitute the minimal generating set of independent field monomials from which any pre-geometric action, including arbitrary functionals thereof, can be constructed. While the most general pre-geometric theory can extend beyond linear combinations, these five irreducible invariants serve as the 'atomic' constituents of all possible pre-geometric dynamics. Upon spontaneous symmetry breaking, they collectively generate an emergent gravitational theory consisting of the Einstein-Hilbert action, the cosmological constant term and all 4D topological invariants: the Gauss-Bonnet, Pontryagin, Holst and Nieh-Yan terms. The unification of gravity's dynamical and topological sectors from a common pre-geometric source represents the central result of this work. We also uncover a see-saw mechanism linking the Planck mass and the cosmological constant, as well as several novel relations for the coupling constants of the topological sector, inclusive of the Barbero-Immirzi parameter. This framework establishes the pre-geometric foundations from which all aspects of gravitation can dynamically emerge, providing a unified starting point for quantum gravity, dark energy phenomenology and the study of topological phases in gravitational theories.
Read the full abstract

History and Overview of Mathematics

Some Early Results by Tutte Regarding the Cycle Double Cover Conjecture in 1948

OpenAI recently announced a proof of the Cycle Double Cover (CDC) Conjecture. Most media reports have characterized it as a 50-year-old open problem. In reality, according to a 1987 letter from Tutte to Fleischner, the Cycle Double Cover Problem has been open for at least 80 years. Two early results regarding the CDC conjecture were established in one of Tutte's 1949 publications.
Read the full abstract

Econometrics

Average Treatment Effect Localization: Projection Methods in Synthetic Control

Many real-world policies and business interventions require assessing short-term effects to inform timely decisions, even though most causal inference methods focus on long-term average treatment effects. In this paper, we introduce average treatment effect localization (ATEL), which captures localized, short-term policy impacts in panel data settings with a single treated unit and provides early indicators of policy impact. To accommodate both time-varying and nonlinear effects of observed and unobserved covariates, we propose a nonparametric model for untreated outcome, interpreted as a time-varying factor model via sieve approximation. Estimating the time-varying factor model is challenging due to the boundary bias and identification. Our estimation method based on diversified projection can effectively address these issues. We develop an asymptotic distribution theory to facilitate inference for the ATEL estimator. In an empirical application, we apply our proposed methodology to assess the impact of right-to-carry laws on violent crime rate.
Read the full abstract

Functional Analysis

A note on the Browder S-spectrum of a bounded quaternionic operator

Let $A$ be a bounded operator on a separable right quaternionic Hilbert space and let $\CA$ be the set of compact operators commuting with $A$. We establish the formula \[ \sigb(A)=\bigcap_{K\in\CA} \sS(A+K), \] a quaternionic analogue of the classical characterization of the complex Browder spectrum. The proof rests on a factorization lemma (Lemma~\ref{lem:fact}), a consequence of the Riesz–Schauder theorem and an ascent/descent argument, and on the characterization $\sigb(A)=\sS(A)\setminus\sd(A)$ in terms of S-eigenvalues of finite type, due to Arzini and Jaatit \cite{AJ26}, for which we give here an independent proof (Lemma~\ref{lem:A}). We also record two consequences of independent interest: the intersection defining $\sigb(A)$ may be restricted, without loss, to finite-rank operators commuting with $A$ (Corollary~\ref{cor:finite}); and, by a separate argument resting on the ascent/descent stability results of \cite{KD24}, the stronger pointwise statement $\sigb(A+K)=\sigb(A)$ holds for a fixed $K\in\CA$ and arbitrary $A\in\B(\V)$ (Theorem~\ref{thm:pointwise}), known in the complex case.
Read the full abstract

Econometrics

ACT, WAIT, or EXPERIMENT: A Causal Governance Framework for Retail Price Optimization Under Abstentions

This paper presents a causal decision framework for list-price elasticities in intermediated retail channels, where estimates are confounded by promotions, competitor moves and shopkeeper pass-through. Rather than forcing a number when the evidence is ambiguous, the system treats abstention (Wait) as an active diagnostic output. Built on Double Machine Learning and conformal prediction, it evaluates whether conditions exist to act on a recommended price or whether withholding is preferable, and it states when: acting beats withholding exactly when the calibration slope of the true headroom on the recommended move exceeds one half, a condition that predicts the sign of the gain in 37 of 38 synthetic groups. When the system abstains, it classifies the cause and flags which products are candidates for a designed pricing experiment. Aggregating estimates to the brand or category level, where independent price designs average out, cuts the root mean squared error from 0.488 to 0.159, although aggregation alone does not make the intervals honest. Tested on synthetic data with known truth and on public scanner data, the study reports its costs: the system abstains on 52% to 87% of presentations that were identifiable, and a single-window estimate forecasts real price changes poorly (calibration slope 0.22). What it offers is a way to audit any abstention rule and to publish at the unit the evidence supports; no commercial panel or pilot was used, so it does not measure commercial uplift.
Read the full abstract

Superconductivity

Generating is not discovering: a pre-registered physics judge for AI-proposed superconductors, calibrated on six known superconductors and one negative control

Generative models now propose millions of "stable" crystal structures, and their own referees have shown that stability is not discovery; no public benchmark asks whether a candidate has the physics to do a job. For superconductors the job is twofold: to pair, with a definite gap symmetry, and to become wire. We present a physics judge that answers both from a structure alone (DFT+U, a gated Wannier model, the full rank-4 RPA susceptibility on a $24^3$ mesh, the linearized gap equation, an irreducible-representation classifier, a manufacturability funnel) and measure what it says about what generators propose. With answer keys pre-registered before computing, the judge was calibrated on six known superconductors (Nb, Nb$_3$Sn, MgB$_2$, BaFe$_2$As$_2$, La$_{2-x}$Sr$_x$CuO$_4$, YBa$_2$Cu$_3$O$_7$), recovering the expected class in all six after two dated corrections. On the ruthenium analogue of the iron pnictide, with two runs and their criterion pre-registered the day before, it returns a null verdict at the same physical interaction where iron gives s$\pm$ ($\lambda_1 \le 0.0007$ versus 0.096; critical U 4.17 versus 0.97 eV); the pre-registered equal-alpha run fails the letter of the criterion, and we declare the resolution in favour of absolute U as post-hoc. The funnel reproduces the industrial map with thresholds fixed before running. Three generation audits follow: a descriptor sweep over 150 Materials Project metals re-finds the canon; 1,248 MatterGen structures yield 0 candidates that are new, stable and carry a pairing motif; mechanism filters over 47,893 compounds, validated by a blind hold-out, re-derive the community's analogies: geometry plus d-count is necessary but not sufficient. Generation re-finds what is known; judgment is the bottleneck. We propose a public benchmark over the eleven lists of AI-proposed superconductor candidates, none of which classifies gap symmetry.
Read the full abstract

Cryptography and Security

Understanding In-Context Multimodal Jailbreaks via Posterior Reweighting

In-context learning (ICL) jailbreaks reveal a critical vulnerability in multimodal large language models (MLLMs): harmful demonstrations in the prompt can induce unsafe outputs without modifying model parameters. Despite extensive empirical evidence, existing work lacks a principled understanding of why such jailbreaks reliably succeed or how their effectiveness scales with context composition. We propose a posterior reweighting framework that models a safety-aligned MLLM as implicitly operating over competing behavioral modes, and interprets in-context demonstrations as inference-time evidence that dynamically shifts the model's posterior preference between safe and harmful behaviors. This view formalizes jailbreak as a process of evidence accumulation, yielding predictive scaling laws with respect to demonstration count, harmful ratio, adversarial strength, and semantic diversity. Guided by this framework, we introduce a posterior-aware inference-time defense that adaptively injects benign counter-evidence based on estimated risk, effectively suppressing harmful posterior drift while preserving model utility. Compared to existing in-context defenses, our method achieves a significantly improved robustness-utility trade-off under a fixed intervention budget. Together, our results establish posterior reweighting as a unifying and predictive framework for understanding and mitigating ICL jailbreak in MLLMs.
Read the full abstract

All authors Xu Zhang, Dev Mistry, Xiang Xu, Ren Wang

Cryptography and Security

PEARL: A Task-Aware Framework for Evaluating Differentially Private Synthetic Educational Data

Personalized learning systems rely on real learner data, including performance, behavior, and demographic information, but these data are highly privacy-sensitive. Differentially private (DP) synthetic data can support system development and educational research while reducing exposure of individual learners. Existing evaluations, however, assess privacy and predictive usefulness separately, without determining whether synthetic learner data remain usable for the intended personalized learning task. We introduce PEARL (Privacy-Equivalence Audit and Release Ledger), which approves a DP synthetic educational dataset only when it passes all required checks of validity, privacy protection, predictive usefulness, and suitability for the intended educational task, while recording why each rejected dataset fails. Across 96 study settings, each defined by a dataset, data-generation method, privacy budget, and random seed, only 12 produced synthetic datasets that passed all applicable PEARL checks. Many privacy-protected datasets were rejected for omitting important outcome groups, such as withdrawn students, or for failing to preserve the order of learning activities. Fairness analysis further showed that some datasets passing privacy and predictive-usefulness checks still yielded unequal at-risk prediction performance across groups defined by disability and socioeconomic background. Moreover, Deep Knowledge Tracing and Self-Attentive Knowledge Tracing learned no meaningful next-response patterns from any tested synthetic knowledge-tracing dataset, showing that privacy protection alone does not guarantee usefulness for dropout prediction, knowledge tracing, or adaptive tutoring.
Read the full abstract

Cryptography and Security

Black-Box Membership Inference via Word-Level Probability Estimation

Membership inference attacks (MIAs) have emerged as critical tools for auditing privacy risks in large language models (LLMs), aiming to determine whether a given text was included in a model's training corpus. However, most existing MIAs require access to per-token logits or probabilities, making them inapplicable in practice to proprietary LLMs that expose only textual continuations. To address this underexplored setting, we propose Word-level Probability MIA (WPMIA), a statistically principled MIA for strict black-box privacy auditing. WPMIA estimates word-level generation probabilities via Monte Carlo sampling with local kernel smoothing, then aggregates these estimates into a sequence-level likelihood estimator. Furthermore, WPMIA constructs the likelihood conditioned on different prefixes, thereby amplifying the distributional differences between members and non-members. We evaluate WPMIA across various open-source LLMs and find that it consistently outperforms existing black-box baselines. Importantly, we also evaluate WPMIA on modern proprietary LLMs, including GPT-5-Chat, Gemini-2.5-Flash, and Claude-4.5-Haiku, achieving an average TPR@5%FPR of 42.0 across these models. These results offer a sound foundation for future research on strict black-box membership inference. Code is available at \href{https://github.com/niusj03/WPMIA}{https://github.com/niusj03/WPMIA}.
Read the full abstract

Software Engineering

On the Relation between Code Quality and Machine Learning Performance: A Large-scale Empirical Study

Context: Computational notebooks are the standard environment for machine learning (ML) development. Within the ML community, model performance is often the primary considered metric, and code quality is treated as a secondary concern. This prioritization relies on a largely untested assumption that code quality and ML performance are unrelated. Practitioners also reuse existing code that may come from notebooks selected through social signals (popularity, author expertise) whose reliability as quality proxies has never been assessed. Objective: We empirically investigated the relationship between code quality and ML performance in notebooks, and evaluated whether popularity and author expertise give indication on code quality or performance. Method: We conducted a large-scale empirical study of 265,363 Python notebooks submitted to Kaggle competitions. We assessed code quality with two static analysis tools: Pylint, capturing general Python code quality, and SonarQube, configured with a profile of 34 rules targeting data-science and ML-specific practices. Results: The relationship between code quality and performance depends on the notion of quality considered. General Python code quality is decoupled from ML performance, showing negligible or non-significant correlations across all observations. In contrast, ML-specific violations exhibit a consistent, small negative association with performance that persists across all observations. The popularity of a notebook does not give information on the code quality or performance. Code expertise provides no information on quality or performance, but competition expertise correlates with better performance, fewer ML-specific violations, and slightly more Python errors and refactoring violations.
Read the full abstract

Combinatorics

Real stability of layer-refined antichain polynomials for three-chain products with a two-element factor

For all positive integers $n,k$, we prove that the layer-refined antichain polynomial of the product poset $[2]\times[n]\times[k]$ is real stable. Jacobi-polynomial interlacing further shows that its diagonal specialization, the ordinary antichain polynomial of the same poset, has only simple, strictly negative zeros. For the special family $[2]\times[m]\times[m+1]$, explicit reciprocal identities give palindromicity; reciprocal pairing of the simple negative zeros then shows that every coefficient in the gamma expansion is strictly positive. Thus we prove Conjecture 4.3 of Ding and Dong and resolve all parts of their Conjecture 4.5, while strengthening its stated gamma-positivity consequence. The enumerative input is an explicit first-crossing reflection for two lattice paths, specialized from work of Krattenthaler and Sulanke.
Read the full abstract

Cryptography and Security

Adaptive Diffusion Freezing: Privacy-preserving Diffusion Models Against Membership Inference Attacks

Diffusion models have achieved remarkable success in generative tasks across various areas, however their training process raises significant privacy concerns, particularly under membership inference attacks (MIAs). Prior studies on privacy-preserving of diffusion models fail to balance privacy, utility, and efficiency. To address this gap, we propose a novel framework of privacy-preserving diffusion models, Adaptive Diffusion Freezing (ADF), which can defend against MIAs with better trade-off. By leveraging cross-timestep adaptive freezing training, ADF explicitly control the participation of different data subsets across diffusion timesteps via a mask matrix, which reduces the over-memorization and leads to more uniform model behaviors between member and nonmember samples. To construct a freezing mask matrix that effectively reduce membership leakage without unnecessarily harming generation quality, we introduce a pretraining-based risk-aware freezing policy to estimate MIA risk based on memorization tendency, and suppress the contribution of the subset-timestep pairs with higher risk. Evaluations on multiple datasets demonstrate that ADF provides effective defense performance as well as state-of-the-art privacy-utility-efficiency trade-off performance compared to various baselines.
Read the full abstract

Probability

Gumbel convergence for maximal packing distances from uniform random samples

Let $(X _n )_n$ be a sequence of i.i.d. $R^d$-valued random variables uniformly distributed on a compact subset $M$ of $R^d$ . In this work, we study the asymptotic behavior of maximal packings of this compact subset $M$. Under some assumptions on $M$ (in particular, that it is a smooth submanifold of $R^d$ without boundary), we establish a precise extreme value law (of Gumbel type) for the minimal distances from the sample $(X_1 , \cdots, X_ n)$ to the centers of maximal packings. This result leads to explicit asymptotic confidence bounds for this maximal packing and thus to the support $M$. A distinctive feature of our contribution is the explicit derivation of the scaling sequences in the Gumbel convergence, depending only on the geometry of the support M. These formulas provide interpretable confidence bounds for $M$, which represent a novel complement to previous approaches such as those by Fasy et al. (2014). Our approach bridges geometric and combinatorial probability arguments and relies on analytic tools such as the Lambert $W$ function. In addition, we provide examples of main sample spaces in directional and circular statistics (circle, sphere, torus), along with simulations that illustrate and support the theoretical findings.
Read the full abstract

Quantum Physics

When Measurement Constraints Favor Quantum Computational Sensing for Stealthy Power-Grid Attack Detection

Power-grid defenses that rely on digital telemetry remain vulnerable to stealthy attacks that preserve plausible reported states while altering the underlying physical system. We study when Nitrogen-Vacancy (NV) sensing provides a useful independent physical channel, and when coherent processing before measurement adds value. Across IEEE 14-, 30-, and 118-bus simulations with Lindblad NV models, we evaluate standard FDIA, BDD-stealth, statistical-stealth, and concealed topology attacks. The results reveal an observability hierarchy: evidence shifts from digital telemetry, to reported-versus-physical consistency, to the physical NV state. Quantum Computational Sensing (QCS) follows this selectivity, becoming informative only when the physical state itself carries attack evidence. We then compare QCS, conventional 4-setting NV readout, and tomography under matched measurement budgets. At a matched total budget of only 40 physical trials per sensor on case14, QCS reaches AP $0.944$, versus 0.800 for conventional NV readout and 0.751 for tomography; the same low-budget ordering holds on case30 and case118. Multi-setting methods recover as additional measurements become affordable, showing that the QCS benefit is a measurement-efficiency advantage rather than a universal accuracy advantage. Finally, we ask where the benefit varies across the three case simulations. Although the quantum-to-classical Fisher-information ratio increases from 1.21x to 1.54x, Normal–Attack Helstrom separation collapses in the harder regimes, and interleaved control substantially increases that separation only on case14, thereby quantum sensitivity does not necessarily imply task-relevant distinguishability. Our results show that realized QCS utility depends on the full chain from physical perturbation to state separation, coherent processing, and measurement under the resource constraints of the task.
Read the full abstract

All authors Saisubramaniam Gopalakrishnan, Supreeth B S, Pranav Sarda, Ashesh Xalxo, Dagnachew Birru

Analysis of PDEs

Confined elastic wires of large length: the sharp second-order asymptotics

We study the least bending energy $m_L$ of a closed embedded wire of prescribed length $L$ confined to the closed unit disk. It is classical that the bending energy of any closed curve in the disk is at least its length, with equality only for multiply covered unit circles. Such a circle is the energy limit of a stack of disjoint embedded circles, but not of a single embedded loop, and the size of the resulting defect $m_L-L$ has remained open. We prove that \[ m_L \;=\; L+c_1\,L^{4/9}+O\bigl(L^{1/3}\bigr)\qquad\text{as }L\to\infty , \] with a sharp constant $c_1=5.223049\ldots$ given in closed form by two one-dimensional model problems. In particular $m_L-L\asymp L^{4/9}$, which improves the previously known upper bound of order $\sqrt{L}$ and provides the first lower bound beyond the trivial one. This paper was produced with substantial assistance from large language models; Section 7 sets out in detail what they contributed and what I verified.
Read the full abstract

Popular Physics

The Epistemic Risk of the 31st Spore: If Planets Aren't Fine Tuned, We're Doing Life Detection Wrong

Exploration of planetary bodies within our solar system will be essential for understanding the origin of life on Earth and the distribution of life in the universe. Planetary protection policy is concerned with balancing this desire for exploration against the risks of contaminating alien planets with Earth life, and contaminating Earth with alien life. However, at present, we have no fundamental scientific understanding of life's emergence or its nature beyond Earth. Given this nearly complete ignorance about the possibility of alien life, or Earth life's capacity to expand beyond our planet, it is difficult to reason about the real risks of space exploration. Here we contend that contemporary understandings of the risks of forward contamination are based on arguments which inconsistently apply our incomplete knowledge to the problem. We reason that a more assertive posture towards space exploration, focused on determining whether other planets in the solar system are inhabited, is warranted and explain why such a posture may not increase the epistemic risks of planetary contamination. Finally, we explore the consequences of our arguments for planetary protection protocols, and life detection efforts.
Read the full abstract

Econometrics

When is statistical evidence strong enough? Using hypothesis tests to value data collection

Policy decisions often hinge on conventional p-value thresholds, which ignore economic costs and benefits of further data collection. This paper recasts statistical significance as a choice between making an immediate policy recommendation and deferring it until further evidence is collected. The welfare-optimal decision corresponds, under minimax regret, to a statistical test whose level depends on the cost and precision of additional evidence. Inverting this rule, we introduce and recommend reporting the abstention-value (A-value) to determine where additional data collection is most needed. The A-value defines the break-even welfare cost of abstaining and recommending further experimentation given the initial evidence. Computing the A-value only requires a point estimate and its standard error, imposes no prior assumptions on policy effects, and can be converted to a monetary research budget using inputs already frequently used by regulators and funding agencies. When experimentation capacity is limited, we show that prioritizing additional data collection where A-values are the largest yields strong finite-sample guarantees. Applications to anti-poverty programs illustrate the practical benefits of using A-values.
Read the full abstract

Astrophysics of Galaxies

Phase-Spirals Across Galactic Disks II: Using large-scale "macro-spirals" in phase-spiral amplitude to derive perturbation times

Phase-space spirals in the Milky Way disk are a key observable remnant of recent perturbations to the Galaxy. They provide insight into the disk's potential, dynamical evolution, past interactions, and even substructure. However, the complex dynamics of phase-spiral formation and evolution have made clear interpretations challenging. For example, recent work has shown that the "winding time" – measured from how tightly wound a phase-spiral is – is a biased estimate of the true time since the inciting perturbation due to the complex effects of self-gravity. In this paper series, we present an alternative approach by looking at correlations in phase-spiral morphology across the Galactic disk. Here we show that following a localized perturbative event, the ridgeline connecting the largest amplitude phase-spiral at each radius winds up into a "macro-spiral" at the rate expected for differential rotation. This means the macro-spiral can 1) be unwound to constrain the origin – time and location – of correlated phase-spiral properties and 2) indicate the delay in individual phase-spiral winding as a novel diagnostic of disk dynamical properties. Applying these ideas to the Milky Way's phase-spirals, we estimate a perturbation time of $\simeq 1$ Gyr ago, consistent with the penultimate passage of the Sagittarius dwarf galaxy through the disk, and delay times of up to 800 Myr for phase-spirals in the inner disk. While the current application of this method to Gaia DR3 data is limited by the available radial velocities, future Gaia data releases will enable stronger constraints using a much larger area of the Galactic disk.
Read the full abstract

All authors Kiyan Tavangar, Kathryn V. Johnston, Jason A.S. Hunt, Axel Widmark, Vandana G. Kaushik, Adrian M. Price-Whelan

Number Theory

Integer group determinants for the elementary abelian group of order 49

Let $G=C_7\times C_7$, and let $S(G)$ denote the set of integer values of its group determinant. We determine $S(G)$ by classifying the values divisible by $7$; the coprime values are already known. Every nonzero divisible value has valuation at least $10$, and every multiple of $7^{12}$ occurs. At valuations $10$ and $11$, we give necessary and sufficient conditions on the cofactor in terms of two additive invariants of ideals in $\mathbb{Z}[\zeta_7]$. The first condition is a bounded signed sum of prime-ideal invariants. The second requires a prime ideal with nonzero invariant pair whose norm divides the cofactor. Neither cofactor set is a union of congruence classes modulo any positive integer. The least positive divisible value is $43\cdot7^{10}$, and the least positive value of valuation $11$ is $8\cdot7^{11}$. The proof combines integral reconstruction from character values with a calculation of the global-unit image modulo $7$. We conclude by identifying the additional local conditions and realization problems that arise at primes at least $11$.
Read the full abstract

Optimization and Control

A Function-Space Framework for BDDC Preconditioning in Control- and State-Constrained Sparse Optimal Control

We develop a Balancing Domain Decomposition by Constraints (BDDC) preconditioner for the interface systems arising from active-set semi-smooth Newton linearizations of elliptic optimal control problems with box-constrained controls, $L^1$-sparsity, and Moreau–Yosida regularized state constraints. The domain decomposition and BDDC construction are formulated directly at the infinite-dimensional level. Eliminating the interior variables yields a Schur complement for the state and adjoint traces assembled from local subdomain operators. We prove a global–local equivalence result, establish well-posedness of the local and partially assembled problems under verifiable conditions on the primal constraints. The BDDC operator admits a two-level additive Schwarz representation with independent local solves and a finite-dimensional coarse solve. We implement the preconditioner using conforming finite elements and test it for distributed and Neumann boundary control. For a fixed ratio of subdomain diameter $H$ to mesh size $h$, the GMRES iteration counts are nearly independent of mesh refinement and grow only moderately as this ratio increases. Enriching the coarse space substantially improves convergence and reduces sensitivity to the control regularization and sparsity weights. For the state-constrained problem, scaling the Moreau–Yosida penalty parameter proportionally to $h^2$ yields nearly constant GMRES iteration counts as the mesh is refined.
Read the full abstract

Probability

Second-order fields for stochastic partial differential equations

In this article, we study the second-order fluctuation of the solutions of one-dimensional polynomial stochastic partial differential equations (SPDEs) of the form \begin{equation*} (\partial_t - \Delta) \Phi_{\varepsilon} = -P(\Phi_{\varepsilon}) + \xi_{\varepsilon}, \end{equation*} where $P$ is a polynomial of degree greater than or equal to $2$, $\xi_\varepsilon$ is the white-noise after being convoluted (in space) by the heat kernel $K_\varepsilon = e^{\varepsilon \Delta}$. More precisely, taking advantage of the local solutions of pointwise well-posedness of limits $\Phi= \lim_{\varepsilon \to 0}\Phi_{\varepsilon}$, we characterise the limit of $\Phi^{err}=\lim_{\varepsilon \to 0} \varepsilon^{-1}(\Phi_{\varepsilon}-\Phi)$ as the solution of a more irregular stochastic partial differential equation. This is performed by applying the Da Prato–Debussche decomposition in the non-linear equation and characterising the limit of each of the terms. We also discuss possible characterisations of second-order fluctuations performed for higher-dimensional SPDEs in the weakly-coupled regime.
Read the full abstract

High Energy Astrophysical Phenomena

First Detection of Radio Polarization During Jet Formation in the Changing-Look AGN 1ES 1927+654

We present a multiwavelength radio study of the changing-look AGN 1ES 1927+654 following its renewed X-ray brightening since mid-2022, combining VLBA imaging with multi-band and time-resolved VLA observations. Our main results are: (i) VLBA observations continue to reveal the bipolar jet structure first reported in Meyer et al. (2025), with new X-band detections confirming the morphology at larger separation and K-band imaging revealing a bridge of emission consistent with a continuous outflow; (ii) the VLBA core exhibits a GHz-peaked spectrum, and VLA/VLITE observations from 340 MHz to 45 GHz show emission from an inhomogeneous synchrotron source, self-absorbed below ~3 GHz and optically thin above; (iii) VLA and VLBA flux densities are consistent above 8 GHz, indicating emission dominated by a compact core; (iv) linear polarization is detected above ~3 GHz, with fractional polarization increasing with frequency from ~0.2% to ~6% and EVPAs exhibiting non-linear lambda^2 behavior; (v) long-term VLA X-band monitoring reveals a large, smooth rotation of the EVPA of ~137 degrees over ~81 days in 2026, followed by a partial return, at stable total intensity and suppressed fractional polarization <~1.4%, consistent with a propagating disturbance crossing the compact emitting region; and (vi) circular polarization is not detected at any epoch. These observations demonstrate that 1ES 1927+654 has transitioned from an X-ray-dominated, non-jetted state to a radio-loud AGN (Laha et al. 2025) hosting a newly launched relativistic jet, a unique laboratory for studying jet birth and early magnetic field evolution.
Read the full abstract

All authors Onic I. Shuvo, Eileen T. Meyer, Sibasish Laha, Agniva Roychowdhury, Dev R. Sadaula, Tracy E. Clarke, Wendy M. Peters, Mitchell C. Begelman, Markos Georganopoulos, Rostom Mbarek, Amelia M. Hankla, Alexander Philippov, Navin Sridhar

Statistics Theory

Central limit theorems and bootstrap for sparse regularized multimarginal optimal transport and barycenters

Wasserstein barycenters are a fundamental tool for the analysis of distribution-valued data, but their computation and statistical analysis become challenging in high-dimensional and multi-sample settings. The unregularized barycenter problem is computationally demanding and suffers severely from the curse of dimensionality, making regularization essential in many applications. While regularized optimal transport methods are widely used, approaches based on entropic penalization tend to overspread mass and may blur the geometric structure of the solution. Motivated by barycenter estimation, we propose a more general methodology based on sparse regularized multimarginal optimal transport (RMOT). Our formulation creates sparse transport plans and is designed to avoid this overspreading effect. We establish consistency, central limit theorems, and bootstrap validity for the optimal potentials and couplings for the empirical RMOT problem. Specializing to sparse regularized barycenters, our results yield a uniform weak limit and bootstrap consistency. Numerical simulations show that sparsity plays a fundamental role in preserving the structure of the solution and avoiding the overspreading behavior typical of entropic regularization.
Read the full abstract