Robotics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 3 September
Autonomous unmanned aerial vehicles (UAVs) increasingly operate in cluttered environments where global planners such as RRT* are not directly deployable at control rates. This paper presents a real-time local planning and obstacle avoidance module for an omnidirectional multirotor (omnicopter) by extending the Dynamic Window Approach to six degrees of freedom (6D-DWA). Our method achieves real-time feasibility through (i) local-map voxelisation, (ii) a compact sphere-based approximation of the vehicle geometry, and (iii) adaptive velocity sampling in the 6D search space. To improve reactivity to unknown obstacles, we introduce a context-aware "Agile Mode" that adjusts scoring weights online to trade-off between goal progress, clearance, and heading/facing constraints during evasive manoeuvres. We evaluate our approach in simulation across computational stress tests, dense-waypoint path tracking, and static/unknown obstacle scenarios. Our planner runs consistently within a 0.2s control loop, tracks waypoint-dense global paths with < 0.1m average cross-track error and 13deg average heading error, and avoids collisions in static environments. For unknown obstacle avoidance, Agile Mode achieves 79.3% success for an off-centre obstacle and 41.4% for a centred obstacle, highlighting both the effectiveness of adaptive weighting and remaining limitations in highly constrained geometries.
Computer Vision
EraseSAE: Surgical Concept Erasure in Text-to-Video Diffusion Models via Sparse Autoencoders
Recent advances in text-to-video (T2V) diffusion models have demonstrated remarkable generative capabilities, yet their reliance on loosely curated training data raises pressing safety and copyright concerns. Concept erasure offers a principled remedy by removing unwanted semantics from pretrained models while preserving remaining concepts. However, existing approaches typically operate at a coarse granularity misaligned with the fine-grained, distributed nature of concept representations, leading to incomplete removal or degraded generation quality. We argue that surgical erasure fundamentally requires intervention at the level of monosemantic features, where each unit encodes a single interpretable concept. To this end, we propose EraseSAE, a novel framework that leverages sparse autoencoders to achieve surgical concept erasure in DiT-based T2V diffusion models via a principled decompose-attribute-erase pipeline. We first introduce the Partitioned Convolutional Sparse Autoencoder, which decomposes dense spatiotemporal activations into disentangled, interpretable sparse features while preserving spatiotemporal coherence. A contrastive attribution mechanism then contrasts activations from paired prompts to isolate concept-specific feature kernels. At inference, timestep-resolved spatiotemporal masks derived from the identified kernels confine erasure to regions where the target concept is active, leaving unrelated content intact. Extensive experiments across diverse diffusion models and concept erasure tasks demonstrate that EraseSAE achieves precise and robust concept removal with minimal quality degradation, substantially outperforming state-of-the-art methods. The code is available at https://github.com/HiDream-ai/EraseSAE.
Read the full abstract
Computational Complexity
Random Garbage Creates Exponential Quantum Query Separations
We prove an exponential separation between standard (XOR) quantum queries and forward-erasing queries for a Boolean property of injective functions, answering a question of Aaronson~\cite{Aaronson2021Open}. We also obtain an exponential separation for a Boolean promise problem on permutations, where forward in-place queries are unitary but inverse queries are not supplied. Our construction follows Aaronson's suggestion of appending fixed random garbage to a Simon function. Let $X=\{0,1\}^n$ and $N=2^n$. We use the injection $f_{h,r}(x)=(h(x),x,r_x)$, where $h:X\to X$ is promised to be either a permutation or a Simon function, and $r:X\to X$ is an arbitrary tag table that is reused on every query. The resulting problem is solvable with at most $n+2$ standard queries, but requires $\Theta(\sqrt N)$ forward-erasing queries, matching classical collision sampling. Embedding these instances into permutations, with the copy of $x$ replaced by a single bit that keeps the embedding injective, gives a promise problem on permutations of $L=4N^2$ elements that requires $O(\log L)$ standard queries versus $\Theta(L^{1/4})$ forward in-place queries. The lower bounds use the recording method. We replace each query by a nearby contraction that records cells of a purified tag table. After $T$ queries, tracing out the purified tag table yields a sum of positive-semidefinite contributions, each depending on $h$ at no more than $T$ addresses. A query-by-query hybrid bounds the trace-norm difference between the actual and auxiliary final operators by $O(T/\sqrt N)$. Random permutations and random Simon functions have the same restriction distribution on such a set unless it contains a hidden Simon pair, an event of probability $O(T^2/N)$. Consequently, the actual final states in the two cases remain close unless $T=\Omega(\sqrt N)$, giving the required lower bounds.
Read the full abstract
Soft Condensed Matter
Pseudospectral Methods and Critical Phenomena
We employ pseudospectral methods to solve the homogeneous Ornstein-Zernike (OZ) equation for a model fluid in the vicinity of the critical point. Focusing on the Mean-Spherical Approximation (MSA) as a closure to the OZ equation, we obtain numerical estimates for the critical exponents $\eta$, $\delta$ and $\gamma$ for a system of hard-core Yukawa particles both in two and three dimensions. The three-dimensional MSA exponents are already well-known from an analytic solution and are recovered by our numerical methods. The two-dimensional exponents are a new output of this work. The pseudospectral method allows for rapid and highly accurate solution of liquid-state integral equation theories, and enables calculations on truely infinite domains, as needed for highly correlated states. In addition, we analyse the standard Picard iteration scheme and propose a variation of it which provides increased stability and speed of convergence.
Read the full abstract
Numerical Analysis
Residual neural networks overcome the curse of dimensionality for semilinear heat equations
Rigorous results show that feedforward neural networks can overcome the curse of dimensionality in the numerical approximation of high-dimensional partial differential equations (PDEs), but comparatively little is known about residual neural networks (ResNets) in the nonlinear PDE setting. We prove that ResNets overcome the curse of dimensionality in the numerical approximation of solutions of semilinear heat equations with globally Lipschitz continuous, gradient-independent nonlinearities: under polynomial growth and network approximability hypotheses on the PDE data, there exist $\eta\in(0,\infty)$ and ResNets $\Psi_{d,\varepsilon}$, $d\in\mathbb{N}$, $\varepsilon\in(0,1]$, with at most $\eta d^{\eta}\varepsilon^{-\eta}$ parameters whose realizations approximate the solution in dimension $d$ with an $L^2$-error of at most $\varepsilon$. The proof represents one deterministic realization of a multilevel Picard estimator by a ResNet whose shortcut connections transmit the spatial variable and a scalar accumulator, while the residual branches successively add the summands of the estimator. For ridge-sum initial conditions, admissible sigmoidal activations, and globally Lipschitz truncations of the nonlinearity, we obtain, for every $\xi>0$, the explicit bound $C_\xi d^{4+\xi}\varepsilon^{-(3+\xi)}$ on the number of parameters.
Read the full abstract
Quantum Physics
Quantum Quasi-Monte Carlo: a window for pre-asymptotic quantum advantage
Numerical integration with Monte Carlo methods is a central computational task in many scientific and industrial applications, including financial derivative pricing and risk management. Classical Monte Carlo algorithms are computationally demanding: achieving an accuracy $\epsilon$ typically requires a number of function evaluations scaling as $O(1/\epsilon^2)$. Quantum-accelerated Monte Carlo methods based on quantum amplitude estimation can in principle quadratically improve this dependence. However, quasi-Monte Carlo methods have not been explored in the quantum context. In this work, we introduce a quantum quasi-Monte Carlo algorithm that combines low-discrepancy nets with quantum amplitude estimation. The proposed method prepares the quasi-random point set coherently in superposition. The method does not yield an asymptotic improvement over classical quasi-Monte Carlo, since the total error separates into a discretization error, determined by the finite net, and a quantum estimation error. Instead, we explore a pre-asymptotic advantage window: for a target accuracy that would classically require $2^q$ low discrepancy points, one can prepare a higher-resolution net of size $2^Q$, with $Q>q$, in superposition and reach the same accuracy using significantly fewer function queries. This window can be controlled by tuning the circuit resolution and amplitude-estimation parameters, making the approach relevant for practical regimes where the number of queries is finite rather than asymptotically large.
Read the full abstract
High Energy Astrophysical Phenomena
NuSTAR View of the 2025 Mini-Outburst of the Black Hole X-ray Binary GRS 1739-278: Spectral and Timing Evolution in the Soft State
We present a dedicated timing and spectral analysis of three NuSTAR observations of the Galactic black hole candidate GRS 1739-278 obtained during the decay of its 2025 mini-outburst (September 18-30, 2025). The source is found in a soft, disk-dominated state throughout, with a declining $3$-$79$ keV count rate (from $112.4$ to $99.4$ cts s$^{-1}$), a low and only mildly evolving hardness ratio ($\approx0.14$-$0.16$) and a fractional variability that drops from $3.2\%$ to $1.1\%$; no periodic or quasi-periodic signal is detected in the power spectrum or the Lomb-Scargle periodogram of any observation. Joint FPMA+FPMB spectral fits with an absorbed disk-blackbody plus thermal-Comptonization continuum, together with a broad, high-significance ($11.7$-$15.0\sigma$) curvature feature near $10$-$12$ keV (an empirical proxy for a strong disk-reflection/returning-radiation component) describe the $4$-$30$ keV spectra well ($\chi^2/{\rm dof}=1.04$-$1.15$). The inner-disk temperature cools steadily from $1.021\pm0.001$ to $1.002^{+0.003}_{-0.002}$ keV while the disk normalization and hence the derived inner radius, $R_{\rm in}\simeq27.7$ km $\simeq0.95\,R_{ISCO}$ remains constant to within $0.3\%$, consistent with a disk anchored at the innermost stable circular orbit throughout. The unabsorbed $4$-$30$ keV luminosity declines from $1.35\times10^{37}$ to $1.15\times10^{37}$ erg s$^{-1}$ ($0.67\%$ to $0.57 L_{Edd}$). These results of the spectral-timing characterization of the 2025 mini-outburst are broadly consistent with a high-spin black hole accreting at a low, steadily declining rate in the canonical soft state.
Read the full abstract
Robotics
QLAUN: A Research-Oriented, Robust, Agile, Modular, and Affordable Torque-Controlled Quadruped Robot
QLAUN Bot (Quad-Legged Adaptive Unmanned Navigator Robot) is a torque-controlled quadruped robot that is research-oriented, cost-effective, and aimed at achieving simultaneous robustness and agility while being completely 3D-printed. It is a quadruped robot that is aimed at empowering robotics research at universities and research institutes in Lebanon and the MENA region. Using a novel electronics-free leg design strategy, we present a modular robot with interchangeable and easily replaceable legs. The 15 kg robot possesses 12 DoF (Degrees-of-Freedom) with three per leg, each paired with a completely 3D-printed Quasi-Direct Drive (QDD) actuator that consists of a brushless DC motor and a low-ratio gearbox transmission that is connected to a belt transmission system for significantly increasing the torque outputs at the joints. We present legs that have decoupled hip and knee actuators to improve the overall modularity of the robot. QLAUN is almost completely 3D-printed using polylactic acid (PLA) and assembled using off-the-shelf parts to create a robust, agile, and affordable robot for legged robot locomotion research. The legs possess joints with wide ranges of motion, including a continuous hip flexion-extension joint. A compliant foot, printed using TPU-95A is also implemented for alleviating hard impacts and handling terrain uncertainties. This extended abstract aims to introduce QLAUN, a novel platform for robotics research, emphasizing the design concepts and principles that underpin its development to the academic and research communities in the field of robotics.
Read the full abstract
Sound
Test-time adaptation for speech enhancement with an autoregressive speech prior
Test-time adaptation (TTA) offers a promising direction for improving speech enhancement models under mismatched acoustic conditions, without requiring access to labeled target data. In this work, we propose a single-utterance TTA method that regularizes a pretrained speech enhancement model using an autoregressive prior trained on clean speech latent representations extracted from a neural audio codec. Adaptation is performed by minimizing the Kullback-Leibler divergence between the enhanced speech distribution and the clean speech prior. Experiments across multiple noisy speech datasets show consistent improvements in speech quality, particularly under training-testing noise mismatch conditions. Code and audio examples are available online.
Read the full abstract
Artificial Intelligence
A computable representation of the physical laboratory enables verifiable workflows
Making science computable requires representations of both scientific knowledge and the physical world in which scientific claims are tested. A computable representation of the physical laboratory is established through typed research objects, capability-bound operations and a compositional workflow algebra. It provides the physical-world counterpart to machine-readable knowledge, expressing workflows as programs over evolving laboratory states with explicit dependencies, decisions, iteration and concurrency. The representation was implemented in a modular agentic robotic laboratory by binding formal operations to executable Function Skills. For diverse scientific intents, capability-relative workflows were generated, while stateful simulation propagated object transformations and verified operation preconditions and laboratory constraints before dispatch. The proposed representation and its engineering framework jointly establish a general computational interface between agent reasoning and capability-bound physical transformations, providing a foundation for end-to-end autonomous scientific discovery.
Read the full abstract
Audio and Speech Processing
ToolDF: Tool-Integrated Reasoning for Mixed-Authenticity Audio Deepfake Detection
Audio deepfake detection is commonly formulated as clip-level binary classification of single-domain audio. However, real-world manipulated audio can exhibit mixed authenticity, where genuine and manipulated cues coexist across temporal transitions, overlapping sources, or both. This setting requires not only detecting manipulated audio but also localizing the components that provide evidence for the decision. We propose ToolDF, a tool-integrated reasoning framework for mixed-authenticity audio deepfake detection. ToolDF employs an audio large language model as an orchestrator trained with supervised tool-use trajectories. It adaptively analyzes the audio scene, selectively performs source separation, routes components to domain-specific experts, and aggregates their evidence into an interpretable verdict. We further introduce a mixed-authenticity ADD benchmark covering temporal transitions, acoustic overlaps, and hybrid mixtures. Experimental results show that ToolDF achieves the best overall performance on composite-type detection, achieving macro-F1 gains of 3.72 and 14.39 points over the strongest monolithic baseline and a fixed pipeline, respectively, while providing interpretable evidence localized to temporal regions and acoustic sources. Our source code and dataset are publicly available online.
Read the full abstract
Computation and Language
Remember and Reweight: Enhancing Multi-Agent Debate with Experience Memory and Confidence Estimation
Multi-agent debate (MAD) improves the reasoning capabilities of large language models by having multiple agents iteratively refine their responses through discussion. However, MAD suffers from a critical vulnerability known as shared misconception: when a majority of agents initially converge on an incorrect answer, the debate process tends to amplify rather than correct the error. Existing methods primarily address peer skew but leave the agents' inherently biased concept priors unaddressed. To mitigate this systematic weakness, we propose R$^2$-MAD (Remember and Reweight for Multi-Agent Debate), a framework that equips agents with an experience memory accumulated from past debates. R$^2$-MAD intervenes on both failure modes through two complementary mechanisms: A debate-state-aware retrieval policy dynamically calibrates the concept prior by retrieving relevant historical evidence based on the current consensus level. Then these retrieved experiences provide a basis for estimating per-agent reliability, yielding confidence weights to modulate peer influence. Experiments on various benchmarks show that R$^2$-MAD achieves consistent improvements over existing single-agent and MAD baselines.
Read the full abstract
Disordered Systems and Neural Networks
Spectral properties of deterministic matrices multiplied by rotationally invariant random non-Hermitian ensembles
In this paper, we study spectral properties of multiplicative deformations of non-Hermitian random matrices. We consider matrices of the form $\mathbf{A}\mathbf{B}$, where $\mathbf{A}$ is a deterministic $N\times N$ matrix (not necessarily Hermitian) and $\mathbf{B}$ is a rotationally invariant random matrix. We show that, as $N\to\infty$, the boundary of the complex eigenvalue distribution of $\mathbf{A}\mathbf{B}$ is governed by simple equations involving the $\mathcal{R}_1$ and $\mathcal{R}_2$ transforms of $\mathbf{B}$.
Read the full abstract
Optimization and Control
On the $\Omega(n)$ SDP relaxation gap for Submodular Box-Constrained Quadratic Programming
Continuous submodular optimization is an important class of nonconvex problems with global guarantees, for which the submodular box-constrained quadratic programming (BCQP) forms a fundamental subclass. It arises naturally as quadratic subproblems of general twice-differentiable submodular problems and also has direct business and management applications such as pricing. This paper studies the worst-case relaxation gap of semidefinite programming (SDP) relaxations with a finite family of valid linear cuts for submodular BCQP. We show that in dimension $n\geq4$, for any SDP relaxation formulated with a finite family of linear cuts, there always exists a submodular BCQP instance with a strictly positive relaxation gap. By restricting to the normalized BCQP instances that remove the problem's scaling factor, we further quantify this limitation with explicit lower bounds. In general dimension $n\geq4$, when the SDP relaxation is formulated with only Boolean-quadric-polytope (BQP) valid cuts, including the commonly used McCormick and triangle inequalities, we establish an $\Omega(n)$ relaxation gap lower bound. For an arbitrary family of $m$ valid linear cuts, we establish an $\Omega(n/m^2)$ lower bound. These results capture the fundamental limitations of SDP relaxations, especially SDP with BQP valid cuts, and provide guidance on what additional constraints may be needed for more effective convex approximations of submodular BCQP.
Read the full abstract
Optimization and Control
Nash-Bargaining HalpernSGD via Limited-Memory MultiLRSGA: A Two-Phase Optimizer for Multi-Objective Learning
We propose NB-HalpernSGD via LM-MultiLRSGA, a two-phase optimizer for multi-objective optimization. The process starts with a competitive optimization phase by applying a limited-memory variant of MultiLRSGA, named LM-MultiLRSGA, designed to reduce the memory footprint of the original optimizer while preserving its rotational correction mechanism. It addresses multi-objective tasks by considering the competitive game associated with the original multi-objective problem, approximating a Nash equilibrium. This point defines a Nash-equilibrium-induced disagreement point: we formulate the Nash bargaining problem associated with the original losses and rewrite its Nash product as a logarithmic minimization surrogate. This minimization problem is solved using HalpernSGD, anchored at the computed competitive reference point. Therefore, the method uses the Nash equilibrium as a principled reference point for the bargaining stage and then moves toward a Pareto-oriented solution of the original multi-objective problem. We validate the proposed optimizer on a PINN-inspired neural model, where it outperforms established multi-objective optimizers, including PCGrad, MultiAdam, and DualConeGD. Finally, while the convergence properties of HalpernSGD have been extensively studied, we discuss the convergence and stability properties of the proposed LM-MultiLRSGA phase.
Read the full abstract
Computer Vision
Auditing Patient Privacy in Medical Generative Models: Scalable Memorization Detection with DeepSSIM++
While deep generative models offer new opportunities for medical image synthesis and data sharing, their ability to memorize and reproduce training samples raises serious concerns about patient confidentiality. Detecting such memorization at scale remains challenging: traditional pixel-based metrics are sensitive to generation artifacts, whereas generic embedding-based metrics often lack the anatomical sensitivity required for medical data. To address this challenge, we introduce DeepSSIM++, a self-supervised similarity metric for scalable memorization auditing in medical generative models. By leveraging multi-scale feature aggregation and anatomy-preserving augmentations, DeepSSIM++ learns an embedding space where cosine similarity approximates the Structural Similarity Index (SSIM), eliminating the need for exact pixel-level registration. Compared with state-of-the-art baselines, DeepSSIM++ achieves an average Macro F1 improvement of 33 percentage points under ideal alignment and 46 percentage points under realistic spatial and intensity perturbations. Furthermore, it accelerates large-scale similarity computation by several orders of magnitude compared with analytical SSIM. By combining anatomical sensitivity and computational efficiency, DeepSSIM++ provides an open-source tool for scalable memorization auditing in medical generative AI. Code and data are publicly available at: https://github.com/brAIn-science/DeepSSIM.
Read the full abstract
Statistical Methodology
Change-point analysis: a new perspective for unstable financial markets
We introduce two new classes of nonparametric change-point tests for sequences of univariate non-negative random variables. The proposed procedures are based on the empirical modified Hankel transform and the Laplace transform, respectively, and provide new transform-based tools for detecting distributional changes. We derive the asymptotic null distributions of the corresponding test statistics and use a permutation bootstrap procedure to obtain $p$-values, since the limiting null distributions are not distribution-free. Through a finite-sample simulation study, we show that the proposed tests are well calibrated, generally more powerful than empirical-characteristic-function-based competitors, and capable of accurately estimating the change-point location, particularly in asymmetric settings. The usefulness of the proposed methodology is further demonstrated through applications to Argentina rainfall data, as well as U.S. GNP and S&P 500 absolute log-returns. The real-data examples illustrate that the proposed Hankel- and Laplace-transform-based tests are effective tools for detecting meaningful distributional changes in non-negative data.
Read the full abstract
Graphics
TileGS: Tile-Local Depth Binning for Gaussian Splatting Rasterization
Real-time 3D Gaussian Splatting (3DGS) achieves high rendering quality, but standard rasterization still traverses a globally sorted tile stream that creates long per-tile ranges and heavy geometry-attribute traffic. We present TileGS, a tile-local reorganization of Gaussian splatting. TileGS turns each long tile range into a sequence of shorter depth-local ranges, rasterizes those ranges in front-to-back order, and applies selective repair where coarse ordering is insufficient to match baseline compositing. Across a 9-scene benchmark on desktop and laptop Ada GPUs, our default No-GW (No Geometry-Write) variant delivers a mean 1.44x raster-kernel speedup on RTX 4090 and mean end-to-end frame speedups of 1.069x on RTX 4090 and 1.094x on RTX 1000 Ada over gsplat–a widely used optimized open-source 3DGS implementation–while matching the gsplat output up to numerical noise (|Delta PSNR| < 0.001 dB, |Delta SSIM| < 0.001, |Delta LPIPS| < 0.001). Full-suite RTX 4090 Nsight Compute profiling reveals TileGS is faster despite lower SM throughput, lower active-warp occupancy, and higher DRAM traffic, while total SASS thread instructions fall by 1.26x. Source-attributed profiling confirms that geometry attributes dominate the remaining memory pressure (85.8% of total raster traffic and 88.6% of excess sectors). Together, these counters support the interpretation that TileGS improves raster performance by reducing effective raster traversal work, rather than by reducing byte volume, improving coalescing, increasing occupancy, or directly reducing measured warp divergence.
Read the full abstract
Optimization and Control
The Traveling Tournament Problem: An Overview
Over the past 25 years, the Traveling Tournament Problem (TTP) has become one of the most extensively studied optimization problems in sports scheduling. At its core, the TTP seeks to minimize the total travel distance incurred by teams that travel directly between opponents' venues during consecutive away games. The problem originated from the scheduling challenges faced by Major League Baseball, where it was identified as the central computational difficulty. This paper provides a comprehensive overview of the literature on the TTP. We review the principal problem variants and benchmark instances, and summarize the current state of the art in lower bounds, approximation guarantees, and exact and heuristic optimization algorithms. Moreover, we contribute to the continued development of the field by tracking and validating lower and upper bounds, while succeeding the repository originally established by Prof. Michael Trick as part of the RobinX sports timetabling project. Finally, we identify several open questions and outline promising directions for future research.
Read the full abstract
Robotics
FailBench: How Reliable are VLMs at Judging Robot Task Success?
Vision-Language Models (VLMs) are increasingly used to evaluate robot manipulation outcomes, but existing benchmarks offer limited evidence of cross-domain generalization. We introduce FailBench, a benchmark for robot failure detection comprising 2,197 manipulation attempts across 14 public sources (12 real-world, 2 simulated). In FailBench, 75% of failures occur naturally, and six real-world sources come from non-failure-detection datasets. Evaluating 13 VLM-based detectors, we find the best model achieves only 0.77 mean balanced accuracy. Notably, models fine-tuned for failure detection consistently underperform general-purpose VLMs and their own pretrained baselines. Performance depends heavily on required visual evidence: models approach saturation when outcomes depend on observable object motion, but degrade to near-chance (<0.60 balanced accuracy) on contact-intensive assembly tasks. Error analysis reveals a systematic bias toward predicting success under ambiguous evidence, which persists even with increased reasoning effort. Finally, we show that input-level intervention–spatially localizing and cropping outcome-relevant regions–improves the top detector by 2.4 percentage points without extra training.
Read the full abstract
Dynamical Systems
Dynamical (non-)recurrence of escaping exponential maps
We show that exponential maps could have complicated measurable behaviours by giving two examples in the exponential family both with singular value escaping to infinity “slowly", one of which is recurrent (in the sense that every positive measure set will return to itself infinitely many times), while the other is non-recurrent. This also complements earlier results of Lyubich, Rees, Urbański-Zdunik, and Hemke. We also give an example of exponential maps with singular value escaping to infinity arbitrarily slow, which is a parametric analog of a result of Rempe.
Read the full abstract
High Energy Astrophysical Phenomena
KM3NeT/ARCA stacking search for high-energy neutrino point sources: the case of ultra-luminous infrared galaxies and blazars
A stacking search for high-energy astrophysical neutrino emission is presented using three candidate source populations: ultra-luminous infrared galaxies and two categories of blazars, specifically high-synchrotron-peaked BL Lacs and extreme blazars. The analysis is based on 336 days of data from the Astroparticle Research with Cosmics in the Abyss component of the KM3NeT research infrastructure (KM3NeT/ARCA), collected with detector configurations comprising 19-21 detection units, yielding 1544 selected track-like neutrino candidates. The sample of ultra-luminous infrared galaxies consists of 75 sources, for which conventional power-law spectra are assumed in modelling their signal emission. The blazar samples comprise 232 high-synchrotron-peaked BL Lac objects and 88 extreme synchrotron-peaked emitters, respectively. Their expected neutrino fluxes are obtained from numerical tools constrained by multi-wavelength observational data. No statistically significant excess above the atmospheric background is found for the three catalogues and 90% confidence level upper limits are set.
Read the full abstract
Astrophysics of Galaxies
C3PO: A Ly{\alpha} Emitting Galaxy at $z \sim 9$
We present deep, $>10$ hr, JWST NIRSpec G140M and G395M observations of a previously known galaxy C3PO-8447 at $z=8.99781$. Our new NIRSpec data from the Carbon-3 plus Oxygen (C3PO) survey detect Ly$\alpha$ emission with an escape fraction of $1.5\pm0.3\%$ and a velocity offset of $210\pm80$ km s$^{-1}$, making C3PO-8447 the third Ly$\alpha$ emitting galaxy confirmed at $z\gtrsim 9$. The galaxy has a low stellar mass of $2.2\pm0.1\times 10^7\ M_\odot$, a young stellar age of $1.6\pm0.2$ Myr, and a low oxygen abundance of $12+\log(\mathrm{O/H}) = 7.50\pm0.08$. UV diagnostic diagrams incorporating the C IV, O III], and C III] emission lines indicate that the ionization of C3PO-8447 is dominated by the recent starburst. The low Ly$\alpha$ escape fraction of C3PO-8447 implies that C3PO-8447 resides in a small ionized bubble with radius $R \sim 0.07$ to 0.20 pMpc, which can be produced by C3PO-8447 alone. We find no evidence for either a large-scale galaxy overdensity around C3PO-8447 or a significant contribution from nearby galaxies to its ionizing photon budget. Taken together, these results suggest that C3PO-8447 likely represents an ionized bubble powered by a single low-mass, vigorously star-forming galaxy during the early stages of cosmic reionization.
Read the full abstract
Probability
Scaling limit for the pinning model in correlated Gaussian environment beyond the $L^2$-regime
In this paper, we study the scaling limit of the pinning model in correlated Gaussian environment. The tail probability of the underlying renewal process of the model has a polynomial decay with exponent $\alpha>0$. The covariance of the Gaussian environment $\{\omega_n\}_{n\in\mathbb N}$ is given by $\text{Cov}_{\mathbb P}(\omega_n,\omega_m)\sim |n-m|^{2H-2}$ with $H\in(0,1)$. Assuming $\alpha\in(0,\frac12]$, $H\in(\frac12,1)$ and $\alpha+2H>2$, we show that the partition function of the disordered pinning model, under the appropriate scaling, converges in distribution to the $L^1$-solution of the fractional stochastic heat equation driven by Gaussian noise correlated in time and localized at the origin. In particular, it is known that the solution is not $L^2$-integrable when $\alpha<\frac12$.
Read the full abstract
Statistical Mechanics
Frenkel line of Yukawa fluids within the self-consistent relaxation theory
For a simple model fluid, the Yukawa fluid, the condition for dynamic crossover, known as the Frenkel line, is defined. The condition is related to the fact that the roton minimum in the dispersion relation of longitudinal acoustic-like excitations exists only when the collective vibrational dynamics of particles dominates in the liquid. Based on the self-consistent relaxation theory for the Yukawa fluid, thermodynamic states are determined in which the roton minimum disappears. The obtained values of the state parameters for the Frenkel line are consistent with the results of studies in which the position of this line on the phase diagram of the Yukawa fluid was determined using molecular dynamics simulations. It is shown that the Frenkel line in this system can be determined directly from the structural characteristic – the static structure factor. A physical interpretation of the roton minimum frequency for the simple liquids near the Frenkel line is proposed.
Read the full abstract
Soft Condensed Matter
Patterned Substrates Unlock Self-Electrophoretic Phenomenon in Active Janus Microswimmers
Inert colloids half-coated with platinum (Pt) are a standard model of chemically powered active particles, yet the microscopic origins of their propulsion in hydrogen peroxide (H2O2) remain difficult to dissect experimentally. Whereas self-diffusiophoresis was the prevailing theory, self-electrophoresis has been more recently suggested as the main mechanism of propulsion. According to the latter mechanism, the pole-to-equator Pt-thickness gradient produced by directional metal deposition is sufficient to create anodic and cathodic regions on the metal cap and thereby generate an electric field sustained by H2O2 decomposition. Enhancing self-propulsion performance of such particles thus requires precise control over the Pt thickness distribution, which is currently not achievable with standard methods (e.g. evaporation or sputtering). Here, we propose a method to fabricate Janus active particles by assembling silica microspheres on patterned substrates containing spherical grooves whose depth and spacing set the degree of particle coating while simultaneously suppressing proximity-led defects (Pt bridges). The resulting particles exhibit a tunable platinum-thickness contrast, as verified by Focused-Ion-Beam cross-sections. In 2.5% H2O2, our results suggest that this control can significantly increase propulsion efficiency, while providing evidence indirectly supporting the hypothesis that self-electrophoresis is the dominant mechanism. These results demonstrate that our patterned-substrate route can enhance control over the catalyst deposition and enable novel Janus morphologies, allowing for more precise engineering of active colloids.
Read the full abstract
Machine Learning
On the Interaction Between Model Compression and Test-Time Adaptation
Deep neural networks deployed in the wild must be both efficient and adaptable, requiring model compression and test-time adaptation (TTA). While both are well studied in isolation, their interaction remains poorly understood. We systematically analyze how structured compression affects a model's ability to adapt under distribution shift. Using ResNet-18 and ViT-Base on CIFAR-10-C and ImageNet-C, we evaluate multiple compression methods combined with standard TTA techniques. We introduce a diagnostic framework that examines representational expressivity and adaptation subspace compatibility. Our results reveal a consistent gap: although compressed models retain high accuracy under supervised adaptation, their TTA performance degrades significantly with increasing compression. We show that this stems from reduced representational diversity and structural constraints that limit recoverability. These effects strongly depend on the compression method, highlighting the need to design compression strategies that preserve adaptability.
Read the full abstract
Machine Learning
Neural-Network Maxent: a general extension with learned nonlinearity, applied to time-series for Desert Locust distribution modelling
Species Distribution Modelling (SDM) is essential for understanding how environmental conditions shape biodiversity, particularly for destructive pests such as the Desert Locust (Schistocerca gregaria), whose breeding dynamics are tightly coupled to rapidly evolving environmental conditions. Maxent has become the dominant method for presence-only data, but its reliance on a linear combination of hand chosen feature transforms limits its ability to capture the nonlinear, temporal relationships common in ecological monitoring, where covariates such as precipitation, soil moisture, and vegetation indices evolve meaningfully over time. Standard implementations flatten time-series covariates into independent features, discarding sequential structure that carries critical signal. We introduce RNN Maxent, an extension of the Maxent framework that replaces the fixed feature dictionary with a neural network, specifically a Gated Recurrent Unit (GRU), trained end to end via backpropagation. The approach preserves Maxent's presence only statistical foundations, background normalization, and probability calibration, differing only in that the nonlinearity is learned from data rather than fixed in advance. We apply RNN Maxent to map suitable habitat for the Desert Locust using 50 day environmental time series derived from ERA5 Land, MODIS, and Sentinel 3, maintaining a 7 day gap between covariates and presence records to yield forecasting behavior. Compared against standard Maxent, RNN Maxent improves performance across metrics (ROC AUC 0.862 std 0.036 vs. 0.792; F1 0.671 std 0.056 vs. 0.590).
Read the full abstract
Computer Vision
SV-WAM: An Efficient Surround-View World-Action Model for End-to-End Autonomous Driving
World models (WMs) have demonstrated strong potential for end-to-end autonomous driving by learning predictive representations of future scene dynamics. However, generating future videos during inference introduces substantial computational overhead, leading many recent driving WMs to adopt a single front camera as input for efficient deployment. This design restricts spatial coverage in safety-critical maneuvers such as lane changes, merges, and turns. To address this limitation, we propose SV-WAM, a surround-view world-action model (WAM) that preserves full six-camera observations while maintaining efficient inference. SV-WAM leverages future-video prediction as dense training supervision for action learning within a shared generative model, rather than as an inference-time output. At the core of this design is an action-centered causal mask that prevents action tokens from attending to future-video tokens during joint action-video denoising. Consequently, the video branch can be discarded at deployment, enabling efficient action-only planning. Furthermore, we introduce a differentiable drivable-area compliance regularizer that penalizes vehicle-footprint corners approaching or crossing drivable boundaries, improving planning safety and boundary awareness. Extensive experiments on the closed-loop NAVSIMv2 benchmark and the open-loop nuScenes benchmark demonstrate that SV-WAM achieves state-of-the-art planning performance with low inference latency and competitive zero-shot transfer capability.
Read the full abstract
Category Theory
Beth companions of finitary essentially algebraic theories
For categories, being balanced (meaning that every arrow that is both epic and monic is an isomorphism) can be regarded as a strong tameness property which plays an important role, for example, in algebra and logic. We study the problem of associating a balanced companion category, called a Beth companion, to a locally finitely presentable category (equivalently, the category of models of a finitary essentially algebraic theory). We show that, if it exists, the Beth companion is unique and can be described in terms of saturated objects. Under additional assumptions, we prove that Beth companions can be computed as orthogonality classes, and admit a syntactic presentation via Gabriel–Ulmer duality. Furthermore, we establish conditions for the transfer of properties from a category to its Beth companion. A number of examples of categories that do or do not admit a Beth companion are provided. In particular, we show that the category of commutative rings does not admit a Beth companion, and employ Pierce sheaf representations to describe the Beth companion of other classes of rings.