Systems and Control
Infrastructure-Aware Runtime Protection of Co-Deployed AI Services at the Industrial Edge
Nitya Sathyavageeswaran, Dibyajyoti Nath, Ziran Min et al. (+1 more)
In modern production lines, a single edge device simultaneously runs multiple AI services like a torque-control model, a
vision inspector, a predictive maintenance routine. All these services run on the same CPU and same shared network segment. When the shared network segment saturates, the services do not crash. The services just produce outputs that are inaccurate or stale, and no longer reflect the plant's conditions. Our paper targets this class of silent, partial failure scenarios in Industrial Edge Systems. We treat each co-deployed service as an autonomous
agent. We introduce the Runtime Criticality Score (RCS), which each service
agent uses to express at any moment, its intrinsic utility and survival urgency under resource constraints. The score is computed from six normalized dimensions covering physical consequence, inference quality, input freshness, recent availability stress, actuator coupling, and available protective actions on the host. Its dimension weights are adjusted online through a lightweight
Bayesian update driven by the type of disturbance currently observed, so the same service is ranked differently under a network delay than under thermal throttling.
Agents publish their RCS to peers over an open, peer-to-peer discovery channel, and use it to negotiate three things without a central controller: (i) local threshold enforcement on their own host, (ii) migration to a willing peer when the current host can no longer protect them, (iii) reassignment of the coordination role when the incumbent coordinator is itself impaired. Beyond the immediate benefit of catching silent degradation, the framework contributes a concrete instance of decentralized, physically grounded
agent coordination.
Agents belonging to different owners can advertise capabilities, bid for protection, and reorganize across an open network.
Read the full abstract ▾
All authors Nitya Sathyavageeswaran, Dibyajyoti Nath, Ziran Min, Pavan Athreya Narasimha Murthy
Probability
Critical oriented percolation enhanced along random supercritical horizontal lines
Lucas Leproult and Maria Eulália Vares
We consider the oriented site percolation model in a random environment introduced by Kesten, Sidoravicius and Vares (2022), in which the horizontal lines of the
lattice are independently declared bad or good, sites being open with a probability that depends only on the type of their line. We study the boundary case where the bad lines are critical and the good ones supercritical, so that the model is critical oriented site percolation enhanced along a random set of good lines. We prove that percolation occurs for any positive density of the good lines. An arbitrarily sparse set of arbitrarily weakly enhanced lines suffices, the density of good lines and the strength of the enhancement being decoupled. The proof adapts the renormalization scheme of Archer, Hartarsky, Kolesnik, Olesker-Taylor, Schapira and Valesin (2026). We show that the edge of the process travels at a positive speed, which yields crossing estimates for tilted boxes; during runs of bad lines, we rely on the box-crossing property of Duminil-Copin, Tassion and Teixeira (2018); and the renormalized model dominates the oriented percolation with
geometric defects of Hilário, Sá, Sanchis and Teixeira (2026).
Read the full abstract ▾
Cryptography and Security
BALLAST: Physics-Aware Moving Target Defense for Industrial Control Systems Using a Digital Twin Safety Oracle
Oscar Mogollon-Gutierrez, Mohammadhossein Homaei, Alberto Lopez-Trigo et al. (+2 more)
Industrial Control Systems run on static network configurations, which leaves an attacker unlimited time to map the plant. Moving Target Defense answers this by shuffling addresses, but almost every published scheme never checks whether the reconfiguration will itself disturb the physical process it protects, so the defense can destabilise a control loop or trip an actuator. We present BALLAST, which places a learned Digital Twin between the security policy and the plant as a safety oracle. A physics-informed LSTM surrogate forecasts the physical effect of a candidate network change, a reinforcement learning
agent trades security
entropy against process stability, and a connection-draining handover lets in-flight Modbus transactions complete before the old mapping is withdrawn. An explicit shield rejects any action whose predicted trajectory would leave the safe operating envelope. On the SWaT water treatment testbed, Ballast cut the success rate of an adversary's reconnaissance sweep by 85 percent and recorded no safety violation across 50 runs, a 95 percent upper bound of 5.8 percent on the violation rate, against five violations and a tracking error 19 times larger for RL-MTD, the closest published analogue to our design. Measured against the strongest baseline on the physical metrics, a threshold-triggered scheme that itself records one violation rather than RL-MTD's five, Ballast still improves the tracking error by a factor of 2.5 and removes that remaining violation. The gain costs a small and deliberate concession in raw deterrence. Twin inference meets the 50 ms budget on a workstation core and an embedded GPU, and is projected to meet it on both edge CPUs when the three envelope stages are evaluated in a single batched call; the complete decision-to-enforcement path was measured only on the workstation.
Read the full abstract ▾
All authors Oscar Mogollon-Gutierrez, Mohammadhossein Homaei, Alberto Lopez-Trigo, Andres Caro, Mar Avila
Analysis of PDEs
Global well-posedness for a damped two dimensional fluid–beam system
Weihua Wang and Lu Yang
We study the normalized surface system of an irrotational incompressible fluid of infinite depth beneath an elastic Euler–Bernoulli beam with structural damping. For every fixed damping coefficient $\gamma>0$, we establish global well-posedness for arbitrary initial data in $H^2(\R)\times\bigl(\dot H^{1/2}(\R)\cap\dot H^1(\R)\bigr)$. The main novelty is that the admissible data include finite energy initial potentials for which neither $\psi_0$ nor $u_0=\psi_0+G(\eta_0)\psi_0$ has an $L^2$ representative. We also prove the exact energy dissipation identity at the regularity of the constructed solutions. The main analytical difficulties are the possible failure of $L^2$ integrability at low frequencies and the justification of energy differentiation at the solution regularity. To handle the low-frequency difficulty, we fix the low-frequency part of the initial potential and evolving an $L^2$ remainder. Homogeneous Dirichlet–Neumann estimates and a high-frequency
Fourier symmetrizer yield local well-posedness with $H^{-1}$ forcing. Continuous Fréchet differentiability of the energy in these affine coordinates, together with a chain rule on $H^1\subset L^2\subset H^{-1}$, justifies the dissipation identity. The Energy identity and coercivity provide uniform bounds that yields global continuation tharough the local theory. Iterating the local uniqueness and stability estimates gives global uniqueness within the stated affine solution class and Lipschitz dependence on bounded sets over every finite time interval.
Read the full abstract ▾
Computation and Language
Prompt or Policy? Optimizing Physician Slate Presentations with LLM User Simulators
Andrei Niculae, Adrian Cosma and Emilian Radoi
Choosing the right physician is a key challenge for patients seeking medical care. Physician recommendation systems identify relevant doctors, but patients still find it difficult to compare their profiles to make a decision. We formulate slate presentation optimization: given a patient's medical case and a fixed, ranked list of physicians, we generate short, case-specific summaries that make doctors easier to compare, while remaining grounded in the information in their profiles. To optimize physician presentations without relying on real patient interactions, we develop an offline framework that uses persona-conditioned LLMs to simulate patient choices. We generate patient personas from attributes extracted from each medical case and evaluate presentations for factual accuracy, relevance, naturalness and simulated patient choice. We also use a second LLM judge, unseen during optimization, to assess simulator-specific overfitting. We compare automatic
prompt-level optimization with GEPA, parameter-level optimization with GRPO, and their combination across two model sizes. GRPO substantially outperforms GEPA in satisfying the full set of constraints, and its gains in simulated conversion persist under the unseen judge. GEPA alone brings smaller improvements, but using its optimized
prompt to initialize GRPO speeds up training and achieves the highest simulated conversion. These results show that
prompt-level and parameter-level optimization play complementary roles in improving how physician recommendations are presented.
Read the full abstract ▾
Computation and Language
SimInter: Evidence-Grounded User Simulation with Interview-Based Profile Enrichment for Conversational Recommender Systems
Erfan Shahabi, Amirhossein Roshandel, Mohammad Sadegh Sadeghi et al. (+1 more)
Conversational Recommender Systems (CRSs) require reliable user simulators for cost-effective evaluation. LLM-based simulators have made realistic simulation possible but typically condition on profiles derived from explicit ratings, demographics, and static sources capturing surface preferences while omitting implicit behavioral characteristics such as decision-making style, aspect-level preferences, and domain expertise. We propose SimInter, a pipeline that elicits these attributes via synthetic interviews between two LLM
agents, grounded in the user's statistical profile and historical conversational evidence. Before profile enrichment, an audit
agent explicitly flags potential fabrication and stereotyping at the turn level. We instantiate and evaluate SimInter on movie recommendation. We simulate 300 ReDial users and, on a matched subset of 100 users, compare SimInter with a statistics-only baseline: SimInter significantly increases the diversity of expressed preference aspects (filtered aspect
entropy: 1.619 vs. 0.859; paired d = 0.96, Holm-adjusted p < 0.001). This gain persists when the simulator converses with two trained recommenders, KBRD and UniCRS, under two independent evaluation judges (+0.28 to +0.56 nats; all p < 0.001). Human evaluation finds 94.3% of enriched-attribute claims supported or partially supported, while human judgments agree with 95.2% of the auditor's binary filtering decisions on a stratified sample.
Read the full abstract ▾
All authors Erfan Shahabi, Amirhossein Roshandel, Mohammad Sadegh Sadeghi, Azadeh Shakery
Robotics
I2O2I-Nav: A Benchmark with a Modular Agent for Long-Horizon Indoor-to-Outdoor-to-Indoor Navigation
Zhengjie Shu, Wanyu Li, Di Xiao et al. (+11 more)
Service and delivery
robots often need to complete long-horizon navigation missions spanning indoor and outdoor environments. Existing embodied navigation
benchmarks primarily evaluate isolated stages or partial cross-domain journeys, providing limited support for assessing how heterogeneous navigation capabilities compose under continuous execution. We introduce I2O2I-Nav, a physics-enabled
benchmark of 20 real-world composite scenes spanning over 115,000 square meters across multiple buildings and floors, with aligned 3DGS rendering and collision geometry. Its three-level evaluation progresses from independent primitives to two-stage compositions and complete four-stage I2O2I missions. Each stage starts from its predecessor's actual endpoint, while stage-conditional metrics track completion and localize failures along the mission. We further propose the MODULAR-
Agent that coordinates frozen navigation policies through completion verification, stage handoffs, and bounded recovery. Evaluation across 1,374 episodes highlights the difficulty of continuous composition. The MODULAR-
Agent achieves 40.5-76% success on individual primitives, but full I2O2I success remains low. Compared with a fixed policy chain using the same underlying policies, it more than doubles success on both Indoor-to-Outdoor (I2O) and Outdoor-to-Indoor (O2I), while improving full-mission success from 1.15% to 4.02%. Stage-wise analysis shows that the MODULAR-
Agent enables more missions to reach the final object-navigation stage, where conditional success remains low, revealing challenges in both long-horizon stage execution and completion decisions.
Read the full abstract ▾
All authors Zhengjie Shu, Wanyu Li, Di Xiao, Lingxin Jiang, Lai Jiang, Kaidi Hu, Xincheng Tang, Wenkang Hu, Haoran Wang, Jinchuan Zhang, Chang Chen, Wei Li, Jia Pan, Ruigang Yang
Machine Learning
Bayesian Optimization of CVaR with Unknown Hyperparameters
Valentin Margraf, Jonas Hanselle, Viktor Bengs et al. (+3 more)
We study
Bayesian optimization of the Conditional Value-at-Risk (CVaR) of an expensive black-box function that depends on a decision and an environmental variable. Existing no-regret
algorithms assume that the correct Gaussian process (GP) kernel and its hyperparameters are known. In practice, however, they must be estimated from limited data, often yielding a misspecified surrogate. This is especially damaging for CVaR because misspecification can distort the estimated tail-risk and thus lead to catastrophic decisions. We propose HE-CV-UCB, the first no-regret
algorithm for CVaR
Bayesian optimization with unknown GP hyperparameters. Given a finite candidate set containing a well-specified model, HE-CV-UCB adaptively eliminates those inconsistent with the observed data, actively queries where the survivors disagree, and returns a design that is robust across the surviving candidates. We prove that it attains cumulative regret matching the oracle rate of CV-UCB up to a factor depending only on the candidate set's size. Across synthetic
benchmarks and three real-world applications, HE-CV-UCB achieves competitive performance and frequently outperforms existing CVaR
Bayesian optimization methods.
Read the full abstract ▾
All authors Valentin Margraf, Jonas Hanselle, Viktor Bengs, Serafima Lebedeva, Eyke Hüllermeier, Marcel Wever
Computer Vision
AffectWorld: Benchmarking Emotion-Driven Future World Prediction
Changhao Song, Yacong Chen, Hui Gao et al. (+2 more)
The world encompasses both physical environments and human actors. Predicting its
evolution requires understanding both physical laws and how emotions shape human actions. However, existing world model
benchmarks primarily assess physical plausibility and perceptual quality, leaving emotion-driven world
evolution largely untested.We introduce AffectWorld, a
benchmark that evaluates the chain Emotion-Action-World State through three tasks: World-State Consistency, Emotion-Action Consistency, and Causal Consistency. It comprises 1,498 everyday
video samples annotated with plausible emotions, action tendencies, and atomic world-state constraints. A counterfactual subset of 233 samples tests whether changing an
agent's emotion under fixed scene and event conditions yields corresponding changes in actions and consequences. We define eight metrics across the three tasks and an additional counterfactual metric, combining automatic visual assessment with human-calibrated VLM/LLM judgments. Evaluations of five SoTA world models (e.g., Cosmos3) required world-state change scores of 0.100-0.280 and counterfactual emotion sensitivity scores of 0.027-0.070 on a [0, 1] scale. These results reveal substantial gaps in generating the consequences of emotion-driven actions and adapting future predictions to emotional interventions. AffectWorld provides a testbed for advancing world models that capture how emotions shape future physical and social states.
Read the full abstract ▾
All authors Changhao Song, Yacong Chen, Hui Gao, Yazhou Zhang, Peng Zhang
Artificial Intelligence
Efficient Safety Guardrails through Latent Dual-Hypothesis Reasoning with On-Demand Trace Reconstruction
Md Asiful Islam, Fahmida Alam and Mihai Surdeanu
Guardrail approaches that rely on
reasoning-based large
language models (LLM) improve safety decisions, but come with substantial inference overhead due to the explicit
reasoning. We introduce LR-Guard (Latent
Reasoning Guardrail), an efficient and explainable guardrail that jointly learns structured latent
reasoning and on-demand
reasoning reconstruction. Our method distills a frozen teacher's analysis of safe and unsafe
prompt interpretations and their comparison into compact latent representations. Joint training optimizes these representations for both safety prediction and reconstruction of the teacher's explicit
reasoning traces, enabling human-readable explanations without requiring intermediate text generation for every decision. Experiments show that LR-Guard outperforms its explicit-
reasoning teacher in most in-domain and out-of-domain settings on safety classification and explanation extraction, while achieving a 19.4x inference speedup at batch size one. Further, reconstructed
reasoning traces from the latent information exhibit high semantic similarity to the teacher's explicit
reasoning.
Read the full abstract ▾
Materials Science
Transition-Metal-Induced Uniaxial Anisotropy in YFe$_3$
M. Nur Hasan, Bruno L. Fertig, Santa Pile et al. (+13 more)
Achieving uniaxial magnetocrystalline anisotropy while preserving high magnetization and thermal stability remains a central challenge in the development of Fe-rich permanent magnets. YFe$_3$ is a promising candidate owing to its saturation magnetization and Curie temperature, but its intrinsic easy-plane anisotropy limits its potential for permanent-magnet applications. Here, we show that selective transition-metal substitution drives an easy-plane-to-easy-axis transition in YFe$_3$. Systematic first-principles screening identifies V, Nb, and Ta as effective dopants, with YFe$_{2.33}$Ta$_{0.67}$ exhibiting a large uniaxial magnetocrystalline anisotropy energy of 2.3~MJ/m$^{3}$. Calculated exchange interactions and atomistic Monte Carlo simulations show that the enhanced anisotropy is achieved while preserving robust ferromagnetic order and high thermal stability. Micromagnetic simulations further reveal that YFe$_{2.33}$Ta$_{0.67}$ exhibits the highest coercivity and strongest resistance to irreversible demagnetization, whereas YFe$_{2.33}$V$_{0.67}$ achieves a comparable maximum energy product at 300~K. Experimental characterization of V-alloyed YFe$_3$ shows that the YFe$_3$ phase can be retained at low nominal V concentrations (1-3 at.%) under the employed synthesis and heat-treatment conditions, whereas higher V contents lead to increasingly multiphase samples. Concentration-dependent MAE calculations reveal a sign reversal between approximately 11 and 14~at.% V, marking the transition from easy-plane to uniaxial easy-axis anisotropy. Consistent with this concentration dependence, electron backscatter diffraction and Kerr microscopy reveal in-plane
magnetic domain structures in the experimentally accessible low-V regime. Therefore, our work offers a promising approach to tuning the magnetocrystalline anisotropy of YFe$_3$ while retaining high Curie temperatures.
Read the full abstract ▾
All authors M. Nur Hasan, Bruno L. Fertig, Santa Pile, Rafael M. Vieira, Alena Vishina, Alexander Kovacs, Gotthard Rieger, Stefan Denneler, Samuel J. R. Holt, Martin Lang, Swapneel Pathak, Andrea Petrocchi, Hans Fangohr, Thomas George Woodcock, Thomas Schrefl, Heike C. Herper
Category Theory
Actions in varieties of non-associative algebras: a survey
Manuel Mancini
This article surveys the main results on the representability of internal actions in varieties of non-associative
algebras over a field. We begin by recalling the general framework of internal actions and split extensions in semi-abelian categories, before specialising to Orzech categories of interest and varieties of non-associative
algebras. We discuss action representability, weak action representability, and action accessibility, with emphasis on the relationship between the latter two notions. We then describe the constructions of the universal strict general actor in Orzech categories of interest and the external weak actor in action accessible, operadic varieties of non-associative
algebras over a field, as well as their role in the study of weak action representability. We also give a tensor-domain formulation of partial
algebras and adapt the construction of the external weak actor to this setting. These notions are illustrated by examples arising from associative, Lie, Leibniz, Jacobi-Jordan, anti-commutative anti-associative, alternative, and $2$-nilpotent
algebras. We conclude with a summary of the known results and some open questions for future research.
Read the full abstract ▾
Artificial Intelligence
A dataset of formalised theorem statements from the Annals of Mathematics
Thomas Browning, Katerina Hristova, David Ledvinka et al. (+4 more)
Recent progress in autoformalisation of mathematics using AI has been impressive. Measuring this progress requires
datasets of formal statements that are both hard and trustworthy. We present AnnalsChallenge: 50 formalised
theorem statements written by hand in Lean 4 using Mathlib. Each is the main result of a paper published in the Annals of Mathematics since 2020. Every formalisation was manually reviewed and then audited by AI. We release them without formalised proofs.
Read the full abstract ▾
All authors Thomas Browning, Katerina Hristova, David Ledvinka, Justus Springer, Hang Lu Su, Jack McKoen, Kevin Buzzard
Data Structures and Algorithms
The Complexity of Hamiltonian Cycles and Paths with Prescribed Walk Lengths
Hitendra Kumar, Mathieu Mari and R. B. Sandeep
Fix a set $S$ of positive integers. An $S$-
Hamiltonian cycle is a cyclic ordering of all vertices of a graph such that each consecutive pair is joined by a walk whose length belongs to $S$. We study the complexity of deciding whether an input graph admits such an ordering, together with the corresponding path variants with prescribed or unrestricted endpoints. We complete the complexity classification for undirected graphs. In particular, we prove NP-completeness for every finite nonempty set of odd lengths, resolving an open problem posed by Amarilli, Lombardo, and Monet (WG 2026). For digraphs, all three problems are NP-complete for every fixed finite nonempty $S$ and admit linear-time
algorithms when $S$ is cofinite. For other fixed infinite sets $S$, we establish NP-complete and undecidable cases, with undecidability occurring even when $S$ is recursive.
Read the full abstract ▾
Computation and Language
LegalRewardBench: Benchmarking Reward Models for Legal Preference Evaluation
Hui Huang, Jie Zhou, Hongli Zhou et al. (+7 more)
Reward models are central to preference
alignment for legal tasks, yet existing reward model
benchmarks are still primarily centered on general-domain preferences. In this work, we introduce LegalRewardBench, the first human-verified
benchmark for evaluating reward models in the legal domain. The
benchmark organizes legal preference evaluation into five capability levels: knowledge, understanding, analysis, decision, and writing, and preference examples are constructed through rubric-guided automatic judging combined with human verification. We evaluate three families of judges, covering both general and fine-tuned models, and generative and discriminative architectures. Experimental results show that strong general-purpose generative models are currently the most robust legal reward models, while existing fine-tuned reward models transfer less reliably to legal preference evaluation. We also conduct detailed comparisons of input organizations and training strategies.
Read the full abstract ▾
All authors Hui Huang, Jie Zhou, Hongli Zhou, Qundong Shi, Shuyuan Zheng, Yini Su, Shaochun Wang, Biyuan Lin, Muyun Yang, Tiejun Zhao
Artificial Intelligence
Output Discipline Sets the Floor: A Falsification-Guided, Model-Agnostic Agent Protocol for Auditable Hypothesis-Driven Decision Making
Jing Liu, Yongxing Qi, Ping Yao et al. (+1 more)
What binds large
language models on auditable, hypothesis-driven decision tasks, where every candidate can be checked against the evidence? On the public TCMEval-SDT
benchmark for Traditional Chinese Medicine syndrome differentiation (200 cases), the binding constraint is output discipline and
reasoning structure, not model capability alone: two years of model scaling move zero-shot scores by only -0.03 to +0.08, while output-discipline failures (non-terminating
reasoning, format non-compliance, large run-to-run spread) persist even in frontier models. We test this with a model-agnostic
agent framework formalizing decision making as a five-stage closed loop, enforced by a candidate x evidence
matrix falsification protocol: every candidate is judged against every evidence item, at least half of the candidates must be falsified before a conclusion is licensed, and survivors are emitted in a machine-verifiable single-line format. Across five base models, the effect is sign-conditional on zero-shot output discipline: it rescues models whose discipline fails (DeepSeek-V4-Flash: +0.267/+0.271, entirely termination rescue), reduces a high-variance model's run-to-run standard deviation from 0.218 to 0.007 (Kimi K3), and significantly costs already-disciplined models (-0.087 to -0.120 on budget-matched Qwen3-4B/32B). Ablations attribute the cost to the candidate-bounded format layer and show the effect is independent of injected knowledge content (protocol-knowledge decoupling). The protocol also transfers to misinformation-robust mathematical
reasoning, detecting injected false premises at 89.8% recall with a 6.8% false-positive rate (MCC 0.54; a flag-everything baseline scores 0) and restoring accuracy to the no-misinformation ceiling. Where zero-shot output discipline is weak, the
reasoning protocol sets the floor; base-model capability always sets the ceiling.
Read the full abstract ▾
All authors Jing Liu, Yongxing Qi, Ping Yao, Ting Wu
Materials Science
A Topological Spin on Racetrack Memory
Benjamin Zingsem
Magnetic racetrack
memory requires localized carriers that combine rapid motion, resistance to perturbations and a detectable signal. Here we show how a 360{\deg} domain wall in a hollow ferromagnetic tube combines these properties within a single continuous
magnetic shell. Two successive magnetization reversals form a bound object whose surrounding domains point in the same direction. Analytical energetics and micromagnetic simulations establish its finite separation, recovery from tested perturbations and current driven transport without microscopic Dzyaloshinskii-Moriya interaction or a maintaining field. The texture's winding resists unwinding while radial rotation carries an energy cost, providing a specific control over stability. Relative rotation of the constituent
spins accounts quantitatively for additional displacement during current onset. The wall supports motion at hundreds of metres per second, with transient velocities approaching one kilometre per second at stronger drive. A prepared seven-bit pattern retains its carriers and local readout contrast after pulsed transport and recovery. Curved-track and differential-current tests identify radial stiffness and drive uniformity as controls for intact motion. These results establish a physical basis for fast, locally distinguishable tubular
magnetic bits and identify a continuous switching pathway that can be targeted by tailored excitation.
Read the full abstract ▾
Classical Analysis and ODEs
Weighted Norm Inequalities for Maximal Multi-sublinear Rough Singular Integrals
Bae Jun Park
We continue our study of multilinear rough homogeneous singular integrals by establishing weighted estimates for their maximal truncations. Let $m\ge2$ and let $\Omega\in L^q(\mathbb{S}^{mn-1})$, $1<q\le\infty$, have mean zero. For $1<p_1,\dots,p_m<\infty$ with $p_j\ge q'$ and $1/p=\sum_{j=1}^m1/p_j$, we prove $$\big\Vert \mathcal{L}_\Omega^*(f_1,\dots,f_m)\big\Vert_{L^p(v_{\vec{\boldsymbol{w}}})} \lesssim \Vert\Omega\Vert_{L^q(\mathbb{S}^{mn-1})} \prod_{j=1}^m\Vert f_j\Vert_{L^{p_j}(w_j)}$$ whenever $\vec{\boldsymbol{w}}=(w_1,\dots,w_m)\in A_{(p_1/q',\dots,p_m/q')}$, where $q'$ is the Hölder conjugate of $q$ and $v_{\vec{\boldsymbol{w}}}=\prod_{j=1}^m w_j^{p/p_j}$. The estimate holds under the same multiple weight condition as the corresponding principal value estimate.
Read the full abstract ▾
Computer Vision
Robust Logical Anomaly Detection under Capture Shifts via Synthetic Anomalies
Ufuk Efe and Alptekin Temizel
Detectors of logical anomalies, such as missing, extra or wrongly combined parts, usually score a test image by its distance from normal (good) images. They achieve near-perfect image-level AUROC on standard industrial anomaly detection
benchmarks such as MVTec LOCO, but their robustness to changes in capture conditions remains largely unexamined. Under a heterogeneous perturbation protocol in which each test image is randomly kept unaltered or subjected to a brightness reduction or a zoom-out transformation, the normal-reference methods we tested lose 9.7 to 34.1 AUROC points, and two of the strongest, CSAD and SALAD, fall from 97.7% and 96.4% to 68.1% and 62.3%. We propose a framework in which a classifier, trained on normal images and on synthetic anomalies, scores a test image without comparing it with normal data. From a short description of each product, a
language model derives anomaly types and editing instructions, an image editor applies them to normal images, and the classifier learns to separate the edited images from automatically discovered normal sub-modes. With its full training recipe, the framework reaches 94.9% AUROC on clean images, 1.5 to 2.8 points below SALAD and CSAD, and 94.1% under the perturbation protocol. Trained without brightness or zoom-out augmentation, so that neither test perturbation occurs in training, it still reaches 92.5% (93.2% clean). These results suggest that clean
benchmark scores can hide a strong sensitivity to capture conditions, and that learning the decision from both normal images and rule violations may reduce it.
Read the full abstract ▾
Machine Learning
Which Optimizer Leaks Most? Optimizer Implicit Bias as a Training-Data Privacy Controller
Sagi Habani, Yehonathan Refael and Ofir Lindenbaum
Releasing a neural network checkpoint can leak information about its training data, and existing reconstruction results rely on gradient-descent training. We ask whether the choice of optimizer affects the ability to recover the training data. Extending work connecting Karush–Kuhn–Tucker conditions to training-data reconstruction, we show that the implicit biases of popular optimizers (SGD, Adam,
Muon) leave very different fingerprints of the training data. We argue that the key distinction is the weight-specificity of these implicit biases, namely how much continuous information about the final weights the optimizer-dependent target of the stationarity conditions encodes. The limiting stationarity conditions for models trained with SGD and
Muon retain enough weight-specific information to reconstruct training examples in practice. For Adam in the $\ell_\infty$ implicit-bias regime, the optimizer-dependent part of these conditions encodes only a discrete pattern of maximal coordinates and their signs, and at nondegenerate KKT points the conditions are too few to identify the training set to first order. Experiments on MNIST and CIFAR-10 support these claims: models trained with
Muon can yield stronger reconstructions at much lower training cost than models trained with SGD, although this advantage depends on the training regime and architecture, and models trained with Adam appear consistently more resistant to the evaluated attacks.
Read the full abstract ▾
Differential Geometry
A topological sphere theorem for pinched Legendrian submanifolds
Qing Cui and Linlin Sun
Let $M^n$ be a closed simply connected Legendrian submanifold of $\mathbb S^{2n+1}$, where $n\geq5$, and let $\mathbf{B}$ and $\mathbf{H}=\operatorname{tr}\mathbf{B}$ denote its second fundamental form and mean-
curvature vector. We prove that $M$ is homeomorphic to $\mathbb S^n$ if \begin{equation*} |\mathbf{B}|^2\leq \frac{(n-1)(n+2)}n+\frac{n^2+3n-2}{2n^2}|\mathbf{H}|^2-\frac{(n-1)(n-2)}{2n^2}|\mathbf{H}|\sqrt{|\mathbf{H}|^2+4n}. \end{equation*} The equality can be attained by the Calabi products, so simple connectivity cannot be omitted. For $n\geq4$, we also prove a sphere
theorem under an explicit bound and the weaker assumption $b_1(M;\mathbb R)=0$. The proof combines phase-rotated Lawson–Simons variations with a systolic inequality for finite two-complexes. We also give local constructions showing that the pinching condition allows negative isotropic
curvature at a point.
Read the full abstract ▾
Machine Learning
Reliable Risk Estimation with Kernel Mean Matching under Dependent Sampling
Alexey Zaytsev, Diana Koldasbayeva and Egor Serov
Kernel Mean Matching (KMM) estimates risk under
covariate shift by reweighting labelled source samples so that their kernel mean
embedding matches that of unlabeled target samples. Existing finite-sample guarantees are usually stated for independent samples, although many applications involve temporal, spatial, or clustered dependence. We give a finite-sample KMM analysis under dependent sampling by replacing nominal sample sizes with effective sample sizes in the concentration steps used by the classical proof. The approach uses Bernstein inequalities for geometrically strongly mixing processes and derives a model-based effective sample size. The resulting bound separates overlap, dependence, and smoothness effects with specific constants derived under a Gaussian process assumption. This analysis yields an additional tradeoff: smoother losses preserve more input dependence and therefore reduce effective sample size, while also reducing function complexity and empirical-to-population risk gaps. Applying the result to prediction yields confidence intervals that achieve near-nominal coverage in synthetic experiments and substantially reduce undercoverage on real monitoring data.
Read the full abstract ▾
Computer Vision
Unsupervised and Supervised Approaches for Infrared Small Target Detection : Overview, Challenges and Perspectives (Part 2)
Thierry Bouwmans, Anastasia Zakharova, Belmar Garcia et al. (+3 more)
Since the rise of supervised deep learning modelet dets, researchers has attempt to solve infrared small targection using various models like conventional neural networks, GANS,
Transformers, Mixture of Experts and recent graph/hypergraph neural networks. Moreover, these deep neural networks methods surpassed the traditional unsupervised models and the low-rank and sparse decomposition unsupervised models reviewed in "Traditional and Recent Approaches for Infrared Small Target Detection : Overview, Challenges and Perspectives (Part 1). In this context, the aim of this paper is to provide a complete survey of the deep learning models in infrared small target detection. First, we categorize the different approaches found in the literature in terms of the modeling models of strategies. Then, we have discussed them in terms of the critical situations that they claim to handle. Furthermore, we present the specific performance evaluation. Then, we conclude with several promising directions for future research.
Read the full abstract ▾
All authors Thierry Bouwmans, Anastasia Zakharova, Belmar Garcia, Marwa Chendeb El Rai, Meghna Kapoor, Badri Narayan Subudhi
Systems and Control
On Over-Conservatism Under Low-Probability Extreme Events: A KL-Regularized Wasserstein Distributionally Robust Planning Method
Xinda Shen, Yifan Su, Bo Hu et al. (+3 more)
The increasing penetration of renewable energy makes power systems more vulnerable to low-probability extreme events, elevating system planning to a critical role in ensuring reliability. Distributionally robust optimization (DRO) based on the Wasserstein distance is widely used to address this challenge. However, we find that the conventional Wasserstein ambiguity set tends to excessively amplify the probabilities of low-probability extreme events, leading to over-conservative investment decisions. To mitigate this over-conservatism, this paper proposes a novel ambiguity set that modifies the Wasserstein distance with a Kullback-Leibler (KL) divergence regularization term. By leveraging the KL divergence's sensitivity to relative probability changes, the proposed set effectively curbs the unrealistic magnification of low-probability events. A two-stage distributionally robust planning model built upon this new ambiguity set is developed to optimize the installed capacity of photovoltaic, wind, and energy storage. The complex model is then reformulated as an equivalent exponential cone program with strict theoretical proofs, making it computationally tractable. Numerical results on the PJM 5-bus and IEEE 118-bus systems reveal the over-conservatism of standard Wasserstein DRO under low-probability extreme events and validate the effectiveness and efficiency of the proposed planning model and solution
algorithm.
Read the full abstract ▾
All authors Xinda Shen, Yifan Su, Bo Hu, Yanbo Chen, Changzheng Shao, Tao Wu
Computer Vision
LongForcing: Scaling Temporal Manifold Matching for Long-Horizon World Rollout
Ziyu Zhu, Zhaoxu Sun, Fan Jiang et al. (+5 more)
Autoregressive
video diffusion models have emerged as a foundation for long-horizon world rollout. However, when extended far beyond their training windows, these models suffer from progressive drift: visual quality deteriorates, motion becomes inconsistent, and object identity gradually collapses as generated frames are recursively reused as context. We study a supervision-horizon gap: short-horizon
distillation leaves later rollout states without direct supervision. We show that matching short-window marginals need not determine long-horizon behavior, while longer on-policy trajectories expose delayed failures to training. This motivates temporal
manifold matching at scale for long-horizon world models. Building on this insight, we introduce LongForcing, a
distillation approach that scales teacher supervision to long student self-rollouts. LongForcing preserves the short-horizon teacher-forcing and causal ODE-
distillation stages, while replacing the realscore teacher in the final distribution-matching stage with an extended-horizon bidirectional teacher. The long-horizon teacher supervises extended student self-rollouts using the standard distribution-matching formulation, without modifying the student architecture or increasing inference cost. We evaluate LongForcing on text-to-
video generation and interactive action-conditioned world models, achieving a VBench Quality score of 85.52 and improving WorldRoamBench 60s action score by 23.9 points in first-person ablations. Qualitative stress tests further sample interactive rollouts up to 24 hours, far beyond the training horizon. These results suggest that scaling supervision along the temporal axis is a practical step toward long-horizon world models. The project website and code are at https://github.com/Fantasy-AMAP/LongForcing.
Read the full abstract ▾
All authors Ziyu Zhu, Zhaoxu Sun, Fan Jiang, Mengchao Wang, Junfeng Ni, Ting-Bing Xu, Zhicheng Liu, Mu Xu
Mesoscale and Nanoscale Physics
Role of dissipation dilution in achieving high mechanical quality factors in van der Waals resonators
Lukas Schleicher, Leonard Geilen, Alexander Musta et al. (+5 more)
Dissipation dilution has driven a revolution in the quality factors ($Q$) of top-down nanomechanical resonators. Yet it has not been leveraged for van der Waals (vdW) materials. Here we report quality factors up to $372{,}000\pm37{,}000$ in monolayer MoS$_2$ drums at 4 K, nearly an order of magnitude beyond the previous best for any vdW material at this temperature, enabled by a clean,
polymer-free exfoliation technique. We develop a dissipation-dilution model for circular drums and analyze the contributing intrinsic and extrinsic dissipation mechanisms. The MoS$_2$ monolayer exhibits a dilution factor of order $1,000$, establishing sub-nanometer-thin resonators as ultimate dilution materials. The intrinsic quality factor, however, is limited by surface losses at atomic thickness. Conversely, multilayer drums show a lower dilution factor but a high intrinsic quality factor, on par with established top-down materials of similar dimensions, both resulting from the increased thickness. A two-level-system analysis of the temperature-dependent damping confirms this bulk-versus-surface picture and provides first insight into its microscopic origin. Our results identify dissipation dilution as the governing mechanism for high-$Q$ mechanical vdW resonators. Combined with strategies such as soft clamping or stress engineering, they establish atomically thin resonators as a platform for ultrahigh-$Q$
quantum transduction and sensing.
Read the full abstract ▾
All authors Lukas Schleicher, Leonard Geilen, Alexander Musta, Niklas Walter, Anne Rodriguez, Philipp Bredol, Alexander W. Holleitner, Eva M. Weig
Logic in Computer Science
Finite-on-Top-of-Temporal CSPs in Fixed-Point Logic
Manuel Bodirsky and Philipp A. Grzywaczyk
We study definability in fixed-point logic with counting (FPC) for a class of constraint satisfaction problems (CSPs) that simultaneously generalises finite-domain CSPs and temporal CSPs. The templates for the CSPs are first-order expansions of the structure $([k] \times {\mathbb Q};<_1,\dots,<_k,\sim)$ where $<_i \, := \{((i,x),(i,x')) \mid x<x'\}$ and $\sim \, := \{((i,x),(i,x')) \mid i \in [k], x,x' \in {\mathbb Q}\}$, i.e., $k$ copies of $({\mathbb Q};<)$. Such a CSP is definable in FPC precisely when the template pp-constructs neither the ternary temporal relation $X$, which holds on three elements if the minimum is attained exactly twice, nor a nontrivial finite affine structure. Equivalently, the template has global polymorphisms that witness the tractability of the temporal CSPs on the copies, and that provide 3-4 weak near-unanimity operations on the set of copies of ${\mathbb Q}$. In this case, we prove that the CSP is even expressible in the less expressive logic FP$^*$, which is the restriction of FPC without counting terms (i.e., the extension of first-order logic by a fixed point operator and an ordered auxiliary number sort). Our
algorithm samples finitely many values in each copy and then performs nested constraint-singleton consistency tests (which is a uniform
polynomial-time
algorithm for finite-domain CSPs). The correctness proof uses finite-domain CSP theory to combine consistency certificates by a single polymorphism term, fixing each variable to one copy without destroying the certificates needed for the temporal
algorithm. We then use a known normal form for algebraic products to control the remaining interactions between different copies.
Read the full abstract ▾
Superconductivity
Phonon-Mediated Superconductivity near Phase Boundaries in Rhombohedral Multilayer Graphene
Konstantinos Sourounis and Peter M. Oppeneer
The recent observation of superconductivity in untwisted multilayer
graphene has sparked intense interest in its microscopic origin. As the displacement field and carrier density are varied, superconductivity emerges in several distinct regions with different temperature and
magnetic-field responses. This behavior has been attributed to the presence of distinct electronic phases, raising the possibility that different superconducting regions originate from different microscopic mechanisms. While purely electronic pairing mechanisms have been extensively explored, an in-depth examination of
phonon-mediated superconductivity is still missing. Here, we solve the coupled selfconsistent Migdal-Eliashberg equations using input from first-principles calculations and incorporating interaction-driven
symmetry breaking of the electronic ground state. Varying the carrier density and displacement field, we reproduce the experimentally observed superconducting regions with comparable superconducting properties such as gap and $T_c$ and predict additional, as-yet undiscovered superconducting phases. Our findings exemplify that several superconductivity regions in rhombohedral
graphene can be satisfactorily explained by
phonon-mediated pairing.
Read the full abstract ▾
Astrophysics of Galaxies
EMU detection of mysterious multiple component radio-continuum emission from nearby spiral Seyfert I type galaxy 'Sand-timer' (NGC 7213)
M. D. Filipović, Z. J. Smeaton, B. S. Koribalski et al. (+24 more)
We present an Australian Square Kilometre Array Pathfinder (ASKAP) Evolutionary Map of the Universe (EMU)
radio-continuum detection of the nearby spiral Seyfert I type
galaxy NGC 7213 which we nickname `Sand-timer' due to its morphological appearance. We find intriguing, peculiar emission extending from the central supermassive black hole (SMBH), which we refer to as
galaxy hyperbubbles and/or active galactic
nucleus (AGN) jets. Finally, we find two previously undetected kiloparsec-scale features in the polarised intensity image, which appear most likely also related to the central SMBH, as restarted bipolar
radio jets, though somewhat obscured in the
radio-continuum total intensity image. They resemble the bow shocks often seen in Fanaroff-Riley Class II (FR-II)
radio galaxies, and so we propose they most likely represent an early episode of the `Sand-timer's' AGN jet activity. An alternative scenario is that the core of NGC 7213 may be a supermassive binary black hole (SMBBH) undergoing a tidal disruption event (TDE), potentially causing the peculiar
radio-continuum emission.
Read the full abstract ▾
All authors M. D. Filipović, Z. J. Smeaton, B. S. Koribalski, R. Kothes, A. Ahmad, R. Z. E. Alsaberi, C. Bordiu, A. C. Bradley, E. J. Crawford, A. M. Hopkins, E. Järvelä, S. Lazarević, D. Leahy, J. L. Payne, S. F. ur Rahman, G. P. Rowell, H. Sano, M. Selby, S. S. Shabala, E. J. Stanton, S. Taziaux, K. Tsuge, D. Urosević, N. F. H. Tothill, J. Th. van Loon, V. Velović, G. L. White
Computer Vision
Efficient Scene-Text Refinement via Pixel-Space Grouped Image Generation
Yaoli Liu, Guo-Hua Wang, Yiyang Chen et al. (+4 more)
Generated images often have coherent layouts and visual content, yet local errors in small text and complex characters still limit their readability and practical use. Scene-text refinement requires accuracy, preservation, and efficiency. Existing methods based on variational autoencoders (VAEs) can lose fine strokes during reconstruction, while full-image editing can shift background colors. In this work, we introduce PixText, a model that formulates local text refinement as grouped image generation in pixel space. Conditioned on the original image, PixText generates only the target regions, preserves unedited content, and uses MeanFlow for four-step generation. We further introduce Text Refinement Bench to evaluate text correction and image preservation in generated images. On this
benchmark, PixText achieves 83.06% glyph pass and 94.66% exact target-character OCR accuracy, versus 36.77% and 52.31% for the strongest baseline. Its four-step MeanFlow variant retains an 81.84% pass rate with less than one GPU-second of amortized generation cost per image.
Read the full abstract ▾
All authors Yaoli Liu, Guo-Hua Wang, Yiyang Chen, Qing-Guo Chen, Zhao Xu, Weihua Luo, Yao-Xiang Ding