Cryptography and Security
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Saturday 7 November
Security fixes usually reach users as compiled updates, but deployment delays leave unpatched systems exposed. During this window, comparing patched and unpatched binaries can reveal how to exploit a vulnerability, turning a defensive update into actionable information for attackers. Existing benchmarks do not measure this capability: they expose source code, rely on human-crafted challenges, or evaluate binary comprehension without requiring exploitation. We introduce CyberBinGym, a benchmark containing 100 tasks derived from real vulnerabilities in 53 widely used projects, spanning six vulnerability families and three levels of analysis complexity. Agents receive stripped vulnerable and patched binaries without source code or vulnerability information, and must combine static and dynamic analysis to recover and exploit the vulnerability. Evaluation separately measures vulnerability recovery through differential crash testing and end-to-end exploitation through flag retrieval, with trajectory audits to reject reward hacking. The strongest evaluated agent recovers the target vulnerability in 88% of tasks and exploits it in 20%. Providing the patched binary raises its exploit rate from 10% to 25% on a matched 20-task subset; with standard binary hardening enabled, it still exploits two of its 20 previously successful tasks. These results demonstrate that current agents can turn released binary patches into working exploits in a controlled setting, although exploitation remains unreliable. These capabilities lower the barriers to targeting unpatched software and compress defenders' effective response window, increasing the urgency of timely patch deployment.
Applied Physics
Bifurcation-based phononic frequency combs in a forced van der Pol-Duffing oscillator
Here, we analyze the dynamics of a forced Van der Pol-Duffing oscillator in search of phononic frequency comb (PFC) generation. When the drive amplitude is small, quasi-periodic self-oscillations are dominant, and we obtain a PFC spectrum response of the oscillator. When the drive amplitude is sufficiently large, we obtain frequency locking. We observe that near the transition to frequency locking the PFC spectrum is strongest and has more peaks. With increasing drive amplitude, the peaks coalesce, and the PFC spectrum disappears at an infinite-period saddle-node or homoclinic bifurcation. We also study the stability of the frequency-locked responses and find that a PFC spectrum can also occur near a subcritical Hopf bifurcation. We find that the first-order averaged equations of our nonlinear dynamical system are identical to the equations of motion of the Stuart-Landau oscillator. We use averaging techniques up to second order to characterize the quasi-periodic solutions. Very good agreement with numerical integration results are obtained. We complement our analysis by proposing a phenomenological model of nonlinear dissipative coupling between two modes that accounts for the negative dissipation term in the van der Pol-Duffing equation of motion.
Read the full abstract
Computation and Language
Behavioral Evaluation of Large Language Models Under Cognitive Biases in HAZOP Studies
Hazard and Operability (HAZOP) studies safeguard industrial processes against catastrophic accidents. Large language models (LLMs) increasingly assist HAZOP drafting, yet evaluations rely on lexical overlap. Because LLMs internalize human heuristics, prompts may trigger biased judgments. Whether frontier LLMs withstand cognitive perturbations in safety-critical workflows remains unknown. Here we show across 420 trials that three frontier LLMs resisted 96.8% of cognitive manipulations, rendering overall susceptibility statistically inconclusive. Contrary to assumed broad fragility, models rejected confirmation and availability cues. However, safety-critical risk downgrades emerged under low anchors and gain frames. Thus, automated hazard analysis requires disciplined human-in-the-loop validation.
Read the full abstract
Computer Vision
Locating and Repairing Domain Shift in VLM Trajectory Planning
Vision-Language Models (VLMs) are increasingly adopted as end-to-end driving planners, yet they suffer from cross-city domain shift. Existing methods treat this shift as a single quantity addressed by uniform alignment or single-site editing. We challenge this view on two counts: (i) domain shift in VLM trajectory planners is inherently structured across modality and layer, and (ii) a locate-then-repair recipe matched to this structure outperforms every uniform-objective baseline. We analyse domain shift via activation patching and uncover two regularities. (F1) Modality axis. Image and text tokens exhibit distinct shift patterns in each VLM. (F2) Layer axis. The layer-wise shift partition is architecture-dependent and dataset-stable. These findings recast domain shift as a two-dimensional tensor indexed by modality (image vs. text tokens) and layer, motivating a simple principle: locate where shift occurs, and repair only there. We instantiate this principle in MoLaRx (Modality-Layer Repair), a locate-then-repair framework. The Causal Importance Score (CIS) estimates the domain shift as a modality-layer tensor and partitions the layers into functional zones. CIS-Guided Selective Repair (CISR) then applies a matched operator (Maximum Mean Discrepancy (MMD) / low-rank adaptation (LoRA) / freezing) to repair each zone, with the composition adapting automatically to each architecture. We evaluate MoLaRx on three architecturally distinct VLMs (LFM2-VL-1.6B, Qwen2-VL-2B, and PaliGemma2-3B) across two cross-city benchmarks (nuScenes and Argoverse 2). MoLaRx attains the lowest cross-city trajectory-error gap among adaptation methods in all six main-table settings, while using 50-60% fewer trainable parameters and 8.4-10.9 ms lower inference latency. In contrast, uniform MMD-only adaptation collapses in all six settings, highlighting the importance of structure-guided repair.
Read the full abstract
Artificial Intelligence
No Second Run: Does RSI Stop at Frontier Pre-training?
Pre-training is where every foundation model begins; bringing recursive self-improvement (RSI) there would let AI improve the process that creates AI. Yet public RSI work has so far targeted only post-training, which starts from an existing base model, runs trials costing GPU-hours, and returns verifiable scores. Pre-training breaks all three conditions at orders-of-magnitude higher cost: a run consumes hundreds of thousands of GPU-hours over months and cannot be repeated; its objective, post-training headroom, cannot be read off any single score; and what must improve is not a model but the whole production system, from data pipelines and training infrastructure to the teams that run them. To our knowledge, we are the first to apply RSI to an entire frontier pre-training run from scratch and to report it in detail. RSI does not stop at frontier pre-training, but it changes form: since no end-to-end loop can close over a single unrepeatable run, we decompose it into smaller loops that can, each repairing a step the full run breaks. The result is 801-30B-A3B-Base, a 30B-A3B MoE trained from scratch on 11.8T tokens in three months by a newly formed team. We release it as a base model; the lightweight SFT reported here (6B tokens, no RL) only probes post-training headroom. As a base model, 801-30B-A3B outperforms Qwen3-30B-A3B and Nemotron-3.5-Lightning-30B-A3B, trailing only the larger Qwen3.5-35B-A3B, and keeps this position under identical SFT; with lightweight SFT alone, it achieves the highest average on four agentic benchmarks against fully post-trained models, including Qwen3.6-35B-A3B. Since post-training RSI presupposes a strong base model, bringing RSI into pre-training acts on the foundation every later stage builds on. We open-source 801-30B-A3B-Base and the lightweight SFT data at https://github.com/GAIR-NLP/801-30B .
Read the full abstract
Machine Learning
FedSGN: Federated Sparse Grassmann Newton Learning for IoT Anomaly Detection
Anomaly detection plays a critical role in securing the Internet of things (IoT). Although existing federated principal component analysis (PCA) methods can learn the subspace of normal samples without sharing raw data, they often suffer from feature redundancy that may degrade detection performance. To address this limitation, we propose a novel and effective federated sparse Grassmann Newton (FedSGN) learning method for IoT anomaly detection. Technically, our FedSGN leverages the $\ell_{2,0}$-norm constraint on the global projection matrix to eliminate redundancy and introduces an orthogonality penalty to preserve the intrinsic structure. Moreover, we develop an augmented Lagrangian solver integrated with a semismooth Newton scheme to enable parallel client updates and Grassmann projection for robustness. Numerical experiments demonstrate that the proposed method achieves favorable overall detection performance while reducing communication overhead. The source code is publicly available at https://github.com/xianchaoxiu/FedSGN
Read the full abstract
Computer Vision
Disentangled Stylized Image Generation via IP-Cross-Attention Rearrangement
With a solid theoretical foundation and pretraining on large-scale datasets, diffusion-based methods have achieved controllable generation leveraging text and image conditions, and generating excellent stylized images is possible using diffusion models. However, previous methods face significant challenges. First, because diffusion generation is iterative, the common approach of learning implicit style representations via cycle-consistency strategies is too costly and not well-suited to diffusion, even for LoRA-based methods. Second, decoupling style is difficult in the latent space of diffusion models. Representations of attributes are coupled in the latent space of diffusion models; thus, extracting style attributes with diffusion models usually suffers from the issue of content leakage. Furthermore, because style is hard to represent in natural language, capturing style using text prompts or textual inversion is deficient. In this paper, we propose a method for achieving disentangled stylized image generation through IP-cross-attention rearrangement, based on our experimental observations. We also propose negative content guidance, inspired by classifier-free guidance, to reduce content leakage in stylized image generation.
Read the full abstract
Combinatorics
The $\nu$-core label order and noncrossing partitions
We prove that the poset of $s$-noncrossing partitions is isomorphic to the core label order of the $\nu$-Tamari lattice introduced by Préville-Ratelle and Viennot.
Information Retrieval
From User Behaviors to Satisfaction Judgments: Reasoning about User Satisfaction with AI Overviews
The growing presence of AI overviews in everyday search makes their effective evaluation essential. Response quality has been the primary focus of existing evaluation research, with only a small number of studies examining the user perspective. These evaluations rely on queries and response content, leaving the information conveyed by user behavior insufficiently utilized. This work investigates satisfaction prediction for AI overviews from a user-centered perspective and incorporates behavioral signals into the prediction process. Applying existing behavior-based satisfaction prediction methods from traditional information retrieval to this setting, however, presents limitations. We address these limitations with STaRS, an LLM-based approach designed specifically for predicting satisfaction with AI overviews. At its core is a structured reasoning procedure: individual behavioral signals are interpreted first, and their implications then inform the prediction of user satisfaction. Reinforcement learning further optimizes this procedure, while a reward shaping mechanism alleviates reward sparsity during training. Evaluation on two real-world datasets demonstrates that STaRS outperforms the comparison methods in satisfaction prediction, providing search providers with a practical means of accurately estimating users' satisfaction with AI overviews.
Read the full abstract
Computer Vision
Permute-then-Adapt: Weak-to-Strong Contrastive Image-Text Adaptation
Image-text alignment models such as CLIP are typically trained with contrastive learning, where unpaired examples are treated as strict negatives. While this assumption introduces noise by penalising semantically related pairs, existing solutions often rely on heuristic similarity thresholds that lack statistical grounding and sensitivity to dataset-specific noise distributions. In this paper, we propose Permute-then-Adapt (PTA), a framework that leverages efficient weak models to identify missed positives while controlling Type-I error relative to a conditional permutation reference, up to calibration approximation error. Unlike standard distillation or arbitrary thresholding, PTA estimates a random-pairing reference distribution from the weak teacher's similarity scores and uses its upper tail, determined by $\alpha$, to select auxiliary positives. The calibrated positive set is then trained against by a single multi-positive contrastive objective: each anchor maximises the total probability mass it assigns to its positive set. Offline calibration with the weak model adds negligible training overhead. Experiments show gains over heuristic soft-label approaches in object recognition, cross-modal retrieval, and data efficiency. Project repository: https://github.com/JinhaoLee/Permute-then-Adapt.
Read the full abstract
Computer Vision
Quadratic Bezier Splatting with Structural Initialization for Efficient Painting Rendering
Stroke-based rendering is often optimized as a pixel-level reconstruction task. This over-reliance on global error reduction not only degrades strokes into fragmented curves due to a lack of structured planning, but also allows locally defective primitives to be concealed by overlapping layers. To address these issues, we introduce a model-guided framework for stroke-audited rendering. Our framework uses Quadratic Bezier Splatting to analytically model continuous curved strokes with efficient differentiable alpha compositing. Built on this renderer, a feedforward multi-scale stroke decoder provides a structural initialization with learned stroke-scale allocation, while test-time stroke optimization supplies the reconstruction capacity needed for high-fidelity refinement. We further propose a single-stroke audit mechanism that explicitly evaluates local stroke quality, preventing defective primitives from surviving optimization through layer-wise compensation. Compared with existing methods, our approach achieves coherent layouts, improves PSNR by up to 5.46 dB, yields a 16% more compact representation, and accelerates combined renderer forward and backward computation by a factor of 220.3.
Read the full abstract
Systems and Control
Deep Reinforcement Learning with Massively Parallel Simulation for Low-Thrust Multi-Revolution Rendezvous
The growing demand for on-orbit servicing and constellation management in low Earth orbit requires autonomous and reliable low-thrust rendezvous. Traditional numerical methods struggle with the long-horizon nature of multi-revolution maneuvers under perturbations, while standard deep reinforcement learning (DRL) methods are limited by low simulation throughput and poor sample efficiency. To address these challenges, this paper proposes a massively parallel DRL guidance framework. A fully tensorized, GPU-resident orbital propagator is developed to integrate $J_2$-perturbed orbital dynamics and policy optimization within a zero-copy pipeline, achieving up to a 67-fold increase in simulation throughput over conventional CPU baselines. In addition, the rendezvous problem is formulated in modified equinoctial elements and trained with a progressive curriculum learning strategy to improve convergence and generalization across increasingly challenging rendezvous scenarios. Simulations show that the proposed framework can learn robust rendezvous trajectories for complex multi-revolution rendezvous over a broad range of orbital configurations. Additional evaluations under a high-fidelity model and eclipse-induced thrust interruptions further demonstrate the robustness of the learned policy under practical mission constraints. Moreover, the trajectories generated by the proposed framework can be used as effective initial guesses for optimization-based trajectory solvers, improving the convergence of high-precision trajectory optimization. The method offers a scalable approach for real-time low-thrust guidance and trajectory optimization.
Read the full abstract
Logic in Computer Science
Applicative theories can count
Applicative theories offer a proof-theoretic framework for studying complexity. In this paper, we introduce a family of applicative theories to characterize classes ranging from constant-depth circuits to polynomial space. Our approach relies on a simple axiom scheme, the small variations of which allows to capture various forms of counting and alternation. This leads to highly uniform results. When applied on notation, the scheme captures bit counting, yielding a characterization of the functions computable by constant-depth, polynomial-size threshold circuits ($\textbf{FTC}^0$); when applied lexicographically, it allows for the counting of accepting computations, yielding a characterization of the functions in the counting hierarchy ($\textbf{FCH}$). Increasing the logical complexity of the induction formulas allowed in the scheme, we obtain characterizations of the functions computable by bounded Boolean circuits of logarithmic-depth and polynomial-size ($\textbf{FNC}^1$) and of the functions computable in polynomial space ($\textbf{FPSPACE}$). The latter characterization solves an open problem in the field. Finally, we investigate separation results. While it is unknown whether $\textbf{FNC}^1$ and $\textbf{FCH}$ are different, there is an oracle relative to which (relativized) $\textbf{FNC}^1$ is strictly contained in (relativized) $\textbf{FCH}$. We use this fact to separate the theories characterizing them, by exploring their ability to represent higher-order functionals.
Read the full abstract
Classical Analysis and ODEs
Optimal orders for Hilbert-valued Littlewood–Paley functions of symmetric semigroups
We study the Littlewood–Paley $g$-functions associated with symmetric submarkovian semigroups and Hilbert-valued functions. For every $1<p<\infty$ and $2\le q\le\infty$, we prove \[ \norm{G_q^T(f)}{L_p(\Omega)} \lesssim\max\{p',\sqrt p\}\norm{f-F(f)}{L_p(\Omega;\mathcal{H})}, \] with the sharp order $\max\{p',\sqrt p\}$, where the implicit constant is independent of $p,q$, the underlying space $\Omega$, the semigroup $(T_t)_{t>0}$, and the Hilbert space $\mathcal{H}$. Compactly supported radial functions show that, for each fixed $q>2$, the large-$p$ order cannot be replaced by $p^{1/q}$ in a dimension-free heat estimate. In particular, for the case $q=2$, we answers affirmatively the question posed by Xu in his JFA paper (2022) \cite[Problem~7]{XuJFA}. For every finite $q>2$, our lower bound gives a negative answer to the dimension-free question in Xu's JEMS paper (2025) \cite[Problem~1.8(i)]{XuJEMS}, already for real scalar functions. Together, the upper and lower bounds establish the sharp dimension-free order $\max\{p',\sqrt p\}$ for Hilbert-valued functions.
Read the full abstract
Classical Analysis and ODEs
Nonlocal bilinear differential operators induced by a third order hypermetric in $\mathbb{R}^n$
The third order hypermetric $\rho(x,y,z)$ in $\mathbb{R}^n$ is given by the Euclidean distance of the point $(x,y,z)\in(\mathbb{R}^n)^3$ to the diagonal $\Delta^{(3)}=\{(x,x,x):x\in\mathbb{R}^n\}$ of $(\mathbb{R}^n)^3$. In this paper, we consider local and nonlocal fractional versions of bilinear differential operators generated by the geometry induced by $\rho(x,y,z)=\rho(0,x-y,x-z)$ instead of the usual $|(x-y,x-z)|$, where $|\cdot|$ denotes the Euclidean norm in $\mathbb{R}^{2n}$. In doing so, we obtain a fractional nonlocal version $\mathbb{D}^s(f,g)$, $0<s<1$, of the Dirichlet bilinear operator $\nabla f(x)\cdot\nabla g(x)$ which converges to this classical form when the fractional parameter approaches its endpoint $s=1$. We consider both the kernel approach and the Fourier multiplier version.
Read the full abstract
Networking
Distributed Intelligent Ad Hoc Routing Agents for Communication at Protests During Internet Shutdowns
Governments around the world are increasingly using internet shutdowns to block digital communication. Internet shutdowns are particularly dangerous during political protests, as they inhibit protesters' ability to organize and communicate critical, potentially life-saving information and may be used to prevent awareness of state-based violence. Recent prior works have explored secure, private messaging systems for use during political protests and internet shutdowns but have encountered long delivery delays or low delivery rates under challenging conditions. In this work, we explore the use of distributed, adaptive, social- and congestion-aware heuristics and efficient communication patterns to create an intelligent ad hoc routing protocol in which network agents learn and leverage the network social environment and monitor and respond to dynamic network congestion behaviour to route messages in a time- and resource-efficient manner. Through this exploration, we contribute an intelligent algorithm and protocol which overall achieves comparable or higher delivery rates relative to the leading dynamic clique routing protocol which it extends while also transmitting messages through notably fewer carriers and, when network devices have larger amounts of buffer storage, transmitting fewer message-related packets. We additionally contribute a further understanding of routing limitations and challenges and conceptualizations of extensions of intelligent dynamic clique routing to overcome them.
Read the full abstract
Cryptography and Security
Towards Lightweight IoT Intrusion Detection: Feature Optimization and Explainable Machine Learning for Resource-Constrained Environment
Machine learning-based intrusion detection systems (IDSs) can improve IoT security but often rely on large feature sets that increase computational overhead and limit deployment on resource-constrained devices. This paper presents a resource-efficient and explainable feature optimization framework for lightweight IoT intrusion detection using the CICIoT2023 dataset. A Random Forest model was used to rank feature importance and identify compact feature subsets. Experimental results show that a three-feature model achieved an MCC of 0.9622 while reducing the feature space by 92.1%, whereas a six feature model achieved an MCC of 0.9675 with an 84.2% reduction. Furthermore, a fifteen-feature model achieved an MCC of 0.9744, preserving 99.97% of the performance of the full 38-feature baseline model while reducing the feature space by 60.5%. SHAP-based explainability analysis validated the selected features and provided interpretable insights into model decisions. The results demonstrate that lightweight, explainable, and high-performance IoT intrusion detection can be achieved using only a small subset of carefully selected features, significantly reducing resource requirements with minimal loss of detection effectiveness.
Read the full abstract
Mesoscale and Nanoscale Physics
Probing Chern Topology through Hall-Sensitive Photonic Spin Hall Shifts in a d-wave Altermagnet
In this paper, we investigate the photonic spin Hall effect in a two-dimensional Rashba-coupled $d$-wave altermagnetic Chern system subjected to a uniform out-of-plane exchange field. We first analyze the energy-band structure and the Dirac mass distribution at the high-symmetry points, where $m_{\Gamma}=m_M=\Delta_x, m_X=\Delta_x-4 t_J$, and $m_Y=\Delta_x+4 t_J$. Band inversions at high-symmetry points generate both a topologically trivial regime $(C=0)$ and Chern-insulating phases $(C= \pm 1)$. Employing the Kubo formalism for the optical conductivity together with the reflection matrix for a substrate-supported conducting sheet, we compute helicity-dependent finite-waist displacements of a reflected Gaussian beam. The resulting optical response exhibits a minimum interband absorption threshold, a Van Hove-type spectral feature, and distinct transitions at the $X$ and $Y$ points arising from the momentum dependence of the exchange field. Under the specified Hall-tensor and polarization sign conventions, the longitudinal helicity half-separation is odd upon reversal of the optical Hall conductivity and correspondingly changes sign between opposite-Chern phases, whereas the transverse half-separation is even in the Hall response. The maximal spin-dependent splitting occur in the vicinity of the pseudo-Brewster reflectance minimum. The longitudinal spin splitting reverses sign between opposite Chern phases, reflecting the exchange-controlled optical Hall response, while its pronounced spectral variations reveal the influence of inequivalent interband transitions induced by the $d$-wave altermagnetic exchange. These findings establish polarization-resolved photonic beam shifts as a sensitive probe of Hall-active topological phases and their characteristic optical spectra.
Read the full abstract
Machine Learning
AbductionBench: Benchmarking Abductive Reasoning in Large Language Models
Abductive reasoning is the inference of plausible explanations for observed evidence. Evaluating this capability requires more than a single task: a system may need to generate an explanation, discriminate among competing hypotheses, revise its hypotheses as evidence arrives, or decide what evidence to seek next. Yet these capabilities are typically evaluated through isolated benchmarks with different task formulations and scoring procedures, so claims that a model is “good at abduction” may reflect only a narrow aspect of this capability. We introduce AbductionBench, a unified framework whose taxonomy characterizes abductive tasks. We instantiate this framework on 27 datasets spanning ten domains and evaluate eight LLMs from four model families evaluated on single-choice selection tasks. Beyond harmonizing heterogeneous tasks, AbductionBench separates final-answer evaluation from analysis of the observable reasoning traces and interaction histories produced alongside those answers. We analyze written traces as observable model outputs rather than faithful representations of internal computation, characterizing evidence use, hypothesis comparison, uncertainty, contradiction handling, and answer commitment. Jev matches or exceeds small LLMs when selecting among explanations over compact evidence but falls substantially behind on longer cases, suggesting that candidate discrimination and extended evidence integration place different demands on a system. In interactive tasks, the proportion of actions judged relevant declines at later steps, and a relevant first question is associated with a correct final answer; in the passive task, models rarely revise their initial hypothesis as new evidence arrives. Together, these findings highlight the value of evaluating abductive reasoning through both final answers and observable evidence use, hypothesis revision, and information gathering.
Read the full abstract
Artificial Intelligence
R-BED: A Consequence-Aware Framework for System-Level AI Evaluation and Continuous Assurance
Evaluations of AI systems usually report properties of a model, while decisions to deploy concern a configured system: a model together with prompts, tools, permissions, data, safeguards and an operating context. We ask how evidence about technical failures, deployment-control effectiveness and real-world consequences can be connected to defensible authorization decisions throughout the lifecycle of a deployed system. We present R-BED (Risk-Based Evaluability Design), a specification-level methodology with a fixed 6+4+3 architecture of thirteen diagnostic dimensions (technical failure, deployment controls, consequence and impact); a failure-mediated expected-loss identity that is kept separate from intended-behavior and misuse pathways; bounded evidence statuses with independence levels; a control assurance ladder; five authorization dispositions governed by eight rules; and a Continuous System Evaluation (CSE) layer in which evidence carries a configuration pin, an expiry and declared dependencies, so that a change invalidates evidence in proportion rather than wholesale. We work through a synthetic refund-agent walkthrough, with a deterministic script that reproduces every number, to show how a task-success view can miss a consequential control weakness that the procedure identifies as missing evidence. The contribution is a reproducible methodological specification. Most components are standard practice integrated into one evidence-to-decision record; the Evidence Validity Register, the authority envelope and the tier rules are proposals. None has been empirically validated, and we state seven hypotheses for independent testing.
Read the full abstract
High Energy Astrophysical Phenomena
Probing the origin of the LHAASO Peanut with a hidden central accelerator
A highly asymmetric, giant Peanut-shaped structure has been detected by the Large High Altitude Air Shower Observatory (LHAASO) in ultra-high-energy gamma rays at high Galactic latitude, $b\approx-17.5^{\circ}$, suggesting anisotropic particle transport around a Galactic accelerator. We investigate whether its spectrum and morphology can arise from a hidden central accelerator embedded in a projected anisotropic halo. The baseline model accounts for the diffuse strip and central excess, while a phenomenological prescription for a localized magnetic field reorientation toward the line of sight (LOS), together with a particle-conserved remapping of the electrons downstream of the far excess into an emitting column of prescribed transverse extent, enhances the emission at the far excess position. We explore various source ages, along with different combinations of ambient magnetic field strength and parallel diffusion coefficient, while accounting for radiative cooling when estimating the viewing geometry. The exploration reveals that relatively rapid parallel transport enhances the far component while preserving a dominant diffuse strip. Representative configurations remain compatible with the observed spectral energy distribution, although the fitted far component remains comparatively weaker than the central component above 100 TeV, even in cases where their component fluxes are comparable at 25-100 TeV. These results illustrate the plausibility of a hidden central accelerator as the origin of the LHAASO Peanut, while identifying the remaining difficulty in reproducing the relative brightness of its compact components within the adopted phenomenological framework.
Read the full abstract
Metric Geometry
Optimal rational spherical designs on the circle
The minimum numbers of points in rational equal-weight spherical designs on the circle are $2304$, $4608$, $9216$, and $9216$ at strengths $4$, $5$, $6$, and $7$, respectively. We construct two nonisometric optimal designs at each strength. At arbitrary strength $t$, we obtain an explicit cardinality divisor $D_t$ with $\log D_t=(1/2+o(1))t\log t$. Applied to the antipodal pairs of a $4608$-point design, Gromov's torus construction gives a flat minimal immersion of $\mathbb{T}^2$ into $S^{4607}$ with $2304$ distinct integral frequencies, attaining Petrunin's sharp Euclidean normal-curvature bound.
Read the full abstract
Discrete Mathematics
Weighted Matroid Intersection under Minimum Rank Oracle
We consider matroid intersection under the minimum rank oracle model, in which two matroids are accessible only through an oracle that returns the minimum of the ranks of a given set in the two matroids. By Edmonds' theorem, this oracle determines the matroid intersection polytope, although it does not reveal the exchangeability graphs on which classical algorithms rely. Recent work by Bárász, Bérczi, Király, Oki, Yamaguchi, and Yokoi (2025) showed that unweighted matroid intersection is polynomially solvable in this model. In this paper, we present a strongly polynomial-time algorithm for weighted matroid intersection in this model. The key to our algorithm is a polynomial-size extended formulation of tangent cones of common base polytopes that can be constructed using the minimum rank oracle. We also give min-max theorems in terms of the minimum rank function via 1/3-submodularity, and present a strongly polynomial-time algorithm to compute an optimal dual solution.
Read the full abstract
Machine Learning
Spectral Bias Reduction in Local Gated ReLU Architectures
Neural networks trained by gradient-based methods learn the low-frequency content of a target function faster than its high-frequency content. This spectral bias is well documented, but it is much less clear which architectural choices mitigate it and which cannot. We study the question for shallow ReLU models linear in their trainable parameters, where each mode is learned at a rate given by a Gram-matrix eigenvalue, and for a general symmetric bilinear form $a(\cdot,\cdot)$, so that least-squares regression ($a=$ the $L^2$ inner product) and variational or physics-informed losses ($a=$ the $H^1$ energy form) are two instances of one analysis. Within this framework we propose gated local ReLU dictionaries, obtained by multiplying each ReLU feature by a hat partition-of-unity weight, in the spirit of partition-of-unity finite elements and overlapping Schwarz methods. The resulting Gram matrices are block-banded for any local bilinear form, with a block-Toeplitz interior whose matrix-valued symbol governs the spectrum. The two forms then behave very differently. For $L^2$ regression, the symbol is bounded away from zero, the ratio is uniform in the number of windows, and the measured spread of frequency-dependent learning rates falls from $1.9\times10^{5}$ to $4.5$ at matched model size. For the $H^1$ stiffness matrix, the symbol vanishes at zero frequency: gating leaves the $\Theta(N^2)$ conditioning intact and inverts the frequency ordering instead of removing the bias. A preconditioner built from the same interior stencil will fix the spectral bias, showing that architecture alone is not enough. Experiments on $1$D regression and Poisson problems confirm the dichotomy, and the regression effect persists when the hidden-layer parameters are trained as well.
Read the full abstract
Computer Vision
PLS-GS: Probabilistic Lidar Scaffolds for Compact, Surface-Aligned Gaussian Splatting
While 3D Gaussian Splatting (3DGS) is typically optimized from images only, many robotics and mapping applications provide both camera and high-definition LiDAR data, combining rich visual appearance with accurate geometric structure. We introduce PLS-GS, a probabilistic LiDAR scaffold for Gaussian Splatting that exploits this complementary sensing setting to construct a compact 3DGS representation aligned to the measured surface. PLS-GS fits a Hierarchical Gaussian Mixture Model jointly over position and color, yielding a continuous and adaptive scaffold that eliminates the need for discrete scene partitioning such as voxel grids or octrees. The fitted mixture then guides the renderable 3DGS Gaussians: each rendering Gaussian is associated with a GMM component and constrained in the local surface frame induced by that component's covariance. A Mahalanobis support loss bounds the deviation of the rendering Gaussians from the corresponding LiDAR-derived surface. Across five large outdoor scenes, PLS-GS places primitives markedly closer to the LiDAR measurements than both image- and LiDAR-initialized baselines, while employing substantially fewer primitives, yielding a considerably more compact model and faster rendering. At matched capacity, PLS-GS places primitives closer to the LiDAR measurements, with comparable rendering quality. Code is available at https://github.com/k-harikumar/PLS-GS.
Read the full abstract
Cryptography and Security
A Case Study of a Phishing Cluster Targeting German Universities
In light of recent ransomware attacks, this study analyzes and measures the impact of cyberattacks against academic institutions in German-speaking areas. First, we provide an overview of 25 confirmed cyberattacks against universities in German-speaking areas between February 2022 and February 2023. In this time frame, we find that 5% of all German universities have experienced a confirmed cyberattack with a measurable impact. Second, we document victimology and Tactics, Techniques, and Procedures (TTP) of a phishing activity cluster spanning more than three years and showing a particular focus on universities in German-speaking areas. Of the 93 targeted universities, 58 (62%) are German, representing a significant fraction with approximately 14% of the total number of universities (423) in Germany. A notable technique is that the phishing emails are sent via university infrastructure, using valid, compromised credentials of university-associated personnel and involving university Virtual Private Network (VPN) tunnels.
Read the full abstract
Artificial Intelligence
Hallucination-Guided Unlearning: Using Hallucination Traces to Reveal Overfitted Memories
Selective unlearning is increasingly important for the safety and compliance of large language models. Harmful or memorized content is often entangled with useful knowledge and difficult to enumerate. We use the model's own failures to identify candidates for unlearning. We propose Hallucination-Guided Unlearning (HGU), which uses hallucinations as diagnostic signals. HGU localizes unsupported spans via evidence and internal consensus, traces a sparse set of high-responsibility MLP neurons, and edits them with compact gates under a utility-preserving KL constraint. On TOFU, HGU achieves the best MU-FE trade-off, with Avg scores of 0.7221 (forget05) and 0.7699 (forget10), and the highest SepS H-Avg (0.4382/0.4715) under mixed forget/retain prompts. On the WMDP auxiliary task, HGU yields the strongest forgetting (e.g., 14.2 on Economics) while improving Retain (52.7) and MMLU (57.2). These results suggest that tracing hallucinations enables targeted, post-hoc unlearning that better balances forgetting efficacy and general utility without full retraining.
Read the full abstract
Materials Science
Scalable MoS2/Si Photodiode Arrays from Roll-to-Roll Mechanical Exfoliation
We present a scalable, all-dry route to MoS2/Si photodiodes based on high-throughput roll-to-roll mechanical exfoliation. The method yields dense films of large-lateral-size MoS2 nanosheets and enables the formation of heterojunctions that operate in the photovoltaic regime when transferred onto locally etched Si regions. Individual devices show robust, fast photoresponses and broadband wavelength sensitivity. We further implemented a linear pixel array demonstrating position sensing capability and a 4x4 array as a proof-of-concept imaging sensor. These results indicate that roll-to-roll mechanical exfoliation can overcome common limitations of solution-processed films to enable low-cost, self-powered optoelectronic detectors for sensing and imaging applications.
Read the full abstract
Optimization and Control
Port-Hamiltonian Systems Beyond Joules: A Review of Unexpected Application Domains
The port-Hamiltonian framework was built for mechanical and electrical systems, where the Hamiltonian is stored energy and the ports carry power. Over the past fifteen years it has spread to chemistry, cell biology, fusion plasmas, musical instruments, district heating, electricity markets, macroeconomics, predator-prey ecology, optimization algorithms and neural-network training. This review maps that spread. We classify each application by what its Hamiltonian really is: physical energy, a thermodynamic potential such as free energy or exergy, or an abstract storage function with no energetic meaning. The review keeps only sources whose bibliographic record we checked against Crossref. The picture that emerges is a continuum. Where a genuine thermodynamic potential exists, the theory is deep and constructive. Where the energy is borrowed, the framework still pays off through compositional interconnection and dissipation inequalities, but the work is younger and scattered. We close with a set of deliberately speculative research directions, including open port-Hamiltonian models of epidemics, opinion dynamics and supply chains, and learned port-Hamiltonian digital twins for structural health monitoring.
Read the full abstract
Computer Vision
Perception, Not Reasoning, Limits Visual Theory of Mind
Vision-language models pass text-based false-belief probes at near-human levels but drop by over twenty percentage points on abstract panels the moment the same problem is conveyed visually. The consensus diagnosis is conceptual, prescribing new belief-reasoning modules. We argue the dominant bottleneck is perceptual: not what the model cannot reason about, but what it cannot see. We propose a five-condition dissociation protocol that holds the reasoning task constant while varying perceptual fidelity from visual only to text only, and quantify the result with the Perceptual Contribution Ratio, the share of the visual-to-text gap a given fidelity closes. We test this on two kinds of scenes: abstract 2D panels in the Heider–Simmel tradition, and naturalistic 3D game scenes from MuMA-ToM and MindPower. The evaluation covers seven VLMs and roughly 159,000 API calls. Ground-truth scaffolds substantially close the gap in every setting, and on the abstract panels and MuMA-ToM they match or exceed the text-only reasoning. Whereas scaffolds built by automatic parsers, including frontier VLMs, fail. The most common error in the visual-only condition is reality bias, which accounts for 60% of false-belief failures. Models report where an object actually is, not where an absent agent believes it is. This points to a perceptual front-end that fails to track which agents were present at each moment. Under the evaluated protocol, structured visual scene-state extraction, rather than belief inference, is the dominant bottleneck for visual ToM.