Peer Revenue Play credits. Real papers.

Papers

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

All of arXiv

Thursday 10 September

Statistical Computation

Extended One-Liners for the Gamma, Poisson, and Binomial Distributions

We demonstrate explicit transformations of three independent uniforms with exact Gamma$(a,1)$, $a>0$, Poisson$(\lambda)$, $\lambda>0$, and Binomial$(N,p)$, $N\geq 1$, $0\leq p \leq 1$ laws using only elementary operations.

Machine Learning

Particle GFlowNets: Rethinking Generative Marginalization Models

Generative Marginalization Models (MaMs) have been recently introduced as efficient neural sampling models for any-order autoregressive modelling of discrete distributions. By learning both the marginal and conditional probabilities of a persistent-block Gibbs sampler, MaMs enable fast posterior evaluation with a single neural network forward pass. While prior work has considered MaMs to be distinct from Generative Flow Networks (GFlowNets), a well-established paradigm for inference in discrete stochastic models, we show that they are equivalent. Then, we also extend MaMs' sampling strategy to non-autoregressive generative processes. In particular, we describe an automatic criterion for full-state rejuvenation of the Gibbs sampler, derived from the Gelman-Rubin statistic, which plays a key role in speeding up learning convergence. Our experiments show that our method, called Particle GFlowNets, markedly accelerates training in large combinatorial spaces.
Read the full abstract

Classical Analysis and ODEs

Shorter proof of dimension-free $L^p$ estimates for maximal Riesz transforms

We provide a shorter and more direct proof of $L^p$ estimates for maximal Riesz transforms (of an arbitrary order) in terms of the corresponding Riesz transforms, with a constant independent of the dimension of the Euclidean space $\mathbb R^d$. This result was originally proved by Mateu, Orobitg, Pérez and Verdera with a constant depending on the dimension, and improved to a dimension-free inequality by Kucharski, Wróbel and Zienkiewicz.
Read the full abstract

Quantum Physics

Impossible to conjugate or reflect an unknown quantum state via a unitary evolution

The no-cloning theorem, the no-deleting theorem, and other no-go theorems play a crucial role in quantum mechanics and quantum information. In this paper, we show that it is impossible to conjugate or reflect an unknown quantum state or conjugate or reflect the bit-flip (the phase-flip) of an unknown state via a universally unitary evolution. We also indicate that the partial conjugating operations are not anti-unitary and it is impossible to universally partially conjugate an unknown state via a unitary evolution. We compute the fidelities of the optimal physical approximations to conjugating and conjugating the bit-flip and phase-flip operations.
Read the full abstract

Cryptography and Security

On Identifying Sound Conditions for Frontrunning Resistance

Blockchains enable decentralized applications through smart contracts—interactive programs executed through consensus. However, the inherently asynchronous nature of blockchain transaction ordering introduces a class of vulnerabilities known as frontrunning attacks, which have caused millions of dollars in losses in major blockchains, such as Ethereum. Frontrunning attacks arise because users interact with smart contracts through transactions, which are added to the blockchain by designated nodes called miners. Miners can exploit their ability to reorder, delay, or insert transactions to gain an advantage over honest users, effectively frontrunning them. Yet, to date, the field lacks a rigorous definition of what it even means for a contract to resist such attacks. Worse, we show that existing dynamic detection approaches are fundamentally inadequate: in a large-scale study comprising 287 smart contract audits, 55% of the 393 reported vulnerabilities identified by leading smart contract auditors fall outside the scope of state-of-the-art detection criteria. To address this gap, we propose the first formal definition of frontrunning vulnerability for smart contracts. Our definition captures a key insight: resistance to frontrunning is not an intrinsic property of a contract alone, but depends critically on how honest users interact with it. Grounded in this observation, we develop a sound algorithm for synthesizing secure interaction conditions, alongside a prototype implementation that we apply to audited real-world contracts—revealing previously undiscovered vulnerabilities in two Ethereum contracts.
Read the full abstract

All authors Sebastian Holler, Anna Piscitelli, Jannik Albrecht, Stephan Dübler, Ghassan Karame, Clara Schneidewind

Combinatorics

Kriesell's conjecture for infinite graphs

Let $G$ be a graph and $S\subseteq V(G)$ be a subset of vertices. An $S$-Steiner tree $T$ of $G$ is a tree of $G$ which contains $S$ in its vertex set $V(T)$. Kriesell conjectured that for every $2k$-edge-connected subset $S\subseteq V(G)$ in a finite connected graph $G$, there exist $k$ pairwise edge-disjoint $S$-Steiner trees. This conjecture is false for infinite graphs. We present a version of Kriesell's conjecture with topological $S$-Steiner trees for countable finitely edge-separable graphs and a version with $F$-limits of trees for rayless graphs. We show that if Kriesell's conjecture holds for finite graphs, then it holds for every connected, rayless and finitely edge-separable graph. We also show that every $2k$-edge-connected rayless and finitely edge-separable graph contains $k$ pairwise edge-disjoint spanning trees.
Read the full abstract

Algebraic Geometry

Campana Rational Connectedness and Weak Approximation of Del Pezzo Orbifolds

We prove that a del Pezzo orbifold of degree less than 7, and del Pezzo orbifolds of higher degree with irreducible boundary are strongly Campana uniruled, and deduce that it is Campana rationally connected. This provides important non-toric examples in dimension two of a conjecture by Campana, in the form required by the weak approximation theorem, and yields weak approximation for Campana sections of Campana fibrations at all places whose general fibre is such an orbifold. In three appendices we classify the del Pezzo orbifolds of degree 6, 7 and 8 with irreducible boundary.
Read the full abstract

Artificial Intelligence

Prompt Revision as a Source of Cultural Bias in Text-to-Image Systems

Commercial text-to-image systems silently revise user prompts before generating images, a step users typically cannot disable or even see. Yet, existing audits of cultural bias examine only the final images and treat generation as a single pipeline, so they cannot tell where the bias originates. We introduce WORLDVIEW, a multilingual benchmark of 8,960 prompts across 15 languages and 31 language-context pairings. Using it, we audit the revision layer in three systems (DALL-E-3, Imagen-4, GPT-Image-1.5) through a three-step analysis of how heavily it marks each cultural context, whether it flattens that context into a narrow vocabulary, and whether that vocabulary is stereotypical. Relative to a no-context English baseline, the US is the least-marked context, while non-Western and non-Anglophone contexts are marked far more heavily, flattened into narrow vocabularies applied across topically diverse prompts, and reduced to recognizable cultural stereotypes. Comparing images from original versus revised prompts on models without a revision layer, we identify the layer itself as a previously undocumented, causal source of this stereotyping. To locate cultural bias, and fix it, we must audit the system as deployed, not the model alone.
Read the full abstract

All authors Aleksandra Urman, Elsa Lichtenegger, Salima Jaoua, Azza Bouleimen, Robin Forsberg, Corinna Hertweck, Stefania Ionescu, Nicolò Pagan, Ancsa Hannak, Joachim Baumann

Solar and Stellar Astrophysics

Evidence for plasmoid formation related to a hot UV burst in the solar atmosphere

Magnetic reconnection in the solar atmosphere drives energetic phenomena such as Ellerman bombs, UV bursts, and surges. We used coordinated observations from the Swedish 1-m Solar Telescope (SST), IRIS, SDO, and the Hinode/X-ray Telescope (XRT) to analyze an active region with strong magnetic flux emergence. This region displayed long duration UV burst activity, occurrence of multiple Ellerman bombs, and recurrent ejection of surges. Chromospheric imaging show sthe presence of a dome-like structure that overarched the region where magnetic flux was emerging. The curved chromospheric fibrils were closely associated with bright and similarly curved extensions from a UV burst in the IRIS 1400 A channel. We observed an episode of formation of a bright sheet along the curved fibrils in H-beta that subsequently fragmented into plasmoid-like blobs with peak emission at a Doppler offset of +43 km/s. IRIS Si IV lines in the same region show complex and non-Gaussian line profiles that display asymmetric extensions to high Doppler offsets. Some profiles have clear components with Doppler offsets in excess of 150 km/s on both sides of the nominal line center. The optically thin SDO/AIA channels, including 94 A, show intermittent and repeated occurrences of the burst. Emission measure and filter ratio analysis indicate that the burst is multi-thermal and can attain temperatures beyond 1 MK. AIA and Hinode/XRT observations reveal a localized EUV and soft-X-ray counterpart of the burst/dome system, supporting the presence of intrinsically multithermal plasma that reaches coronal, and likely multi-MK, temperatures.
Read the full abstract

All authors Mats Ola Sand, Luc Rouppe van der Voort, Souvik Bose, Daniel Nobrega-Siverio, Jonas Thoen Faber, Reetika Joshi, Ignasi Josep Soler Poquet

Neurons and Cognition

pyAvalanches: A Python Package for Analyzing Spatiotemporal Propagation in Neuronal Avalanches

The analysis of neuronal avalanches offers insights into brain dynamics utilizing the framework of criticality, but the reproducibility and comparability of studies are limited by the use of fragmented, lab-specific scripts. To address this issue, we introduce pyAvalanches, an open-source Python package providing a standardized, end-to-end pipeline for avalanche analysis from electrophysiological recordings (e.g., electroencephalography-EEG). Starting from the detection of neuronal avalanches the package provides their core statistical characterization, including size and duration distributions. Beyond this, the main aim of pyAvalanches is to characterize the spatiotemporal organization of activity propagation during avalanches. To this end, the core innovation of pyAvalanches is the compuation of Avalanche Transition Matrices (ATMs) to map spatiotemporal propagation patterns. Building on this, the package derives network-based metrics from the ATMs, bridging the study of the topology and organization of the underlying dynamical interactions with network neuroscience adopting the framework of neuronal avalanches. The entire workflow is encapsulated in a modular and scikit-learn compatible architecture. We demonstrate the utility of pyAvalanches through an illustrative group-level analysis on a public resting-state EEG dataset, comparing propagation patterns across different clinical populations. By providing a user-friendly, tested, and extensible tool, pyAvalanches facilitates reproducible research, enables the development of novel avalanche-based biomarkers, and makes complex avalanche analysis accessible to a broader scientific community. The package is fully documented and distributed via the Python Package Index (PyPI).
Read the full abstract

All authors M. Marzulli, M. Angiolelli, C. Mannino, M. Demuru, P. Sorrentino, M.-C. Corsi

Human-Computer Interaction

Ethics Training Agents: Facilitating Group-Based Ethics Education with Role-Playing and Discussion for Ethical Reflection and Exploration

Group-based ethics training for Science, Technology, Engineering and Mathematics (STEM) students is a complex challenge, requiring substantial resources and expertise. While activity-based teaching methods, such as role-playing and discussions, are commonly employed to simulate real-world scenarios, current practices are often manual and lack integration with effective online platforms for supporting group-based ethical discussions. In this work, we propose Ethics Training Agents, a group discussion system that leverages multiple LLM participants embodying distinct ethical orientations, along with a moderator agent, to enable structured human-AI group ethical discussions for collaborative reflection. We conduct a user study with 45 undergraduate STEM students to evaluate the learning outcomes and user experience. The results show that our system supports engagement, coordination, and perspective-taking in group discussions and has a positive influence on ethical sensitivity. We also discuss practical design strategies for integrating multiple LLM agents into multi-human group settings to facilitate ethics training for STEM students.
Read the full abstract

All authors Youngseok Seo, Sueun Jang, Hyesoo Park, Renz Samuel Gutierrez, Joseph Seering, Uichin Lee

Numerical Analysis

Mean Square Error Analysis of Stochastic Runge-Kutta Integrators

We analyze the mean square error of stochastic Runge–Kutta integrators for overdamped Langevin dynamics whose potential is strongly convex outside a bounded region. A decomposition splits the local error into a mean-zero term and a smaller remainder, and the discrete Poisson equation turns their moments into a bound on the error of a time average. We carry this out for two stochastic Runge–Kutta integrators proposed by Yang & Wang (2026), both of strong order $\frac32$ and weak order $2$, which evaluate the gradient of the potential and no higher derivative. We show that their laws approach the law of the exact solution at second order in Wasserstein-1 distance, up to a logarithm, uniformly in the number of steps. For a test function with bounded derivatives up to third order, we prove that the mean square error over $N$ steps with step size $h$ is $\mathcal O ( \frac{1}{Nh} + h^4 )$, which is the optimal order in the discretization. Experiments measure the strong and weak orders and the sampling bias on a nonconvex potential in $\mathbb R^2$, and compare the integrators on a diffusion model of CIFAR-10.
Read the full abstract

Artificial Intelligence

Lightweight LiDAR-Based Cone Detection Framework Using Random Forest for Formula Student Driverless

Reliable, low-latency perception is crucial for Formula Student Driverless vehicles, yet many existing pipelines rely on deep learning and multi-sensor fusion, often requiring GPU acceleration. This paper presents a lightweight LiDAR-only perception pipeline tailored for CPU execution, combining ground removal, IMU-based motion compensation, DBSCAN clustering, and geometric feature-based Random Forest classification. Feature importance analysis reduced the model input from 12 to 7 features while preserving performance. Evaluated on 2,371 labeled clusters collected from real FSD events, the pipeline achieves an F1-score of 98.33% and an end-to-end runtime of 3.13 ms on CPU-only hardware. The released dataset, labeling tool, and trained models provide a practical and reproducible baseline for other resource-constrained autonomous racing teams.
Read the full abstract

Nuclear Theory

Study of the multiplicity dependence of non-flow correlations in pp collisions at $\sqrt{s}=200$ GeV using PYTHIA

The observation of a near-side enhancement in long-range correlations within high-multiplicity pA and pp collisions, commonly referred to as the ridge, has motivated significant efforts to quantify possible collective effects in small systems. Estimation of non-flow contributions in two-particle correlation measurements in small systems at the LHC typically relies on the assumption that jet-induced non-flow correlations are independent of event multiplicity. Consequently, this assumption allows the high-multiplicity non-flow to be estimated and accounted for by directly using low-multiplicity events. In this paper, the validity of this assumption is studied examining the multiplicity dependence of non-flow contributions estimated via two-particle correlation functions in pp collisions at RHIC energies $\sqrt{s} = 200$~GeV using the PYTHIA event generator. It is observed that selecting any specific multiplicity range inherently biases jet production dynamics. These biases introduce a non-trivial multiplicity dependence of the non-flow contribution at RHIC energies. The dependence of jet fragmentation and $p_{\text{T}}$-differential cross sections on event multiplicity is studied using a variety of commonly used multiplicity estimators, including mid-rapidity charged-particle multiplicity, forward multiplicity, and underlying-event activity. The results indicate that non-flow contributions do not exhibit simple multiplicity scaling and challenge the applicability of standard subtraction procedures utilized at LHC energies at lower RHIC collision energies.
Read the full abstract

General Relativity and Quantum Cosmology

Signatures of charged rotating regular black holes: quasinormal modes, grey-body factors and shadows

We construct and study possible astrophysical signatures of the rotating counterparts of two charged regular black holes, namely the Ayón–Beato–García (ABG) and Balart–Panotopoulos–Rincón (BPR) spacetimes – obtained using the Newman-Janis algorithm. After verifying the regularity of the rotating metrics, we study massless scalar perturbations, compute the quasinormal mode (QNM) spectra, and characterise their dependence on the spin and charge parameters. Grey-body factors (GBFs) are computed, and the QNM-GBF correspondence is verified. The low-frequency superradiant amplification factor is also examined via matched asymptotic expansions. The Lyapunov exponents are used to connect the eikonal QNM damping rate to photon-orbit instability. Additionally, from shadow profiles, we constrain the parameter space of rotating regular black holes using the Event Horizon Telescope (EHT) observations. Across all the probes, we find that the rotating BPR and Kerr–Newman black holes behave almost indistinguishably from one another, while the rotating ABG geometry exhibits distinct perturbative, scattering, and geometric properties, particularly at larger values of spin and charge. Our analysis indicates that differentiating between charged rotating regular black hole models may be possible using gravitational wave and shadow observations.
Read the full abstract

Machine Learning (Statistics)

Risk-Averse Decision Making with Multi-Level Reliability Guarantees

Many applications in engineering, including wireless broadcasting, require designs that provide performance certificates at different target outage levels. This paper studies the problem of maximizing the weighted average of such certificates in the presence of uncertainty about the true system state. The problem is shown to be equivalent to an optimization over nested prediction sets, connecting to the literature on conformal prediction and extending prior art on single-level risk-averse decision making. Furthermore, we derive a dual formulation that decouples optimization across input values. Numerical experiments on a diversity-based wireless transmission system illustrate the cost of enforcing multi-level certificates with a single shared policy and trace the Pareto trade-off between multiple reliability levels.
Read the full abstract

Sound

Post-Training Zero-Shot TTS for Fine-Grained Emotion and Duration Control via Natural Language

Audiobook narration, conversational agents, and audiovisual dubbing require fine-grained control over emotion and pacing within an utterance, yet many text-to-speech (TTS) systems remain limited to utterance-level conditioning. We propose a unified post-training framework for natural-language segment-level emotion and duration control. Supervised fine-tuning (SFT) establishes instruction following, while group relative policy optimization (GRPO) further improves control accuracy using emotion, duration, content, and speaker rewards. Without modifying the pretrained architecture or adding inference-time control modules, our method improves fine-grained controllability while preserving intelligibility and speaker identity, demonstrating post-training as an effective way to extend pretrained TTS models.
Read the full abstract

Quantum Physics

Birth and Death of Entanglement in Hamiltonian-Driven Quantum Games under Decoherence

Quantum game theory investigates the influence of quantum resources on strategic decision-making. In this work, two-player quantum games based on the Transverse Field Ising Model(TFIM) are investigated under amplitude-damping decoherence. The TFIM Hamiltonian naturally produces a family of entangling gates, enabling a physically motivated implementation of quantum games. The effects of noise on Nash equilibria, players' payoffs, concurrence and coherence of the quantum states for different initial states and strategy pairs are analyzed. The results show that decoherence progressively suppresses quantum strategic advantages, with maximum damping driving all outcomes to identical classical payoffs. The concurrence and coherence analysis of the states generated in the quantum game reveal initial state and strategy dependent quantum correlation dynamics, including entanglement sudden birth and death.
Read the full abstract

Machine Learning

Generalised Score Matching on Convex Domains

Score matching avoids computing the normalising constant that maximum-likelihood estimation requires. On constrained domains, its generalised variants weight the Fisher divergence so that boundary terms vanish. We derive generalised score matching on open convex subsets of $\mathbb{R}^{d}$ as the small-neighbourhood limit of minimum probability flow, in which the geometry of the neighbourhoods determines the weight. Every $C^{2}$ positive definite weight arises in this way, including those of classical score matching on $\mathbb{R}^{d}$ and of its variants for non-negative data on $\mathbb{R}_{+}^{d}$. For exponential families, we extend the standard convexity, consistency and asymptotic normality results to every such weight and show that the estimator converges to the true parameter under certain boundary conditions. For a truncated Gaussian on a polytope and a Dirichlet distribution on the simplex, proposed estimators attain the lowest median error of all methods compared, in at least 42 of 50 ground-truth configurations.
Read the full abstract

Numerical Analysis

Approximate Inversion of Discrete Fourier Integral Operators via Hierarchically Semiseparable Matrices

This paper introduces a novel method for approximating the inverse of discrete Fourier integral operators (FIOs). Given an $N \times N$ matrix representation $K$ of an FIO, the proposed algorithm consists of two stages. In the offline stage, we first construct a butterfly factorization (BF) $\tilde{K}$ of $K$, which enables fast forward matrix-vector multiplication. We then construct a hierarchically semiseparable (HSS) approximation $\tilde{G} \approx G$, where $G = K^{*} K$, using fast applications of $\tilde{K}$ and $\tilde{K}^{*}$ to random matrices. Finally, we apply the ULV factorization to the HSS matrix $\tilde{G}$ to obtain an approximation $\tilde{F} \approx G^{-1}$. Combining these approximations yields an approximate inverse $K^{-1} \approx \tilde{F} \tilde{K}^{*}$. The offline stage has complexity $O(N \log^{2} N)$ for 1D problems and $O(N^{1.5} \log N)$ for 2D problems. In the online stage, the proposed method approximates $K^{-1} u$ for a given input vector $u$ with complexity $O(N \log N)$ for both 1D and 2D problems. The proposed method can be used either as a direct solver or as a preconditioner for iterative methods. Numerical results for 1D and 2D FIOs demonstrate the effectiveness of the proposed method.
Read the full abstract

Computer Vision

Learning Interaction between Image and Layout Priors for Joint Image-Layout Generation in Design Templates

In this paper, we address the problem of graphic design template creation, which generates a background image and a layout of foreground elements over the background to form a harmonious composition from an input text. Prior work on graphic design generation mostly adopts a sequential paradigm, where design elements are generated sequentially. We argue that such a sequential scheme falls short of faithfully capturing the dependency between the background and layout (and thus the joint image-layout distribution), which limits the quality of generated design templates. To overcome this limitation, we propose a model, InterIL, which jointly generates the two modalities, background image and layout, in a single generative process. The novel design of our joint model connects the backbones of pretrained image and layout diffusion models with a learnable communication module to explicitly model bidirectional image-layout interaction. During training, the image and layout backbones are frozen to maintain and leverage the vast pretrained single-modality prior knowledge, while only the communication module is updated, so that the model can focus on learning image-layout interaction and thereby better capture the joint image-layout distribution for improved composition harmony. Our model has no design-specific inductive bias, which allows it to better preserve the original characteristics of realistic designs. We further introduce a test-time guidance strategy to enable users to impose their specific preferences on generated results. Our experiments show that, compared with prior approaches, our model can generate significantly better results in terms of image, layout and image-layout harmonization, producing outputs closer to real samples. We also demonstrate the flexibility of our model in enforcing user preferences at inference without retraining.
Read the full abstract

Neural and Evolutionary Computing

Breaking the Central Bias: Spatially Partitioned Experts for Coordinate-Based Neuroevolution

Evolvable-Substrate HyperNEAT (ES-HyperNEAT), a bio-inspired indirect encoding that determines neuron placement and connection weights from spatial coordinates, exhibits a failure mode on MNIST as a diagnostic benchmark. Because input pixels map to a coordinate space centered at the origin, evolved networks converge on a small central cluster of input pixels, a spatial-concentration bias; prior work observed only 21% mean accuracy in this regime. Is this bias an optimization artifact or an architectural ceiling? Inspired by Mixture-of-Experts (MoE) principles, we partition the input into non-overlapping spatial segments, each assigned to a separately evolved specialist network. With 13 such experts, this design reaches 43% mean accuracy, a 106% relative improvement over the baseline. The architectural gain does not depend on data-driven aggregation: equal-weighted averaging, which uses no validation data, already yields a 70% improvement; the gain comes from partitioning, not the weighting. Receptive-field analysis shows the mechanism: partitioning forces evolution to discover features across the entire image, expanding active pixel coverage from 4% to 79%. Absolute accuracy stays below gradient-trained baselines, but the relative gain points to central bias, not the evolutionary search. Two tools are designed to generalize beyond MNIST: a receptive-field diagnostic for silent input-coverage collapse, and a spatial-partitioning remedy that restores coverage.
Read the full abstract

All authors Romain Claret, Arthur Gygax, Michael O'Neill, Paul Cotofrei, Michael Palma Mendes, Pascal Felber

Analysis of PDEs

A sharp curvature lower bound for the first exterior p-harmonic Steklov eigenvalue

In this paper, we study the first variational Steklov eigenvalue of the $ p$-Laplace equation on exterior domain $ \Omega^{\text{ext}}$ for $ 1< p <n$. If $ \Omega$ is convex and $ \partial \Omega\in C^{1,1}$, we prove a sharp lower bound in terms of $p$-logarithmic mean of the principal curvatures of $ \partial \Omega$. For the linear case $ p=2$, our estimate reduces to the logarithmic mean bound of Bundrock et al. (arXiv:2511.09490). We also derive an upper bound in terms of a boundary isocapacitary constant. The analysis relies on the established finite energy theory on exterior domains, together with a decay estimate for $p$-harmonic extensions.
Read the full abstract

Computer Vision

LoopVAE: Recurrent Depth Across Scales for Visual Tokenization

Hierarchical visual tokenizers typically allocate different processing blocks to different spatial scales. We ask how much of this computation can use the same parameters. LoopVAE reuses a scale- and loop-conditioned core within and across scales, while keeping resolution-changing transitions independent. A four-block core executes 28 block applications per encoder or decoder. On ImageNet-256, the 29M-parameter convolutional model reaches 0.28 rFID and 32.54 dB PSNR under an approximately 30-epoch two-stage training budget, using approximately 65% fewer parameters than the 84M reference VAEs. A non-adversarial Transformer ablation with the same execution graph finds competitive PSNR and SSIM under global sharing, although unshared blocks improve LPIPS. Targeted loop interventions show that completing the trained recurrence improves reconstruction and that even small feature updates can have substantial downstream effects. Truncation also exposes output-range errors, distinguishing useful recurrent computation from reliable early exit. Runtime profiling reveals the execution tradeoff: fewer stored weights require more arithmetic and longer runtime in the tested configurations. With convolutional and Transformer operators and single- or multi-resolution latent interfaces, LoopVAE establishes recurrent depth across scales as a parameter-sharing design axis for visual tokenization.
Read the full abstract

Software Engineering

ChurnBench: A Drift-Aware Benchmark Demonstrating That Refresh Scheduling, Not Cache Age, Governs Staleness in Agentic AI

In production, agentic systems answer questions over data that lives in several places and keeps changing: licenses are reassigned, users offboarded, prices changed, contracts renewed. Existing retrieval benchmarks freeze the data, so they cannot ask whether an agent's answer is still true, only whether it found the right passage. We present ChurnBench, an open-source benchmark that generates a four-source enterprise data fabric as a timeline rather than a snapshot. Every change is written to an append-only ground-truth ledger, and gold answers are computed from that ledger, never from the live stores. An answer that was correct when its data was retrieved but wrong when evaluated is therefore detected and labeled a freshness error, distinct from a reasoning error; we validate this by resolving ground truth at both timestamps for every case reported. Using the instrument, we find that when a system refreshes on a schedule, cache age does not predict staleness. Across cache ages of 1, 14, and 28 days, freshness errors were 7, 4, and 4, because scheduled refresh bounds staleness by time-to-live, and no TTL lapse was observed in any window. A controlled ablation confirms the mechanism: disabling tiered refresh raises freshness errors from 4 to 45 at 28 days and leaves them identical at one day. The variable a drift benchmark should sweep is therefore TTL configuration against each entity's rate of change, not drift-window length. ChurnBench, the evaluation harness, and all per-error data are released open source.
Read the full abstract

Computer Vision

Prototype Matters: Modality-unified Prototype Self-distillation for Unsupervised Visible-infrared Person Re-identification

Estimating reliable cross-modality association is crucial to unsupervised visible-infrared person re-ID. While optimal transport is shown to be a practical solution for cross-modality association, it suffers from the rigidness of hard label assignment without considering the impact of cluster noise. Moreover, enforcing only cross-modality contrast is also suboptimal, as it fails to jointly optimize the similarity relation within and across modality. In this paper, we propose a novel framework for cross-modality learning by well exploitation of prototypes: First, instead of contrasting with cross-modality prototypes, we show that modality-unified prototypical contrast facilitates better modality invariance by jointly and simultaneously optimizing similarity relation within and across-modality. Taking self-prototype as a steady teacher, we further refine the instance-prototype online relation through prototype-guided self-distillation. The two components are optimized in a unified framework, leading to a simple yet effective model. On standard VI-ReID benchmarks, we perform extensive comparison and analysis, validating the effectiveness of our proposed method. Code is available at: https://github.com/Terminator8758/PoSeD.
Read the full abstract

Quantum Physics

One-clean-qubit spectroscopy of simulated Kitaev chains

Spin qubits in gate-defined quantum dots provide a highly programmable platform for simulating condensed-matter phenomena. In this work, we introduce a digital-analog quantum simulation protocol for extracting the single-particle spectrum of a Kitaev chain. The Kitaev chain is mapped onto qubits via the standard Jordan-Wigner transformation and implemented as a drive-engineered, $N$-site transverse-field Ising model (TFIM) in a linear array of quantum dots. We show that periodically toggling the analog-simulation parameters conditioned on the state of a control qubit causes the dynamics of this control qubit to stroboscopically match the output of the one-clean-qubit (DQC1) model of computation, thereby yielding the full spectrum of the TFIM from measurements of a single spin. Classical postprocessing can then be used to isolate the $N$ single-particle energies of the Kitaev chain from the $2^N$ eigenenergies of the TFIM. By varying the strength of the Rabi drive used to engineer the synthetic transverse field, the spectral signature of the crossover from the trivial to the topological regime of the Kitaev chain could then be mapped out with measurements of just one spin.
Read the full abstract

Commutative Algebra

Quadratic Gröbner bases for cut ideals of cycles and ring graphs

Let $C_n$ be the cycle of length $n\ge3$ and let $I_{C_n}$ be its cut ideal. We show that $I_{C_n}$ has a quadratic Gröbner basis with respect to an explicit weight order. Since the defining configuration consists of $(0,1)$-vectors, the initial monomials of such a basis are automatically squarefree. As the cut polytope of a cycle is the parity polytope, the result gives a regular unimodular flag triangulation of this classical polytope. Together with the known tree case and the clique-sum theorem for cut ideals, the cycle result also yields a quadratic Gröbner basis for the cut ideal of every connected ring graph with at least one edge, thereby supplying the missing cycle input and establishing the result for connected ring graphs.
Read the full abstract

High Energy Astrophysical Phenomena

Cosmological Evolution of Primordial Black Holes: UV/IR Decoupling and the KM3NeT 220 PeV Neutrino Anomaly

The recent observation of a 220 PeV neutrino event (KM3-230213A) by the KM3NeT observatory presents a formidable challenge to standard astrophysical source models. We investigate the hypothesis that this ultra-high-energy signature originates from the terminal evaporation burst of a Primordial Black Hole (PBH). Since PBH evolution spans cosmic history, static vacuum approximations fail to capture early-universe dynamics. Embedding the PBH in a cosmological background via the McVittie spacetime, we demonstrate that early-universe cosmological accretion and expansion-suppressed Hawking emission shift the required initial mass window for a terminal burst occurring today. We show that although the early universe environment dictates the black hole's overall lifespan, its final explosion today ($z \approx 0$) is governed by standard Schwarzschild thermodynamics. This mechanism naturally produces the intense 220 PeV local flux while suppressing early emissions, thereby satisfying diffuse isotropic background limits. Consequently, this dynamical framework alters the mapping between current ultra-high-energy neutrino observables and the primordial curvature perturbations that seeded them.
Read the full abstract

Astrophysics of Galaxies

Generating the wide sequence of Diffuse Galaxies with de Broglie waves of Dark Matter

Extensive Euclid satellite imaging at low surface brightness has revealed that most nearby galaxies are diffuse-looking spheroids, where the stellar radius increases for over three decades in luminosity. We argue this Diffuse Galaxy sequence results from internal stellar diffusion by Wave Dark Matter ($\psi$DM), as wave energy is transferred to star orbits over time. In particular, the soliton random motion scatters central stars onto radial orbits that become enhanced with each passage through the centre, slowly “puffing up" the stellar profile. Heating is greater within massive galaxies as $\psi$DM fluctuations are stronger and more frequent, reproducing both the Diffuse Galaxy sequence and the rising velocity dispersion along the sequence, from Ultra-Faint to Dwarf Spheroidal and Ultra Diffuse galaxies, favouring a light boson, $m_\psi=2.88^{+0.14}_{-0.13}\times10^{-22}$eV. Winding back this diffusion, we predict the stellar content of Diffuse Galaxies, including globular clusters, formed near the centre, as anticipated by $\psi$DM simulations, where gas cools efficiently within the dense soliton. This predicted $\psi$DM evolution from compact beginnings towards diffuse spheroidal galaxies today can now be fully charted from JWST to Euclid.
Read the full abstract

All authors Jiashuo Zhang, Tom Broadhurst, Alvaro Pozo, Tzihong Chiueh, Justin Schive, Bo-Ren You, Jeremy Lim, Francine Marleau, Tomomi Sunayama, Keiichi Umetsu, Mariam Bouhmadi-Lopez, Masamune Oguri