Peer Revenue Now accepting positions

Papers

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

All of arXiv

Saturday 7 November

High Energy Physics (Phenomenology)

Search sensitivity for type-III seesaw heavy leptons in simulated $pp$ collisions at $\sqrt{s}=13.6$ TeV at the LHC

A search for the pair production of TeV-scale heavy leptons as predicted by the type-III seesaw mechanism is presented. The analysis is based on a simulated signal and Standard Model background data corresponding to an integrated luminosity of $500~\mathrm{fb}^{-1}$ of proton-proton collisions at $\sqrt{s}=13.6$ TeV. The ATLAS detector response is modeled with Delphes. Events are categorized based on the number of final state leptons and large-$R$ jets into nine mutually exclusive search channels. Multivariate analysis techniques are employed to separate the signal from the Standard Model backgrounds. Under the flavor-democratic mixing scenario ($\mathcal{B}_e=\mathcal{B}_\mu=\mathcal{B}_\tau=1/3$), the expected 95% confidence level upper exclusion limit on the heavy lepton mass is approximately $1.6$ TeV.
Read the full abstract

Space Physics

Effect of Turbulence and Inhomogeneities on the Propagation of Kinetic Scale Waves

Context. Electromagnetic ion-scale waves are well known to occur in the solar wind. Eikonal (WKB) like effects may play an important role in the interplay between the propagation and local kinetic plasma dynamics of ion cyclotron waves. Aims. This Letter aims to explore the interplay of ICW propagation with turbulence and local kinetic plasmas using ray tracing equations for ICWs in a moving medium, in which turbulent gradients are present. Methods. Parker Solar Probe observations of ICWs in the solar wind are studied using ray equations in a moving medium. Gradients in the dispersion relation responsible for WKB evolution of ICW are constrained and compared with a Parker spiral model. Results. The empirical gradients, which contain turbulent motions, greatly enhance the efficiency of WKB effects. However, the typical refraction rates are significantly slower than expected quasilinear evolution of the observed wave populations, which suggests that local kinetic physics dominates over nonlinear expansion effects. Conclusions. Turbulence can effectively increase WKB effects due to gradients caused by nonlinear interactions. In the solar wind these gradients are not sufficient to substantially change an individual ion-scale ICW over its lifetime. Incorporating kinetic effects of expansion may help advance the understanding of generation and absorption of ICW and their role in heating in the solar wind.
Read the full abstract

Machine Learning

Constrained Look-ahead Guidance for Interference-Aware Flow Editing

Diffusion and flow models enable powerful generative modeling, but controlled editing remains challenging due to the need to balance source fidelity and edit compliance along the generative trajectory. We propose a principled, inversion-free framework for flow-based image editing from a trajectory centric perspective. Rather than relying on inversion or heuristic guidance, we introduce a one-step look-ahead mechanism that evaluates the effect of candidate updates on both edit and fidelity rewards estimated via denoising, and adjusts the current update based on this evaluation. At each step, our method applies reward-driven corrections guided by the look-ahead evaluation under an explicit fidelity constraint, resulting in stable and coherent trajectory evolution. This look-ahead design mitigates gradient interference between edit and fidelity objectives by anticipating their interaction in future states, leading to improved trade-offs. Experiments demonstrate that the proposed approach achieves stronger semantic edits while better preserving source content compared to existing strong baselines.
Read the full abstract

Machine Learning

AERO: Adaptive Ensemble-Disagreement Routing for Oracle Feedback in Sample-Efficient Online RLHF

Aligning Large Language Models (LLMs) with human preferences via Reinforcement Learning from Human Feedback (RLHF) has driven substantial improvements across a wide range of tasks, but the cost of acquiring preference labels remains the dominant bottleneck in the alignment pipeline. While many recent alignment methods improve sample efficiency through active exploration over candidate responses, they typically query the oracle uniformly across contexts, even for contexts where the learned reward estimates are already reliable enough to provide synthetic labels. We propose Adaptive Ensemble-disagreement Routing for Oracle feedback (AERO), a fully online RLHF algorithm built on a Dyna-style model-based RL architecture, in which an ensemble-based Epistemic Reward Model (ERM) serves both as a reward posterior for exploration and as a learned preference-feedback model for synthetic labeling. AERO uses ensemble vote entropy, a Query-by-Committee disagreement measure, combined with an adaptive thresholding mechanism to decide which contexts require oracle feedback and which can instead receive ERM-based synthetic preference labels, concentrating expensive labels where they are most informative. Empirical results across different model families and scales show that AERO achieves highly competitive alignment performance while reducing the required oracle-query budget compared to online DPO and strong active-exploration baselines.
Read the full abstract

Analysis of PDEs

Nonlinear stability and time-decay rates of planar rarefaction waves for the full compressible Navier–Stokes–Fourier system in $\mathbb{R}^3$

We prove the global-in-time nonlinear stability and establish explicit algebraic time-decay rates of planar rarefaction waves for the three-dimensional full compressible Navier–Stokes–Fourier system of an ideal polytropic gas in $\mathbb{R}^3$. Given a smooth initial rarefaction profile of small wave strength \(\delta>0\), we first construct the corresponding global one-dimensional Navier-Stokes-Fourier solution \(U_b(t,x_1)\) and show that it converges toward the self-similar Euler rarefaction fan in \(L^\infty(\mathbb{R})\) at rate \((1+t)^{-12/25}\). For the three-dimensional Cauchy problem under small weighted $H^3(\mathbb{R}^3)$ perturbations with zero integral of mass, momentum, and total energy, we prove that there exists a unique global strong solution $U(t,x)$ converging toward $U_b(t,x_1)$ in $L^\infty(\mathbb{R}^3)$ at rate $(1+t)^{-11/25}$, with $\nabla(U-U_b)$ decaying in $H^2(\mathbb{R}^3)$ at rate $(1+t)^{-11/25}$.
Read the full abstract

All authors Renjun Duan, Feimin Huang, Rui Li, Lingda Xu

Robotics

REINS: A Self-Evolving Agent Harness for Real-Time Trajectory Planning

Agent harnesses – runtime scaffolds that wrap an LLM's reasoning loop with tool dispatch, verification, and persistent memory – have become the standard substrate for deploying LLMs as autonomous agents, as demonstrated by Claude Code and Codex in software engineering. In trajectory planning, however, this paradigm has not taken hold: existing LLM-based trajectory planning methods either keep the LLM in the decision loop, exceeding the millisecond-level control budget, or offload it to offline rules or training-time supervision – leaving the runtime planner non-evolving. Intermediate hybrids still lack the verification-and-consolidation loop that makes harnesses reliable. We ask: what runtime harness would let an LLM planner meet real-time constraints while self-evolving into a deployable one? We propose REINS, a harness built on three LLM-planner couplings: LLM outputs are (i) grounded to a learned variable-level dynamics model, (ii) verified via calibrated predicates and forward rollout, and (iii) consolidated into an indexed rule memory that serves future scenes in O(log n) without re-invoking the LLM. The harness closes a self-evolution loop: familiar scenes resolve by memory lookup; novel scenes invoke the LLM, whose verified outputs enrich the memory. Without modifying LLM weights, REINS attains 98%+ compliance, nearly 50% collision reduction, and 1.9-4.6 ms latency on simulated and real-world benchmarks at the tactical decision layer, with LLM invocation rate monotonically declining as memory matures. Code: https://github.com/ZhihongCui/REINS
Read the full abstract

All authors Zhihong Cui, Hengyu Liu, Haoran Tang, Shijun Liu, Amir Taherkordi, Tor Skeie

Rings and Algebras

Rigid Lie Algebras from Semisimple Lie Algebras and Their Symplectic Modules

We consider two extensions associated with a nonzero finite-dimensional symplectic module $V$ over a complex semisimple Lie algebra $\mathfrak{s}$, and study their rigidity. The toral extension adjoins a derivation acting as the identity on $V$ and trivially on $\mathfrak{s}$; the Heisenberg extension uses the invariant symplectic form on $V$ to define a central bracket. For the toral extension, we compute the first and second adjoint cohomology groups in terms of $\operatorname{End}_{\mathfrak{s}}(V)$, and obtain rigidity precisely when $V$ is irreducible. These results hold for arbitrary nonzero finite-dimensional modules. For the Heisenberg extension, we express the second adjoint cohomology as the kernel of a map combining the symplectic centralizer with an equivariant Jacobi map. A Killing-form argument shows that every multiplicity-free sum of irreducible symplectic modules gives an algebraically rigid Heisenberg extension. We also treat constructions with independent toral generators or Heisenberg centers. Repeated summands lead to explicit nonrigid examples, including a deformation to $\mathfrak{sl}_3$. The proofs use invariant cochains and recover several previously studied families.
Read the full abstract

Computer Vision

RoadState: A Domain-Prior-Enhanced Vision-Language System for Industrial-Grade Lane-Level Map Updating

Lane-level map updating requires temporally consistent, physically valid, and auditable road-scene semantics despite occlusions, ambiguous lane markings, temporary obstacles, and viewpoint changes. We present RoadState, a production-deployed vision-language system that transforms sequential road-scene observations into structured lane-level semantics for Baidu Maps' map-updating pipeline. Its core model, RoadState-VLM, formulates road-scene understanding as autoregressive state estimation, predicting the current structured road state from the previous observation and symbolic state together with the current observation. We introduce Geometric Chain-of-Thought (Geo-CoT), which uses a directed acyclic graph of semantic dependencies as a domain prior to improve semantic consistency and reasoning fidelity across interdependent road-state dimensions. Extensive experiments on a real-world dataset show that RoadState-VLM achieves 94.33% mean accuracy across six road-state dimensions, outperforming the strongest fine-tuned VLM baseline. Ablation studies further demonstrate the complementary effectiveness of Geo-CoT and temporal transition context. A cascaded routing pipeline directs routine cases to lightweight models while reserving complex cases for RoadState-VLM. Since its deployment in March 2025, RoadState has supported lane-level map updating across more than 360 cities and reduced overall production costs by 30%. Together with human-in-the-loop auditing, the system maintains robust performance under stress-test conditions. These results demonstrate how domain-prior-enhanced reasoning, structured state interfaces, selective routing, and human quality control can bridge the gap between VLM-based road-scene understanding and large-scale lane-level map updating.
Read the full abstract

All authors Deguo Xia, Zihan Li, Jiaqi Wang, Dong Xie, Xiyan Liu, Liuhui Zhang, Jizhou Huang, Mengmeng Yang, Diange Yang, Chen Zhu, Liwei Qian, Xin Pei, Shuanglong Li, Hanmeng Liu

High Energy Astrophysical Phenomena

Stochastic Acceleration Downstream of Astrophysical Shocks

Generation and evolution of turbulence at shock fronts are crucial aspects of the process of particle acceleration in different classes of sources. Here we aim at describing the fate of non-thermal particles and turbulence in the downstream region of an astrophysical shock, both in the case in which turbulence is solely produced upstream and advected through the shock and in the case in which additional turbulence is injected downstream. In particular we will provide an estimate of the effect of stochastic acceleration on particles accelerated at the shock and then advected behind it. The calculations will be specialized to supernova remnant shocks. The temporal evolution of the particle distribution function downstream of the shock and the power spectrum of the turbulent perturbations are both followed by solving numerically the transport equations for the two quantities, accounting for the damping of waves due to the particles that resonantly interact with them, which makes the problem non-linear. Our calculations show that, if waves are only self-generated upstream via Bell instability, they are not able to significantly modify the particle spectrum in the downstream region. Additionally, turbulent cascading in $k$-space quickly drains their energy, further limiting the stochastic acceleration of particles. If, on the other hand, there are additional sources of turbulence downstream, then these waves can contribute meaningfully to second order Fermi acceleration and alter the spectrum of particles. This may reflect in the spectrum of gamma radiation produced through $pp$ scattering downstream. Although second order Fermi acceleration is typically slower than diffusive shock acceleration, particles are expected to stay downstream of shocks for much longer times during which stochastic acceleration can modify their spectrum and affect the spectra of multi-messenger emissions.
Read the full abstract

Computer Vision

Writer-independent handedness identification from offline Bangla handwriting under controlled acquisition

Handedness is a soft-biometric cue for forensic document examination and writer profiling, yet reported accuracies from offline handwriting rarely keep writers, texts and image acquisition apart, so models may recognise the writer or capture conditions rather than the writing hand. Such controlled evidence is missing for Bangla. Using our BanglaHand dataset of 42 writers (21 left- and 21 right-handed), who each wrote six shared prompts and two own texts under controlled collection and photography, we evaluate handedness identification with all images of each test writer held out, whole-writer permutation tests and photometric controls. An ink-gradient descriptor identified the handedness of unseen writers with 77.4% balanced accuracy (95% writer-bootstrap interval 67.7-86.5%; p=0.001) and frozen ImageNet embeddings with 69.5-72.0%. Part of this signal is stroke darkness: in a post hoc analysis, right-handed writers' strokes were darker at their centres for equal stroke width, consistent with higher pen pressure and not explained by lighting drift. With ink density and brightness regressed out, the descriptor still reached 64.5% (p=0.027), whereas writing-only shape features were weaker (60.8%; p=0.090). Handedness is thus identifiable from offline Bangla handwriting under controlled acquisition, and photometric controls are essential to interpret such cues.
Read the full abstract

All authors Tanzim Hasan Prappo, Sayeed Hassan, Sayem Ali, Sukhendra Nath Dhar, M. Shahidur Rahman

Information Theory

Performance Analysis of Entanglement Swapping DV-QKD over MIMO FSO Channels

This paper studies multiple-input multiple-output (MIMO) free-space optical (FSO) quantum communication systems employing polarization-encoded discrete-variable quantum key distribution (DV-QKD), based on entanglement swapping, with the goal of extending secure communications beyond the reach of a direct link. The FSO channels are subject to atmospheric absorption, diffraction, beam wander, pointing errors, and turbulence-induced fading, while background noise and detector imperfections degrade the performance. An eavesdropper controlling the channels and the relay is also considered. To analyze the performance of the considered system, both hops MIMO channels are decomposed into parallel spatial eigenchannels via singular value decomposition (SVD), photon loss is modeled using Kraus operators, and a threshold-detector model is developed for the relay. Based on these models, closed-form per-eigenchannel expressions for the relay gain are derived, which are then used to obtain expressions for the quantum bit error rate (QBER) and the secret key rate (SKR). Numerical results demonstrate that entanglement swapping maintains a low QBER over longer distances than direct transmission, extending the communication range beyond the noise-limited cutoff of the direct link, at the cost of a lower SKR at short distances. The crossover distance beyond which entanglement swapping prevails is determined mainly by the background noise and the detector technology, and it is demonstrated to increase with MIMO configurations. Moreover, the reach gain is shown to be maximized for a midpoint relay. Finally, since each swapping event requires a single-pair emission from both sources, a near-deterministic entangled-photon source is required, which spontaneous parametric down-conversion sources do not provide at practical key rates, whereas MIMO substantially relaxes this requirement.
Read the full abstract

Operator Algebras

Solid factors associated to the linear action of $SL_2(\mathbb{Z})$ on $\mathbb{Z}^n$

Let $n\geq 2$ be an integer. Set $\Gamma_n=SL_2(\mathbb{Z})$ (respectively $PSL_2(\mathbb{Z})$) for $n$ even (respectively for $n$ odd). Then $\Gamma_n$ acts on $\mathbb{Z}^n$ through the restriction of the linear action of $SL_2(\mathbb{R})$ on $\mathbb{R}^n$. We prove that that the von Neumann algebra $L(\mathbb{Z}^n\rtimes \Gamma_n)$ is a solid factor. This generalizes the case $n=2$ proved by N. Ozawa.
Read the full abstract

High Energy Physics (Theory)

Configurational Information and Hadron Spectroscopy in Holographic QCD

Based on the concepts of entropy and complexity, Configurational Information Measures (CIMs) are tools capable of extracting information regarding physical systems from their spatially localized energy density, determining properties such as stability limits and phase transitions. In this thesis, CIMs are employed to investigate hadron spectroscopy. By utilizing phenomenological AdS/QCD models in conjunction with CIMs, cataloged meson and baryon resonances are interpolated into structures analogous to Regge trajectories, from which analytical expressions are derived to estimate the masses of higher-order excited states. These estimates are compared with experimental data to search for possible correspondences that may elucidate the properties of particles that have been experimentally detected but remain unidentified. Using this CIM-based AdS/QCD hybrid method, this work analyzes light mesons within a dynamical model, strange vector kaons in a framework that accounts for constituent quark masses, and light-flavor baryons at finite temperature.
Read the full abstract

Computation and Language

Evaluating Physically Grounded Procedural Reasoning in VLMs through Real-World Execution: Crochet as a Testbed

Vision-Language Models (VLMs) can generate procedures that look convincing on paper, but do they work in practice? In this paper, we propose crochet as an ideal testbed for physically grounded procedural reasoning in VLMs. Crochet cannot be reliably simulated or automated: the only way to know what a pattern produces is to make it by hand, yet verification is tangible and cheap, requiring nothing but a hook and a ball of yarn. We introduce FoundationChain, a deceptively simple test of procedural reasoning: given an image of a basic crochet object, can VLMs infer how it was made and generate a pattern that produces the same result? We crochet 40 patterns generated by four VLMs by hand. Even at this scale, physical execution not only reveals failures that automated proxies cannot detect, but also enables a more precise diagnosis of their causes and consequences: while models write fluent, convincing patterns, only four are free of errors, and our analysis exposes deficiencies in visual, geometric and procedural reasoning.
Read the full abstract

Artificial Intelligence

Pulse waveform morphology encodes haemodynamic signatures of impending stroke

Stroke is a disorder of the circulation, yet the haemodynamic state that immediately precedes it has rarely been observed, because continuous physiological recordings before a stroke are seldom available. Photoplethysmography (PPG), recorded by every pulse oximeter and by hundreds of millions of wearable devices, captures the peripheral pulse beat by beat, but its morphology is largely discarded. Here we ask whether pulse waveform morphology encodes haemodynamic signatures of impending stroke. Using in-hospital stroke as a natural experiment that provides continuous PPG before physician-adjudicated stroke recognition, we studied intensive-care (MIMIC-III, $n=176$) and emergency-department (MC-MED, $n=154$) cohorts and described each beat with interpretable descriptors of pulse timing, amplitude, contour and variability. At a nominal 360-min horizon, morphology separated pre-recognition from stable windows with $F_1$ scores of 0.941 and 0.989 (sensitivity 0.998 and 0.980), with models transferred unchanged to the second cohort, and outperformed heart rate alone on identical windows. The signature was dominated by systolic crest time and its change from each patient's own baseline, feature attribution was concordant across cohorts (Spearman $\rho=0.865$), and morphology recorded in the first two hours of care stratified the subsequent clinical course ($F_1$ 0.714 and 0.691). In high-risk patients without stroke, 29% and 14% of windows were positive, but five-window persistence rules limited alerts to 17% and 2% of patients. The results demonstrate that the routinely recorded pulse thus carries haemodynamic imprint of the period preceding stroke, positioning pulse morphology as a candidate continuous cerebrovascular biomarker.
Read the full abstract

All authors Cheng Ding, Jiaming Liu, Jian Wu, Hongxia Xu, Daoqiang Zhang

Machine Learning (Statistics)

Matchability-Aware Conformal Prediction for Open-Ended Language Model Generation

Large language models (LLMs) often do not produce calibrated uncertainty estimates for their answers. We propose CAELUM, a framework that builds semantic prediction sets while separating aleatoric from epistemic uncertainty. CAELUM samples answers using an ensemble of clarification prompts, clusters them by semantic equivalence, and estimates the aleatoric and epistemic components via kernel-based von Neumann entropy before calibrating nonconformity scores with split conformal prediction. We use a weight $w_\lambda(x)=1+U_{\mathrm{ale}}(x)+\lambda\,U_{\mathrm{epi}}(x)$ to control for the contribution of the epistemic uncertainty of each instance and fit a coefficient $\lambda^\star$ to minimize prediction set size. When the objective is nonflat and selection is stable across splits, $\lambda^\star$ could be a diagnostic, where large values may indicate epistemic dominance (retrieve, abstain) and $\lambda^\star\to 0$ aleatoric dominance (clarify). For individual queries, a log ratio $R(x)$ compares the weighted epistemic component with the aleatoric component. Theoretically, we demonstrate that the coverage is capped by matchability, the fraction of items whose reference answer appears in any sampled cluster. Empirically, this matchability can be predicted from the original question features using a lightweight classifier. We use those predictions to define a stratum and derive a conformal coverage bound within it. Across 30 LLM and dataset combinations (4-32B), mean coverage is 0.74 in the learned stratum versus 0.65 overall, averaged over the 78 evaluations using the learned classifier. Project page: https://unco3892.github.io/caelum/
Read the full abstract

Artificial Intelligence

The Lagrangian Brain: A Constraint Mechanics of Mind and Machine

A brain is in one state at a time, yet most of what decides its next state is absent from it: the patterns its world never produces, the statistics of a lifetime, a goal that lies ahead. In mechanics a wire acts on a bead only through a force that appears when the bead presses on it, and Lagrange computed such forces with one multiplier for each constraint. We propose that a brain follows the same mechanics. Everything a circuit knows, is told or wants is a constraint on its single state and acts through its own multiplier, which accumulates the violation and pushes the state back. Neurons and synapses carry the multipliers, each with its own time constant: inhibitory interneurons carry what never happens, adaptation the cost of staying in one state, synapses the statistics of experience, and feedback to apical dendrites the future. From this hypothesis follow Dale's law, the reciprocal wiring and balance of excitation and inhibition, the paradoxical responses of cortex and the frequencies of brain rhythms. Belief becomes the time one trajectory spends in each state, learning becomes sleep proposing laws and waking refuting the false ones, and planning becomes a wave from the goal meeting a wave from the present. Connectionism and symbolism are the limits of frozen and of instantaneous multipliers, and between them the time a circuit runs becomes the depth of its computation. A machine built on these principles, trained without gradients on 1000 solved puzzles, solves all 422,786 Sudoku-Extreme test puzzles with one trajectory each. It rediscovers the rules of Sudoku by sleeping and waking, plans shortest paths in mazes of 64 times the training area, and tracks permutations exactly on strings 25 times longer than in training, where Transformers fail. A mind, on this view, is life that learns its own constraints.
Read the full abstract

Data Structures and Algorithms

Tunnels: Compressing the Solution Subgraph of Regular Path Queries

Regular path queries (RPQs) are increasingly expected to return the paths that match them, not only their endpoints. Since the number of matching paths can be exponential or infinite, a standard way to deliver them (PMR) is the subgraph $\Hs$ of the product between the graph and the query automaton that contains exactly the nodes lying on some solution path. We show that $\Hs$ can be compressed much further when the graph is available to the decoder: most of $\Hs$ can be regenerated by re-running small parts of the traversal, and only the places where the traversal would waste significant effort on dead branches need to be stored. We formalize this as tunnels: regions of $\Hs$ that are re-expanded from a stored entry point with a bounded amount of extra work, either an absolute budget or one proportional to the output produced. We handle the three classes of RPQs: constant-to-variable, constant-to-constant, and variable-to-variable. In the latter, we study the combination of tunnels with cutting the query at a selective articulation point of the automaton and encoding the answer as a union of Cartesian products of leftward and rightward traversals. Our experimental results show improvements of 2 to 3 orders of magnitude over the baseline space of PMR, for an acceptable increase in traversal time.
Read the full abstract

Combinatorics

A 43-vertex 4-chromatic unit-distance graph in the plane with girth 5

We exhibit a set of 43 points in the Euclidean plane whose unit-distance graph has 83 edges, girth 5 and chromatic number 4. The smallest previously known 4-chromatic unit-distance graph of girth 5, due to Hochberg and O'Donnell (1996), has 45 vertices. The new graph is 4-critical, has trivial automorphism group, and has exactly 2n-3 edges; it is a Laman graph and its unit-distance realisation is infinitesimally rigid. The coordinates are algebraic numbers, specified as the unique real solution of an explicit system of 82 quadratic equations inside an explicit rational box. Every claim is certified by exact computation: existence and uniqueness of the realisation by a rational contraction-mapping certificate (independently re-checked with interval Krawczyk arithmetic), and non-3-colourability both by exhaustive search and by a DRAT proof verified with drat-trim.
Read the full abstract

Computer Vision

Tiny Evidence Region Localization with Selective Contrastive On-Policy Self-Distillation

Multimodal large language models have made progress in general visual tasks. However, limitations remain in fine-grained visual tasks requiring evidence from minute image regions. Existing methods primarily train models using evidence-related regions while overlooking the influence of extensive background regions, resulting in a limited ability to resist interference from irrelevant regions. We propose selective positive-negative on-policy self-distillation (SPN-OPSD), which considers both evidence and background regions. Supervision from both views improves the model's ability to resist interference from irrelevant regions. Furthermore, selective token-level distillation provides more informative supervision, improving the model's ability to attend to evidence regions in Fine-Grained Visual Tasks. On Qwen3.5-2B, SPN-OPSD improves accuracy on V* Bench, ZoomBench, and MME-RealWorld-Lite from 76.44%, 45.92%, and 57.43% to 84.82%, 54.56%, and 59.56%, respectively. Similar gains are also observed on Qwen3.5-0.8B.
Read the full abstract

Machine Learning

Spectrogram-Based CNN Autoencoder for One-Class Anomaly Detection in Machining Vibration Signals

This work presents a one-class anomaly detection framework for vibration-based condition monitoring in machining processes using a convolutional neural network autoencoder. Three-axis vibration signals acquired during industrial broaching are transformed into time–frequency spectrograms and combined into a compact three-channel representation, enabling convolutional feature learning on multivariate time-series data. The autoencoder is trained exclusively on normal process data and detects anomalous strokes based on reconstruction errors. Labeled validation data are used for threshold calibration and an independent test set is reserved for final performance evaluation. To improve sensitivity to short-lived process instabilities such as chatter, a localized residual-based metric termed Segmented Vertical Error (SVE) is introduced. Experimental results on approximately 10{,}000 broaching strokes show a clear separation between normal and anomalous process conditions under strong class imbalance. The processing pipeline is additionally evaluated on an external public turning dataset as a proof of concept, indicating preliminary applicability beyond the original broaching setup. The results suggest that spectrogram-based autoencoder models can provide a practical and scalable basis for one-class anomaly detection in industrial vibration monitoring.
Read the full abstract

Optics

Improving the convergence of coupled quasinormal-mode theory for optical resonators

Quasinormal modes (QNMs) provide a rigorous framework for describing resonances in open optical systems, including radiative and material losses. Recent coupled-QNM theories use the modes of individual resonators to predict the resonances and optical responses of coupled systems, without relying on phenomenological coupling parameters. However, their accuracy can deteriorate when the resonators confine light weakly and radiate strongly, often requiring many modes to obtain reliable predictions. Here, we introduce a reformulation of coupled-QNM theory that substantially improves its accuracy when only a small number of modes of the individual resonators are retained. The approach builds on the same underlying QNM framework but treats the fields mediating the interaction between resonators differently, leading to a new expression for the coupling coefficients. Numerical comparisons show that this extended-domain formulation is particularly advantageous for resonators with strong ra-diative losses, enabling more efficient modeling of a broader range of coupled photonic sys-tems.
Read the full abstract

General Economics

Have wind and solar energy increased electricity price volatility? And what is electricity price volatility, anyway?

It is widely held that wind and solar power increase the volatility of electricity prices. This paper argues that the literature conflates two distinct types of price variation: across delivery time and across trading time. Determined simultaneously in the same day-ahead auction, prices for power to be delivered in distinct quarter-hours differ. This price dispersion, which I call "variability", is rooted in non-storability. After that auction, trading continues in intraday markets until imbalance settlement. The price for the same delivery period changes as new information arises, e.g., weather forecasts. Such price "volatility" is ultimately caused by uncertainty. This paper develops a simple algebraic model of day-ahead and imbalance markets that provides a formal definition of variability and volatility, allows for a decomposition and derives the fundamental drivers, uncertainty and non-storability. It then provides empirical data on price variability and volatility for the past 25 years in Germany, during which the share of wind and solar energy increased from zero to nearly 50 % of generation. Volatility is measured with imbalance prices, intraday price indices and intraday transaction data. Variability followed a U-shaped trajectory, falling from the 2000s to the 2010s but increasing since then. Volatility, in contrast, declined over time. Across 24 European bidding zones, variability has risen since the late 2010s in almost all zones and increases with the share of solar power.
Read the full abstract

Machine Learning

Breaking Precision Walls in First-, Second-, and Zeroth-Order Optimizers: One Constant, Set in Advance

Under power-law signal decay across directions, an optimizer can fit strong directions and miss weak ones while average loss hides it. A precision wall measures the steepest decay where the weakest direction recovers within 5% in a fixed budget. On low-rank matrix sensing and quadratics one constant sets the wall: once it dominates the denominator, Adam or grafted Shampoo is momentum descent at step $\eta/c$, setting the wall; in Shampoo, matching grafting's constant to damping restores the wall after three hypotheses failed. In zeroth-order Adam, $\varepsilon$ acts against $\varepsilon^\ast$ from a mean-square spectral model; below $\varepsilon^\ast$ error hits a noise floor. Set from true curvature or spectrum on new shapes, the rule raised the wall above the same-step default by a forecast amount: on two sensing shapes by $1.96, 1.15$ (Adam) and $1.59, 1.74$ (grafted Shampoo), and on two Hadamard-rotated quadratics by $1.12, 0.73$ (zeroth-order Adam, longer budget), rising on all seeds; fixed-step momentum at that step reaches those walls. A cosine-scheduled default (small constant giving weak modes large steps) is higher on one sensing shape, matched on new seeds of the other, and lower with the rule's constant. A default step retuned per budget catches up on rotated quadratics and population logistic regression, fails at 1 query/step on two diagonal quadratics (catching up at 16), and exceeds the rule on digits logistic regression (full-batch and minibatches, where gradient noise sets the $\eta = 0.01, 0.1$ default walls). On logistic regression the rule did not raise the first-order wall: same-step default Adam was higher (its wall on 4 new shapes at the forecast stability limit), while in zeroth order on the population loss setting $\varepsilon$ raised the wall by $0.98, 0.88$.
Read the full abstract

Computer Vision

LIFT: Local Metric Alignment and Foundation-guided Propagation for Few-shot Depth Completion

Few-shot depth completion aims to recover dense metric depth from sparse measurements using only a small number of target-domain annotations, offering a favorable balance between annotation cost and model adaptation. We propose LIFT, a framework that harnesses both depth and feature priors from a frozen pretrained monocular depth estimation (MDE) foundation model for few-shot depth completion. Specifically, LIFT first estimates local scale and shift parameters on a coarse grid and interpolates them into smooth, dense parameter fields, providing a spatially adaptive metric depth initialization. It then progressively refines this initialization through multi-scale depth propagation guided by foundation features. Finally, a lightweight residual decoder combines multi-level foundation features with propagation correction to refine the metric depth. Extensive experiments on NYUv2 and KITTI-DC demonstrate that our proposed method achieves state-of-the-art few-shot performance. The source code will soon be released.
Read the full abstract

Computer Vision

3D Segmentation in the Clinic: A Grand Challenge

This paper describes the set-up of a segmentation competition for automatic and semi-automatic extraction of the liver from computed tomography scans and the caudate nucleus from brain MRI data. This competition was held in the form of a workshop at the 2007 Medical Image Computing and Computer Assisted Intervention conference. The rationale for organizing the competition is discussed, the training and test data sets for both segmentation tasks are described and the scoring system used to evaluate the segmentation is presented.
Read the full abstract

General Mathematics

A Prey-Predator Framework for Tumor-Immune Dynamics: Circuit-Based Modeling with Hilfer-Katugampola Fractional Derivatives

This study focuses on modeling tumor–immune interactions through a fractional-order prey–predator framework, where tumor cells act as prey and immune effector cells as predators. To capture the hereditary and memory effects inherent in biological systems, we employ the Hilfer–Katugampola fractional derivative, which flexibly interpolates between Riemann–Liouville and Caputo formulations. In addition, a circuit-based analogy is introduced to map tumor proliferation, immune regulation, and therapeutic interventions onto resistive, capacitive, and inductive elements, providing intuitive interpretation of the dynamics. Analytical investigations, including equilibrium analysis and stability criteria, are conducted across varying fractional orders. Numerical simulations reveal that lower fractional orders lead to delayed oscillations, prolonged tumor persistence, and increased sensitivity to treatment strategies compared with classical integer-order models. The integration of fractional calculus with circuit analogies offers novel insights into tumor–immune dynamics and supports the design of memory-aware therapeutic interventions.
Read the full abstract

Number Theory

A quantitative two-interval criterion for the Riemann hypothesis

We aim to study whether the Weil form can detect a zero of the Riemann zeta function away from the critical line using two functions with increasingly narrow, widely separated supports. We prove that the Riemann hypothesis is equivalent to the Cauchy-Schwarz inequality on every such pair at all sufficiently large separations, provided the component widths decrease subexponentially. A zero to the right of the critical line forces the normalized cross term to exceed an exponential at arbitrarily large separations, with any exponent smaller than the horizontal displacement of that zero. In particular, reciprocal-width intervals suffice. Our proof combines a short-support positivity estimate, a smooth partition of fixed test functions and a pole obstruction for a one-sided Laplace transform. We also express the cross term through prime powers in a moving multiplicative interval. The resulting criterion specifies a uniform arithmetic inequality whose proof would establish the Riemann hypothesis.
Read the full abstract

Biomolecules

Symmetry-Induced Conservation Laws, Three-Channel Attribution, and Solid-Phase Recycling in Prebiotic Enantiomeric Excess Amplification

Enantiomeric excess is the primary observable used to quantify chiral amplification across prebiotic chemistry, asymmetric autocatalysis, and multiphase crystallization. Because enantiomeric excess divides net chiral excess by total chiral mass, identical increases can arise either from parity-breaking synthesis of chiral centers or from symmetric phase and oligomer partitioning that depletes soluble mass. We establish a pathwise conservation law, a model-free three-channel attribution identity, and a continuous path-covariant Aumann–Shapley attribution for mirror-symmetric reaction networks. In any closed network lacking autocatalytic feedback or epimerization, net chiral excess remains invariant along every stochastic trajectory, locking liquid-phase amplification inversely to surviving chiral mass while Shear–Kramers–Lorentz relative-entropy balance bounds dissipation. We re-analyze seven experimental benchmark systems and show that standalone phase removal loses $91.7\%$–$98.9\%$ of chiral solute into racemic sediment in closed batches and washes out under open racemic inflow. We show that coupling eutectic partitioning with racemization recycling achieves $96.5\%$ homochiral crystal recovery, whereas evaporative autocatalysis crosses the open-flow bifurcation threshold with $81.9\%$ homochiral yield. Finally, we extend our non-crossing parity invariant across a fifty-nine-point multicomponent peptide ligation and seeded co-precipitation benchmark, evaluate a one-hundred-twelve-row calcite adsorption dataset, and resolve mineral-occlusion non-identifiability using three-block antipodal crystal pairs.
Read the full abstract

Quantum Physics

Dispersion interaction between two bodies of arbitrary shape

Based on the vacuum tensor Green's functions of electrodynamics, the Lippmann-Schwinger-type equations were derived for the correlations of electromagnetic fields. These correlations were applied within the Maxwell stress tensor and the Lorentz force formula to determine Casimir dispersion forces for two arbitrary bodies. The analysis is performed by the excitation of bodies by electric and magnetic dipoles.
Read the full abstract