Peer Revenue Play credits. Real papers.

Papers

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

All of arXiv

Thursday 10 September

Robotics

Gait-Dependent Effects on Quadruped Locomotion for Load-Carrying using Passive Mechanism

Passive mechanical interfaces offer a lightweight alternative to actuated manipulators for quadruped payload carrying, but their impedance directly couples the payload dynamics with the locomotion pattern. This paper analyzes how passive-arm stiffness-damping selection affects payload-carrying locomotion under different gait and payload conditions. We compare damped and underdamped passive-arm impedance configurations in simulation during flat-ground locomotion. For crawl gaits, where the support polygon remains well defined, the results show that underdamped impedance increases passive-joint oscillations and can reduce the ZMP margin with respect to the support polygon. Trot is retained as a dynamic excitation case for the passive arm, but it is not used for direct ZMP-margin stability comparison. The results are summarized in gait-payload-stiffness-damping maps, where ZMP-margin reduction is evaluated for crawl gaits and trot is retained only as a passive-arm excitation case.
Read the full abstract

Machine Learning

EMMI: Edge Multi-Modal Intelligence for Communication-Efficient MLLM Inference via Fused Representation Compression

Recent advances in multimodal large language mod- els (MLLMs) have opened new opportunities for edge intelligence by enabling reasoning across heterogeneous sensor modalities, such as vision, text, and telemetry data. However, deploying these capabilities on resource-constrained edge platforms remains challenging due to the substantial computational, memory, and communication demands of modern MLLMs. Rather than transmitting raw sensor observations or partitioning neural networks at intermediate layers, Edge Multi-Modal Intelligence (EMMI) communicates a compact representation between edge devices and server resources, enabling communication-efficient edge MLLM inference. To achieve this, EMMI performs modality-specific encoding, cross-modal representation fusion, and learned compression at the edge, transmitting only a compact latent representation to server-side resources for high-capacity MLLM reasoning. This representation-centric design reduces communication overhead, preserves local data privacy, and provides a fixed-size interface between heterogeneous edge devices and server-side MLLMs. Evaluation on a representative multimodal benchmark demonstrates that EMMI can reduce the communication payload by 32x while maintaining comparable downstream accuracy, resulting in up to a 3.4x reduction in estimated end-to-end inference latency under bandwidth-constrained edge conditions.
Read the full abstract

High Energy Physics (Experiment)

Search for the process $e^+e^-\to f_1(1285)$ at the SND detector

In the experiment with the SND detector at the VEPP-2000 $e^+e^-$ collider, a search is performed for the direct production of the $C$-even $f_1(1285)$ resonance in $e^+e^-$ collisions. The analysis is based on data with an integrated luminosity of about 200 pb$^{-1}$, accumulated in the center-of-mass energy range of 1.14–1.46 GeV, of which about 72 pb$^{-1}$ were recorded near the maximum of the $f_1(1285)$ resonance. The $f_1(1285)$ production cross section at the resonance maximum $\sigma(e^+e^-\to f_1)=(31\pm 13\pm 2)$ pb and the branching fraction $B(f_1(1285)\to e^+e^-)=(3.5\pm 1.4\pm 0.3)\times 10^{-9}$ have been measured. The significance of the observation of the $e^+e^-\to f_1(1285)$ process is $2.5\sigma$. Since the significance is low, we also present the upper limits at the 90% confidence level: $\sigma(e^+e^-\to f_1)<48\mbox{ pb}$ and $B(f_1(1285)\to e^+e^-)<5.4\times 10^{-9}$.
Read the full abstract

All authors SND Collaboration, M. N. Achasov, A. E. Alizzi, A. Yu. Barnyakov, E. V. Bedarev, K. I. Beloborodov, A. V. Berdyugin, D. E. Berkaev, A. G. Bogdanchikov, A. A. Botov, D. E. Chistyakov, T. V. Dimova, V. P. Druzhinin, R. A. Efremov, L. V. Kardapoltsev, A. S. Kasaev, A. A. Kattsin, V. R. Khamidullin, A. G. Kharlamov, I. A. Koop, A. A. Korol, D. P. Kovrizhin, A. S. Kupich, A. P. Kryukov, A. R. Mamedov, N. A. Melnikova, N. Yu. Muchnoi, A. E. Obrazovsky, A. A. Oorzhak, I. V. Ovtin, E. V. Pakhtusova, I. A. Polomoshnov, K. V. Pugachev, S, A. Rastigeev, Yu. A. Rogovsky, V. A. Romanov, S. I. Serednyakov, D. A. Shtol, Z. K. Silagadze, K. D. Sungurov, I. K. Surin, Yu. V. Usov, V. N. Zhabin, V. V. Zhulanov

Cosmology and Nongalactic Astrophysics

(Re)constructing Accurate Axion Oscillations

The cosmological evolution of ultralight axions typically involves rapid oscillations on the timescale of the inverse mass, making it challenging to resolve the dynamics at both the background and perturbation levels. In this study, we numerically implement a novel approach to this problem based on an effective field theory (EFT) – that is, for the first time, capable of reconstructing the relativistic oscillations intrinsic to the axion field. Compared to other available techniques, the reconstruction of the rapid oscillations is unique to the EFT approach, which describes the axion effective field through a wavefunction representation with relativistic corrections, rather than through effective fluid variables. From a computational standpoint, our approach archives a high level of accuracy – up to a subpercent agreement with the exact solution – while remaining relatively fast compared to alternative methods. As such, it opens up new possibilities for high-precision predictions for axion searches with future cosmological experiments.
Read the full abstract

Cosmology and Nongalactic Astrophysics

A Tale of Two Gauges: Effective Field Theory for Relativistic Behavior of Cosmological Axions

In this work, we present a formalism to model the relativistic behavior of axions. The relativistic behavior of axions is surprisingly difficult to model precisely, as it involves oscillations on timescales much shorter than the Hubble timescale. To overcome this challenge, one typically resorts to some form of effective treatment, focusing only on the time-averaged description of the exact oscillations. Salehian, Namjoo & Kaiser provide a systematic framework for such treatment, based on the effective field theory formalism. While the aforementioned study was formulated for axion perturbations in the Newtonian gauge with no anisotropic stress, we extend the formalism to the synchronous gauge that is more conventionally used for numerical implementation in a realistic cosmological setting. Unlike their work, however, we propose a fluid interpretation in which the axion field can be identified as a perfect fluid at all times, both in the exact and effective regimes. Moreover, we present the effective field theory for the Newtonian gauge with non-zero anisotropic stress, making the original formulation more general and useful for scenarios where the matter content of the universe is multi-component. These results lay the theoretical foundation for a companion paper where we discuss how the axion field should be incorporated alongside other species in common cosmological Boltzmann solvers.
Read the full abstract

Logic

There is no maximal $K$-degree

The Kolmogorov complexity of a string characterize how complex it is to describe the string. If every prefix of a real $x$ is more complex to describe than every prefix (of the same length) of real $y$, then it is seen as $x$ is more complex to describe than $y$. It is wondered if there is a real $x$ so that no other reals are strictly more complex (to describe) than $x$. The behavior of Kolmogorov complexity functions generated by reals (namely $n\mapsto$ the minimal description length of the real) is quite chaos. Therefore, it is widely believed that there are many reals that are maximally complex to describe. For instance, it is conjectured that all random enough reals have maximal $K$-degree. In this paper, it is shown that there is no real with maximal $K$-degree. Actually, for almost all real $x$, we can uniformly computably find another real whose $K$-degree is strictly above $x$. The same construction also derives the following: (a) There is no maximal $C $-degree. (b) Every $n$-random real $x$ admits an $n$-random real $y$ such that $y>_Kx$ and $y>_C x$; moreover, the difference of $K$-complexity and $C$-complexity of $y,x$ (namely $K(y\upharpoonright n)-K(x\upharpoonright n)$ and $C(y\upharpoonright n)-C(x\upharpoonright n)$) do not tend to infinity.
Read the full abstract

Solar and Stellar Astrophysics

Uncovering the Origin of Slow Rotators among Intermediate-Mass Stars in the Star-Forming Cluster Trumpler 14

Intermediate-mass (about 1.5-8 M$_{\odot}$) stars exhibit a wide range of rotation rates and have gained attention for their roles in the extended main-sequence turnoffs (eMSTO) seen in some young and intermediate-age clusters (ages $< 2$ Gyr). Although rapid rotation is expected due to their radiative envelopes, the presence of slow rotators remains puzzling. In this study, we examine disk and X-ray signatures among the intermediate-mass members of the star-forming cluster Trumpler 14. Of the 118 intermediate-mass members, 24 are considered disk-bearing on the basis of their infrared spectral indices derived from spectral energy distributions. The majority of these reside outside the heavily irradiated cluster core and are all slow rotators ($\lessapprox 100$ km s$^{-1}$) except for three fast-rotating Class III objects. Additionally, in the same sample of 118, 37 are X-ray sources, and 22 having available $v \sin i$ data show a systematically slower rotation than X-ray quiet sources. Our findings suggest that, while young stellar disks in intermediate-mass stars contribute to early spin-down, high-energy processes traced by X-ray emission during this phase could be an additional channel for angular-momentum loss in these stars.
Read the full abstract

All authors Khushboo K. Rao, Wen-Ping Chen, Hsiang-Yu Chen, Kristen Dage, Stella S. R. Offner

Information Retrieval

UniRec: Cross-stage Multi-Task Fusion with Preference Optimization for Cascaded Recommender Systems

Industrial recommender systems use cascaded stages with different candidate spaces, data distributions, and latency constraints. Each stage include multi-objective prediction and multi-task fusion (MTF) modules. Existing MTF methods typically design and optimize the fusion module independently for each stage, which can lead to inconsistent preferences across stages. Although cross-stage coordination methods have been proposed, they mainly transfer supervision to upstream scoring models rather than jointly optimizing the fusion modules. We propose UniRec, a Unified Cross-stage Recommendation Fusion framework that jointly optimizes pre-ranking and ranking fusion agents. UniRec employs a coupled dual-agent architecture with partially shared embeddings and stage-specific fusion functions. It further introduces Dual-Axis Preference Optimization to combine stage-local preferences with downstream-to-upstream preference transfer. To enable stable and efficient cross-stage training, we address two practical challenges: reward-induced distribution drift and the computational cost of pairwise preference learning. Extensive offline and online experiments on a production short-video system demonstrate consistent gains over single-stage fusion and cross-stage coordination baselines. In a nine-week online A/B test, UniRec improves total watch time by $0.675\%$ and app usage time by $0.616\%$, and active users by $0.189\%$, and has been fully deployed in production.
Read the full abstract

All authors Lingyuan Kong, Jiaqi Cui, Fanjiao Zeng, Congqi Wang, Yu Li, Yuan Cheng, Jingxin Liu, Xiaoshuang Chen, Kaiqiao Zhan

Signal Processing

Tensor Decomposition Based Mixed-Field Sensing for XL-MIMO AFDM Systems

Integrated sensing and communications enabled by extremely large-scale MIMO (XL-MIMO) and affine frequency division multiplexing (AFDM) is a highly promising paradigm for vehicular networks. However, the near-field spherical wavefront distortions induce severe non-linear parameter coupling, while the highly dynamic scattering environments exacerbate mismatch errors. To address these critical challenges, this paper proposes a novel tensor-based sensing scheme for XL-MIMO AFDM systems. First, the received signals are reformulated into a tensor, followed by an efficient decomposition approach that exploits the inherent Vandermonde structure of the factor matrices. This allows parameters to be directly estimated from the decomposed matrices, effectively avoiding inter-parameter coupling. Subsequently, a symmetric decoupling and real-domain manifold optimization algorithm is proposed for angle of arrival estimation, circumventing the high-dimensional searches typically induced by near-field effects. Furthermore, a baseband reconstruction and analytical gradient-based algorithm is developed to perform delay-Doppler estimation in the continuous parameter domain, fundamentally eradicating the grid-mismatch errors inherent in high-mobility scenarios. With these decoupled factors, the remaining unknown angle of departure can be readily extracted. Extensive simulation results demonstrate that the proposed scheme achieves orders-of-magnitude improvements in delay-Doppler accuracy and eliminates the error floors in angular estimation that severely bottleneck state-of-the-art baselines.
Read the full abstract

All authors Lei Yao, Yin Xu, Aimin Tang, Taohe Chen, Tianyao Ma, Wenjun Zhang

Numerical Analysis

Picard-Based Acceleration of Newton Continuation for Mean Field Game PDE Systems

We develop a method that uses Picard iterations to accelerate Newton continuation for the semi-implicit finite-difference discretization of forward-backward partial differential equation (PDE) systems arising in mean field games (MFGs). We first investigate the computational properties of the Picard and Newton methods, which are widely used separately in the MFG literature but whose comparative cost and robustness across different regimes remain insufficiently explored and documented. The Picard method uses an outer fixed-point iteration that alternates a forward Fokker-Planck solve and a backward Hamilton-Jacobi-Bellman solve. The Newton method instead applies Newton's method directly to the coupled nonlinear space-time system. Across one- and two-dimensional MFG benchmarks, Picard offers substantial computational savings in favorable regimes, but may fail at sufficiently low viscosity or require strong damping under temporal shocks, making Newton continuation preferable. With parameter continuation in the viscosity parameter, the Newton method is more robust in these regimes, at the cost of larger coupled linear systems. We relate these trade-offs to the residuals, Jacobian blocks, and sparsity structures produced by separable, local nonseparable, and nonlocal Hamiltonians. We then demonstrate how to combine inexpensive Picard iterations with Newton continuation in a hybrid method to reduce the total computational cost for a two- dimensional double-well problem.
Read the full abstract

Data Structures and Algorithms

A deterministic $(1+\varepsilon)^n$ approximation for the permanent of a nonnegative matrix

For every fixed $0<\varepsilon\le1$, we give a deterministic strongly polynomial algorithm that, given a nonnegative matrix $A\in\mathbb{R}_{\ge0}^{n\times n}$, returns $Q$ satisfying $\operatorname{per} A\le Q\le(1+\varepsilon)^n\operatorname{per} A$.

Soft Condensed Matter

A Phase-Field Study of Desiccation Crack Pattern Maturation under Drying-Wetting Cycles

The characteristic intersection angle of the desiccation crack relaxes from near \ang{90} toward \ang{120} under repeated drying–wetting cycles. However, the theoretical understanding of this relaxation is insufficient, especially the modeling of the drying–wetting cycles. Here we introduce a phase-field model of desiccation fracture, extending the model proposed in previous studies by adding crack healing and a scar effect left by past cracks. By repeating drying–wetting cycles in a finite element simulation, we find that the angle distribution develops a growing peak near \ang{120} as the cycle number increases, consistent with experiments. The standard deviation of the intersection angle from \ang{120} relaxes exponentially with a characteristic time of about 2.85 cycles. These results are consistent with experiments, except that the characteristic time is slightly smaller than the experimental value. Crack energy dominates the total energy and also relaxes exponentially with nearly the same characteristic cycle as the angle relaxation. This decay is driven mainly by a shortening of the effective crack length rather than a change in effective fracture toughness.
Read the full abstract

Quantum Gases

Transdimensional quantum droplets in an optically trapped Bose mixture

We study quantum droplets in a symmetric two-component Bose mixture with interspecies $p$-wave interactions and a two-dimensional transverse optical lattice. The lattice drives a crossover from an anisotropic three-dimensional gas to weakly coupled one-dimensional tubes. We calculate the ground-state energy and quantum depletion at the Gaussian level and derive their limiting forms. At $y=g_{12}/g=-0.95$, where the bare mean field is repulsive and no free-space droplet exists, the calculated bulk equation of state supports a self-bound minimum across the crossover: a negative lattice contribution at order $n^{2}$ supplies the attraction in the three-dimensional regime, and attractive fluctuations do so in the quasi-one-dimensional regime, with the intermediate, transdimensional range described quantitatively by neither limit. The interspecies $p$-wave interaction modifies only the spin branch. In the parameter range studied, increasing its strength lowers the equilibrium density across the crossover, consistently with a weakening of the induced binding.
Read the full abstract

All authors Xiaoran Ye, Yi Zhang, Ziheng Zhou, Zhaoxin Liang

Probability

Recurrence and transience of random walks with drift $\rho x^{\alpha}/t^{\beta}$

Menshikov and Volkov [Electron. J. Probab. 13 (2008)] studied recurrence and transience of a class of Markovian random walks on $\mathbb R_+$ whose conditional drift depends on both time and position and is of order $\rho x^\alpha t^{-\beta}$ with $\rho>0$. The case on the critical line $2\beta-\alpha=1$, with $\alpha\in(-1,1)\setminus\{0\}$, remained open. We prove recurrence in this remaining case. Furthermore, we establish recurrence and transience criteria that complete the classification for $-1<\alpha<1$ and $\beta\ge 0$, without assuming the Markov property and under weaker assumptions on the increments than those imposed by Menshikov and Volkov.
Read the full abstract

Representation Theory

Rankin–Selberg integrals of opposite conductor–one newforms

Let $F$ be a nonarchimedean local field of characteristic zero and let $n\geq2$. For $r=n,n+1$, let $\Pi_r$ be an irreducible tempered representation of ${\rm GL}_r(F)$ of conductor one and with trivial central character. We evaluate the Rankin–Selberg integral of opposite newforms in $\Pi_{n+1}\times \Pi_n$ explicitly and show that its central value is nonzero. As an application, this implies a case of Disegni–Zhang's conjecture on the nonvanishing of local relative characters.
Read the full abstract

Hardware Architecture

BEACON: A Versatile Accelerator for Computational Pathology Applications

While accelerators for AI have seen great commercial success, it is challenging to replicate that success for other specialized domains due to a number of factors. We make the case that barriers for new accelerators can be lowered by starting with a baseline AI accelerator, and adding minimal logic to support new operators demanded by new specialized domains. This leads to a versatile chip that can be manufactured at high volume and deployed for a range of popular applications. We refer to this as the AI+X approach. This paper explores its potential for the emerging domain of Computational Pathology, which involves analysis of large whole-slide tissue images with a multi-stage pipeline. The pipeline requires support for a number of different kernels and operators - early stages perform segmentation and feature extraction, followed by graph creation with k nearest neighbor (kNN) algorithms, and finally inference with an iterative graph convolutional network (GCN) that alternates between Aggregation and Combination. We show that these stages execute inefficiently on a range of baseline CPU, GPU, AI, and GCN accelerators. That inefficiency is addressed with a combination of software re-structuring and small modifications to a baseline systolic AI accelerator. Many of the above kernels can be mapped to a systolic accelerator by offering a flexible datapath between processing elements and register access mechanisms. We add support for feature aggregation, load balanced execution, Euclidean distance calculation, binning, and counter aggregation. This additional flexibility and logic grows the area of a baseline AI chiplet by 1.1x, but by avoiding the memory wall and offering high parallelism, the proposed accelerator BEACON yields over an order of magnitude higher throughput for Computational Pathology than baseline CPU and GPU platforms.
Read the full abstract

All authors Sumanth Gudaparthi, Ananth Krishna Prasad, Lin Jia, Rajeev Balasubramonian, Srinivasan Parthasarathy

Robotics

LTLDiff: Finite Linear Temporal Logic-Guided Data Generation and Diffusion Policies for Multi-agent Robotic Manipulation

Multi-agent robotic manipulation tasks require coordination among agents to satisfy task-level temporal, logical, and safety constraints. Recently, diffusion policies have been used to perform the task. However, they still suffer from desynchronization, incorrect action ordering, and coordination failures in tasks that require simultaneous or sequential multi-agent interaction. Therefore, LTLDiff is proposed as a framework that combines Finite Linear Temporal Logic (LTLf) specification learning for both the generation of demonstrations and learning via diffusion policies. Each task has a specific LTLf formula that is learned from a set of natural language instructions using a large-scale language model. To enable a fixed-dimensional vector embedding of the learned specification from the language model, LTLf uses an abstract syntax tree representation scheme. This embedding of logic serves as a condition for (i) logic-guided data collection and (ii) diffusion-based policy training, encouraging trajectories that are consistent with the desired ordering and coordination requirements. Experiments on multi-agent LTLDiff manipulation tasks demonstrate improved task success rates compared to the baseline. Together, these contributions demonstrate the effectiveness of LTLDiff for coordinated multi-agent manipulation.
Read the full abstract

Machine Learning

T1: Terminal Agent Reinforcement Learning for Long-Horizon Tasks

Agent usage is shifting toward long-horizon tasks such as coding and scientific discovery, among which terminal tasks are especially important. We introduce T1, a Mixture-of-Experts model of 122B total trained with reinforcement learning, operating a real shell in a cloud sandbox for up to 300+ tool-call turns per task, rewarded by executing each task's own verifier. We provide a comprehensive recipe: First, an aggressively warm-started to stabilize actor-critic training, with a dense process reward scoring trajectories by the absolute number of passing verifiers. Second, stable optimization through TITO construction, training on the exact sampled token identifiers with drift repair at turn boundaries, and rollout routing replay, recording the sampler's per-token expert choices at every MoE layer and replaying them during training. Third, fully out-of-distribution training corpus: isolated seeds and synthesized tasks disjoint from Terminal-Bench 2.1 ensures gains reflect genuine capability transfer over benchmark overfitting. Together, TITO and R3 cut the training-to-inference log-probability difference from 0.021 to 0.013, with exactly aligned zero token drift in the loss region. On Terminal-Bench 2.1, our post-train pipeline raises initial base model from 43.8% to T1 with 64.0% resolved. On Long-Horizon Terminal Bench, T1 reaches 27.9% and surpasses GPT-5.4 and GLM-5.1.
Read the full abstract

All authors Junyao Yang, Yucheng Shi, Zhongzhi Li, Ruhan Wang, Zongxia Li, Haitao Mi, Leowei Liang

Computer Vision

Meta-Learning for Classifier Selection in Image Datasets: A Feature-Driven Framework for Accuracy Prediction

No Free Lunch theorem implies that any performance gains achieved by a classifier on a particular image distribution are necessarily offset by a loss of performance over the set of all possible problems; thus, no single model is universally optimal. Selecting the most suitable classifier for image datasets is a critical yet challenging task due to the intrinsic complexity and diversity of images. This paper proposes a meta-learning framework that leverages a comprehensive set of meta-features capturing dataset complexity to predict classifier performance without exhaustive training. By extracting and selecting features using methods such as autoencoders, pre-trained networks, and dimensionality reduction techniques, we train regression models to efficiently estimate classifier accuracies. Additionally, clustering techniques are employed to group classifiers with similar performance patterns, simplifying the recommendation process. The datasets used span a wide range of concepts, including nature, animals, numbers, motorcycles, medical images, and human bodies, to ensure broad generalization. Evaluated on 56 diverse image datasets, our approach achieves an average ranking prediction accuracy exceeding 86%, demonstrating its effectiveness in guiding model selection. This scalable and interpretable framework provides a practical solution to improve classification performance while reducing computational costs.
Read the full abstract

All authors Zahra Nabizadeh_Shahre_Babak, Farzaneh Koohestani, Nader Karimi, Shahram Shirani, Shadrokh Samavi

Computer Vision

Toward Interpretable Multimodal Fusion: Heat Conduction Modeling for Hyperspectral and LiDAR Joint Classification

The fusion of hyperspectral (HS) and Light Detection and Ranging (LiDAR) data plays a crucial role in enhancing land-cover classification by jointly exploiting spectral, spatial, and structural cues. However, existing multimodal fusion methods still struggle to model long-range dependencies and complex anisotropic interactions while maintaining computational efficiency. This paper introduces M2Heat, a physics-inspired framework that investigates multimodal fusion through the lens of heat conduction. At its core, a physics-driven visual heat conduction module (vHeat) and enhanced Frequency Value Embeddings (FVEs) simulate anisotropic information flow, enabling the capture of global dependencies with sub-quadratic complexity and physical interpretability. This mechanism, combined with a hybrid spatial-frequency fusion strategy named Cross-Frequency Fusion (CFF) module, produces highly discriminative and robust feature representations. M2Heat achieves competitive overall performance on three benchmarks, i.e., Trento, Houston2013, and Augsburg, while providing an interpretable heat-conduction-guided perspective for multimodal feature fusion. These results indicate the potential of heat-conduction-guided neural operators for efficient and interpretable RS multimodal fusion. The source code is publicly available at https: /github.com/Weikan0425/M2Heat_HSI_LiDAR.
Read the full abstract

All authors Kan Wei, Jiahui Cui, Jing Yao, Xinyu Zhao, Lei Wang, Pedram Ghamisi

Probability

Uniform-in-Time Boltzmann Mean-Field Limits for Anchored Binary Opinion Dynamics

We study a continuous-time opinion model in which agents interact in pairs and each agent has a fixed anchor. At each interaction, both opinions are updated according to a possibly nonlinear rule with random inputs. Under suitable stability and moment assumptions, we establish uniform-in-time propagation of chaos: as the population grows, any fixed number of agents become asymptotically independent, with a common law solving a nonlinear Boltzmann equation. The proof combines a graphical approximation on finite time intervals with exponential convergence of the finite system and its nonlinear limit to their respective stationary laws. We examine two applications. For an anchored variant of the Friedkin-Johnsen model with random coefficients, we give conditions for sub-Gaussian or power-law tails of stationary deviations from the anchors. Occasional overreaction can produce power-law tails with Gaussian anchors and noise, even as the nonlinear law converges exponentially fast to stationarity. We also study an anchored model of biased assimilation, determining when a nearby opinion moves toward or away from neutrality with the anchor and incoming evidence held at neutral values. Simulations illustrate the tail predictions and show how a single peak in the smoothed empirical opinion profile can split into two.
Read the full abstract

Numerical Analysis

A Second-Order Method for Locating Critical Points of Prescribed Index on Energy Landscapes

Exploring high dimensional energy landscape is a challenging problem in many applications where both ground states and transition states offer important information about the underlying physical systems. To facilitate the computational exploration, we propose a cubic-regularized second-order method for locating critical points of prescribed index, including both energy minima (with an index $0$) and transition states (saddle points with positive index). At each step of the iteration, the proposed method involves several ingredients, such as updating the eigen-directions of the Hessian according to the prescribed index, reflecting the gradient and the Hessian along selected unstable eigen-directions, if any, and constructing the next step by solving a cubic-regularized subproblem. Under suitable assumptions, we show that, when the eigenspace and the cubic subproblem are solved exactly, the iteration converges locally and quadratically to a critical point of the prescribed index. We further establish convergence results in the inexact setting, where both the eigenspace computation and the solution of the cubic subproblem are performed approximately. We show that the resulting iteration retains local linear convergence, with the convergence rate depending on the accuracy of the inexact computations. Moreover, we prove that the fixed points of the algorithm are precisely critical points of the prescribed index. In addition, we introduce an adaptive strategy for updating the cubic regularization parameter to improve robustness and computational efficiency of the proposed algorithm. Numerical experiments confirm the predicted convergence behavior, demonstrate the effectiveness of the adaptive strategy, and illustrate the capability of the proposed method for locating critical points and constructing solution landscapes.
Read the full abstract

Number Theory

Ordinary 3-Isogeny Graphs and Improvement of Supersingularity Testing for Twisted Hessian Curves over Prime Fields

For any primes $p \neq \ell$, $\ell$-isogeny graphs of ordinary elliptic curves defined over $\mathbb{F}_{p^2}$ have a typical structure called $\ell$-volcanoes, and the structure is the core of Sutherland's supersingularity testing algorithm for elliptic curves. In this paper, by exploiting the properties of $3$-isogenies between twisted Hessian curves, we show that when $p \equiv 2 \pmod{3}$ and $\ell = 3$, every ordinary twisted Hessian curve defined over $\mathbb{F}_p$ lies on the surface of the $3$-volcano. As an application, we give an improved version of Sutherland's supersingularity testing algorithm specialized to twisted Hessian curves defined over $\mathbb{F}_p$ with $p \equiv 2 \pmod{3}$. We also give a generalization of the known fact that any supersingular $j$-invariant is a cube in $\mathbb{F}_{p^2}$; we show that for any twisted Hessian curve $H(a,d)$ defined over $\mathbb{F}_{p^2}$, its $j$-invariant is not a cube in $\mathbb{F}_{p^2}$ if and only if $H(a,d)$ is ordinary and lies on the floor of a $3$-volcano.
Read the full abstract

Soft Condensed Matter

Slow Dynamics and the Geometry of Jammed Packings

Saddle points in the energy landscape of granular packings dominate the discrete steepest descent dynamics and ultimately determine the path that an out of mechanical equilibrium packing will follow and the resulting stable minimum that it will find. The saddle points that ultimately determine the resulting minima tend to be low-index saddle points. For models with an analytic energy landscape, such as the $p$-spin model, the steepest descent minimization path is affected by higher-index saddle points, which pull the system towards saddle points of decreasing index before arriving at the minima. Here, we examine the steepest descent minimization path of granular packings and compare them to the $p$-spin model. We show that the granular packing steepest descent minimization paths act like their smooth energy landscape counterparts and get attracted by saddle points. The index versus time curves for all models follow a shifted, stretched exponential. We further show that the shape parameter for the granular packings is unchanged when the energy landscape is modified to become analytic (Gaussian potential in a harmonic well) or non-local (Mari-Krzakala-Kurchan). The $p$-spin, on the other hand, has a significantly larger shape parameter. The reason is not due to the dimensionality, packing fraction, nonanalyticity, or the locality of the Hamiltonian of the models. The exact reason for the discrepancy in the shape parameter is \st{still} an unsolved mystery.
Read the full abstract

Probability

Moments approach for the elephant random walk

We discuss the method of moments for the one-dimensional elephant random walk (ERW). We first derive a differential recurrence relation for the characteristic function of the ERW, which yields a corresponding system of recurrence relations for its moments. We then obtain asymptotic approximations for the moments in each of the three parameter regimes of the ERW. Finally, by establishing the convergence of the moments and verifying the corresponding moment-determinacy conditions, we identify the limiting distributions of the ERW in each regime.
Read the full abstract

Quantum Physics

Engineered two-photon dissipative confinement of a Kerr-cat qubit using SISIS quantum circuit refrigerator

Kerr-cat qubits realized in periodically driven superconducting nonlinear resonators are a promising platform for quantum information processing with biased noise. Pure dephasing in such systems induces leakage out of the qubit subspace, motivating the use of quantum circuit refrigeration (QCR) to remove excess excitations. While conventional superconductor–insulator–normal-metal–insulator–superconductor (SINIS)-based QCRs can suppress leakage via single-photon absorption, they also enhance QCR-induced phase-flip errors. Here we investigate a QCR based on a superconductor–insulator–superconductor–insulator–superconductor (SISIS) junction coupled to a Kerr parametric oscillator (KPO). We show that a SISIS-based QCR can operate in a regime where single-photon processes are suppressed while two-photon absorption dominates. As a result, the proposed SISIS-based QCR achieves strong suppression of dephasing-induced leakage while substantially reducing the increase in phase-flip errors associated with QCR operation. These results demonstrate that the proposed SISIS-based QCR provides an effective approach for mitigating leakage while limiting QCR-induced phase-flip errors in Kerr-cat qubits.
Read the full abstract

All authors Shumpei Masuda, Tsuyoshi Yamamoto, Shuji Nakamura, Daichi Sugiyama, Akiyoshi Tomonaga

Signal Processing

Load Balancing in Multi-Shell LEO Satellite Networks with Successive Interference Cancellation

Multi-shell low Earth orbit (LEO) networks can increase service opportunities, but altitude-dependent propagation can concentrate traffic on lower shells and create strong inter-shell interference under full frequency reuse. This paper develops a mathematical framework for load balancing in multi-shell LEO satellite networks. Satellites on each shell form an independent spherical Poisson point process (SPPP), and the typical user associates with one of the per-shell serving satellites through a shell-dependent biased received-power rule, with receiver-side successive interference cancellation (SIC) under full frequency reuse. Shell-wise association probabilities, conditioned serving-distance distributions, and the rate coverage probability under shell-dependent traffic loads are derived and validated by simulation. The results show that shell-dependent biasing alleviates lower-shell traffic concentration and improves rate coverage, while receiver-side SIC mitigates the dominant lower-shell interference experienced by users associated with upper shells. Load balancing provides its largest rate-coverage gain in traffic hotspots, while SIC becomes more valuable as receive-side isolation weakens. With a fixed satellite budget, distributing satellites across multiple shells can further improve hotspot rate coverage by adding shell-wise serving opportunities.
Read the full abstract

Statistical Methodology

Bayesian Variable Selection for High-Dimensional Predictors with Missing Psychometric Outcomes

High-dimensional, multimodal predictors and partially observed multivariate outcomes are common in psychometric research. However, existing regularization methods often do not accommodate hierarchical predictor structures and are primarily designed for univariate outcomes. We propose SHIM, a Bayesian framework for structured variable selection that combines hierarchical horseshoe shrinkage with a Bayesian treatment of missing outcomes. The framework jointly accommodates predictor hierarchies, dependence among outcomes, and incomplete multivariate responses. We establish theoretical properties of the proposed prior specification and evaluate SHIM through simulation studies. The results demonstrate that SHIM balances sensitivity with false-positive control while yielding accurate coefficient estimates and well-calibrated uncertainty quantification. We further apply SHIM to data from an Alzheimer's disease cohort to characterize associations between multimodal neuroimaging measures and multivariate neuropsychological outcomes and to generate posterior-based multiple imputations for downstream analyses of the relationships between fluid biomarkers and cognition. An R package, shim, is publicly available to facilitate implementation.
Read the full abstract

All authors Zongyue Teng, Shujie Ma, Timothy J. Hohman, Angela L. Jefferson, Panpan Zhang

Solar and Stellar Astrophysics

A prolonged plateau-to-tail transition in the Type II supernova SN2025abyc

We present optical photometric and spectroscopic observations of the Type II supernova SN2025abyc. During the optically thick phase between approximately 10 and 70 d after explosion, its light curves show strongly wavelength-dependent decline rates of approximately 2.7, 2.1, 0.9, and 0.8 mag/100d in the g, c, r, and o bands, respectively. At approximately 70 d, the light curves begin to depart from their nearly linear plateau evolution and gradually transition toward the radioactive tail. A Fermi-Dirac fit to the well-sampled ATLAS o-band light curve yields a transition midpoint of t_PT ~ 100.5d. The interval between the end of the linear plateau and this transition midpoint is approximately 30 d, indicating a prolonged plateau-to-tail transition. This timescale is comparable to those measured for SN2013by, SN2013ej, and SN2014G. Spectroscopically, at +13 d post-explosion, the Halpha profile appears weak and broad, whereas Hbeta and Hgamma display clear P-Cygni profiles. This morphology can be explained by the normal early spectroscopic evolution of SNe II, although partial filling of the Halpha absorption trough by emission associated with circumstellar interaction cannot be excluded. SN2025abyc otherwise follows the general photospheric velocity evolution of SNe II, while remaining toward the high-velocity side of the comparison distribution in Halpha, Hbeta, and FeII. Exploratory light-curve modelling suggests a synthesized Ni mass of approximately 0.03-0.04 solar mass. We suggest that the extended circumstellar environment, Ni distribution, and hydrogen-envelope structure could all play a role in shaping the observed light-curve evolution, particularly the prolonged plateau-to-tail transition.
Read the full abstract

All authors Luhan Li, Bo Wang, Jujia Zhang, Zhengyang Zhang, Xinjie Luo, Shiyang Dong, Saien Xu, Zhengwei Liu, Zhanwen Han

Artificial Intelligence

The Agent Incident Registry: Toward Preventing Repeated AI Agent Failures

AI agents increasingly act through tools and delegated authority, but general incident repositories rarely capture the mechanisms needed to compare public failures with agent-security evaluations. We present the Agent Incident Registry (AIR) (Project page: https://enkryptai.com/air), a source-linked catalog containing 487 records of agent-related events disclosed from 2022 through 2026. Each record includes supporting evidence, a stable identifier, and missingness-aware labels for causal role, disclosure class, mechanism, and outcome. Among the 336 generative-system records in which the agent acted, 81 involved realized harm (24%). Realized outcomes concentrate in in-the-wild and safety-failure records, while responsible disclosures and research demonstrations are overwhelmingly demonstrated; the aggregate share therefore characterizes collection composition rather than deployment risk. After initial curation, a second human reviewer checked all 487 records and their existing labels for completeness and correctness. In a deployment-analogue audit, InjecAgent's 1,054 cases occupy three of AIR's twelve surfaces and are all attacker-triggered, whereas AIR contains 92 no-adversary safety failures. AIR supports source-grounded case retrieval and evaluation-scope auditing, not failure-rate or control-efficacy estimation.
Read the full abstract

All authors Divyanshu Kumar, Rohith HN, Nitin Aravind Birur, Sahil Agarwal, Prashanth Harshangi