Systems and Control
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Sunday 6 September
Long-duration energy storage (LDES) can mitigate prolonged renewable–load imbalances during Dunkelflaute events, but existing studies rely on zonal or linearized DC network models and inadequately analyze the operational feasibility of the grid across a wide range of full-year renewable, load, and contingency scenarios. To address this gap, this paper introduces a multi-period alternating-current optimal power flow (AC-OPF) formulation that captures the full nonlinear network physics and assesses the reliability value of LDES in high-renewable grids. To evaluate scarcity events spanning multiple days to weeks, full-year operation is modeled as an 8,760-h load-shedding minimization over scenarios sampled from a Gaussian copula model, fitted to 2010–2025 historical wind and bus-level load data, that represents both typical variability and tail events. On a synthetic 200-bus Illinois transmission network, a hybrid fleet of battery energy storage (BESS) and LDES with 50 MW total power reduces annual load shedding by 83.0% on average relative to the base network, versus 68.5% for short-duration BESS alone at equal power. To further account for unexpected line outages throughout the year, the formulation is extended to a multi-day security-constrained AC-OPF. Under N-1contingencies, no feasible operating solution is obtained for the base network, whereas the LDES-equipped network remains feasible in all considered cases, thereby saving the cost of additional generation and transmission capacity. During the contingency period, LDES acts as a backup power supply, requiring only 5.5% more generation on average than the no-contingency base case.
Earth and Planetary Astrophysics
The Effect of X-Ray and Accretion Variability on Mid-Infrared Lines in the Young Disk-Bearing Binary DQ Tau
A star's radiation field is the main energy source of a planet-forming disk. Stellar photons—from X-rays to optical—are reprocessed in the inner protoplanetary disk enabling mid-IR diagnostics of the disk's physical conditions. Most pre-main-sequence stars accrete at moderate rates and emit X-rays efficiently. Such activity is expected to drive stochastic variability in the stellar spectrum. It is not clear to what extent such variability affects the thermochemical structure of protoplanetary disks. We aim to examine the effect of routine accretion variability and X-ray flares on the thermochemical structure and mid-IR line emission spectra of protoplanetary disks. We combine multiepoch, contemporaneous X-ray and NUV/optical observations of the DQ Tau binary with JWST/MIRI spectra of its circumbinary disk. We interpret the data and assess the effect of a varying stellar spectrum on the disk using a thermochemical model illustrative of the system. Synthetic fluxes of CO, CO2, HCN, and H2O are systematically stronger at periastron than at apastron due to an enhanced accretion luminosity. Variations in the integrated fluxes lie within a factor of two between epochs, in agreement with the JWST/MIRI spectra of DQ Tau. Routine stellar variability induces only modest changes in the disk's structure, yet it can still shift the snowline outwards by 17 percent relative to its apastron position. Because the JWST observations did not coincide with X-ray flares, our flare results rely on simulations: moderate-intensity flares leave most mid-IR molecular emission unchanged; in contrast, X-ray flares strongly enhance the abundances and mid-IR luminosities of H I, Ar II, and Ne II in the disk surface. Our results imply that the Spitzer/JWST line variability observed in non-outbursting systems is consistent with moderately variable stellar spectra driven by accretion and magnetospheric activity.
Read the full abstract
Machine Learning
Feature Superposition in Neural Networks: From Theory to Practice
Superposition refers to neural networks representing more features than they have dimensions. It was proposed to explain polysemantic neurons, which respond to several features, and also other phenomena, such as neural scaling laws and adversarial vulnerability. A feature stored in superposition is useful to the extent that its value can be read back despite interference from the other features. The theory of superposition mainly asks how wide a layer must be for the features to be read back, and by which decoder. In theoretical models, a single linear map that reads back every combination of nonzero features needs a width nearly quadratic in their number, while sparse optimization can succeed with a width linear in that number. In practice, one may hope to interpret a trained network by recovering its features in the same way. However, its features are usually unknown, and thus they are estimated from its activations or weights by recovery methods, such as sparse autoencoders and transcoders. In this work, we organize the theory of superposition by the decoder class and its practice by the strength of the claims about a feature supported by each method's results. On the theory side, we review the lower bounds on the overlap between feature directions, the conditions under which trained toy models share dimensions among features, and the separation between storing features and computing new features. On the practice side, we categorize recovery methods into four families, from probes for a named feature to decompositions of the trained weights. We then compare their evaluations, failure modes, and applications, and grade the evidence on one scale of claims about a feature. Finally, we assess previously stated open problems and identify remaining theoretical and empirical questions about superposition in trained networks. Project page: https://eethanshi.github.io/awesome-superposition/
Read the full abstract
General Relativity and Quantum Cosmology
Dissection of a merger-ringdown waveform in the small-mass-ratio limit
Work over the past two decades has unveiled the rich phenomenology of black hole binary mergers and subsequent ringdowns, involving a tapestry of quasinormal modes (QNMs), nonlinearities, tails, transients, and secular effects including gravitational memory. Here we develop a framework for analyzing nonlinear merger-ringdown features in the small-mass-ratio limit, where individual effects can be cleanly isolated. Specializing to the case of a quasicircular, nonspinning black hole binary, we find the waveform sharply divides into a pre-merger extended inspiral phase, a merger phase lasting roughly half a cycle, and a post-merger ringdown dominated by QNMs. We show quadratic QNMs dominate over linear overtones in the ringdown phase for comparable-to-intermediate mass ratios, and we highlight nonlinear effects of gravitational-wave memory, including cubic wave-zone phenomena analogous to horizon absorption effects.
Read the full abstract
Geometric Topology
Bridge numbers of virtual graphs and handlebody-links
We study bridge indices of graphs, virtual graphs, and their handlebody classes. We construct virtual ravels for which the difference between bridge index and weighted overpass bridge index is unbounded. The lower bound uses exponentially many colorings for one fixed flow in a $\Z_2$-family of biquandles. For each negative Euler characteristic we construct classical almost unknotted graphs of one fixed trivalent graph with arbitrarily large bridge index. We also obtain unbounded even bridge indices for ravels of each fixed bouquet rank. A separate almost unknotted family has an unbounded difference between its spatial-graph bridge index and the bridge index of its regular neighborhood. Finally, edge connected sums of Suzuki curves give exact constrained and unconstrained bridge indices whose difference is unbounded.
Read the full abstract
Probability
Normal random variables
This is an advanced introduction to the various types of normal random variables, with most of the needed preliminaries included. We first discuss the probability basics, standard central limits, and the theory of the usual, real normal variables, with examples, illustrations and numerous formulae. Then we go on a similar discussion regarding the complex normal variables, and the Rayleigh variables too. We then move to arbitrary dimensions, with a discussion regarding the Gaussian vectors, and related probability laws, featuring some functional analysis, and geometry and physics too. Finally, we provide an introduction to the quantum versions of the normal variables, and notably to those coming from free probability and random matrices.
Read the full abstract
Materials Science
Magneto-spin-orbit graphene: Chern topology and photoinduced Hall response
Magneto-spin-orbit graphene can host Chern phases due to proximity interactions, but in realistic heterostructures these depend on substrate chemistry, local stacking registry and may be altered by buried-interface reconstruction. Investigation of these effects requires special efforts to take into account the long-range moire patterns that often arise for these heterostructures. We address this question using first-principles calculations and effective four-band models for four out-of-plane magnetized Gr/(Au, Pt)/(Co, Mn) heterostructures in commensurate (2 x 2) and moire (9 x 9) supercells where the latter include misfit dislocations for the Co-based interfaces. The commensurate cells overestimate charge transfer and exaggerate valley asymmetry, whereas moire averaging strongly reduces both effects while preserving sizable spin splittings. The Pt spacer mediates stronger proximity interactions than Au, while replacing Co by Mn substantially changes the balance of the effective exchange interactions in a spacer-dependent manner. Introduction of misfit dislocations into the Co-based structures is selective and tunes the Dirac gap in Gr/Au/Co or spin splittings in Gr/Pt/Co. Graphene in Gr/Pt/Co exhibits effective ferromagnetic Chern phase with C = 2 in all studied supercells. By contrast, the moire models of the other three compositions are ferrimagnetic and topologically trivial, with C = 0. Nevertheless, they exhibit helicity-dependent photoinduced Hall responses, which are strongly enhanced by misfit-dislocation reconstruction in Gr/Au/Co. This work demonstrates that moire-scale stacking controls the effective Chern phase and Hall conductivity, whereas buried-interface defects tune the photoinduced Hall response without changing the topology.
Read the full abstract
High Energy Physics (Phenomenology)
Geometric Phases in Two-Level Mixing: From Neutral Mesons to Holonomic Qubits and Topological Majorana Modes
We investigate the geometric structure of a Hermitian two-level mixing Hamiltonian motivated by neutral meson oscillations and its connections with qubit, fermionic, and topological descriptions. Mapping the Hamiltonian to an effective qubit representation, we analyze the geometric phase associated with cyclic parameter evolution on the Bloch sphere. Without identifying the complex mixing phase with a physical CP-violation observable, we interpret it as a CP-like geometric parameter controlling the azimuthal orientation of the Hamiltonian vector and its reversal under phase conjugation. Third-order Bargmann invariants constructed from the eigenstates yield an associated discrete phase whose dependence on the mixing parameters is examined alongside the continuous geometric construction. The cyclic evolution is also represented as a single-qubit $R_z$ phase operation and expressed through Majorana bilinears. Extending the construction to the momentum-dependent Bogoliubov-de Gennes Hamiltonian of the Kitaev chain, we relate momentum-space winding to the topological regimes and the quantized Berry (Zak) phase in the real-parameter model, while the corresponding Bargmann construction approaches the global phase in the continuum limit. Together, these results provide a geometric perspective connecting phase structure, qubit operations, fermionic representations, and momentum-space topology within related two-level Hamiltonians.
Read the full abstract
Logic
Vopěnka's Principle without Choice: Preservation under Symmetric Extensions
We give a detailed proof that every set-sized symmetric extension of a model of $\mathsf{ZF}+\mathsf{VP}$ again satisfies $\mathsf{ZF}+\mathsf{VP}$, where $\mathsf{VP}$ is formulated for arbitrary set-sized languages. For each standard $n\geq 1$, the same preservation theorem holds for $\mathsf{VP}(\Pi_n)$, with set parameters and arbitrary set-sized languages. We also prove that $\mathsf{ZF}+\mathsf{VP}$ and $\mathsf{ZFC}+\mathsf{VP}$ are equiconsistent, and that $\mathsf{ZF}+\mathsf{VP}$ proves the existence of a proper class of Löwenheim-Skolem cardinals. Relative to $\operatorname{Con}(\mathsf{ZF}+\mathsf{VP})$, the axiom $\mathsf{DC}$ is independent of $\mathsf{ZF}+\mathsf{VP}+\neg\mathsf{AC}$. The preservation theorem also yields Feferman-Lévy and full Solovay models satisfying $\mathsf{VP}$. We consider whether every countable model of $\mathsf{ZF}+\mathsf{VP}$ has a class-generic extension satisfying $\mathsf{ZFC}+\mathsf{VP}$.
Read the full abstract
Cryptography and Security
A Queryable Graph-Based Security Analysis Framework for O-RAN
The Open Radio Access Network (O-RAN) replaces vendor-locked RANs with a modular and interoperable architecture that fosters competition and accelerates innovation. With this openness comes increased complexity and a larger attack surface, making security a critical concern. Today, assessing O-RAN security requires manually cross-referencing dozens of specifications, vendor whitepapers, and academic studies, which is error-prone and static. In this paper, we present a graph-based framework that transforms this static corpus into a single, queryable database. Our graph representation contains over 350 nodes and more than 1,250 relationships, distilled from specifications, academic papers, open-source projects, and vulnerability databases. To keep this resource current, we integrate a hybrid data extraction pipeline that couples deterministic parsing of structured specifications with Large Language Model (LLM)-assisted extraction for evolving specifications and unstructured literature. Querying the graph reveals three actionable findings within our curated corpus: critical infrastructure such as the O-DU, SMO, and O-Cloud carries dozens of specification-level threats yet has little or no empirical coverage; memory-safety weaknesses account for 11 of the 21 CWE occurrences associated with the analyzed CVEs; and fuzzing uncovered 18 of the 20 CVEs attributed to research papers. We provide the database, pipeline, and queries as open-source artifacts.
Read the full abstract
Statistical Methodology
Cluster-Based Dimensionality Reduction by Nonparametric Distributional Screening
We consider dimensionality reduction for high-dimensional observations accompanied by a supplied partition into two or more clusters. The objective is not to construct a low-rank projection, but to retain an interpretable subset of the original coordinates that preserves the distributional information distinguishing the clusters. For each coordinate, the proposed procedure compares the cluster-specific empirical distribution functions through a several-sample Kolmogorov-Smirnov separation statistic. We formalize the resulting marginal cluster support and establish simultaneous finite-sample concentration over all coordinates, explicit bounds for false inclusions and omissions, and exact support recovery when the minimum distributional separation dominates the high-dimensional stochastic error. We also quantify the dimension inflation induced by using an unadjusted testing level and give a familywise-error-controlled version. Under a conditional sufficiency condition, sure screening preserves the full-data posterior cluster probabilities, mutual information, and Bayes risk; an additional result characterizes robustness to imperfectly estimated cluster labels. The procedure is invariant to strictly increasing coordinate transformations and can retain low-variance cluster signals that principal components may discard. We further develop average dual information, a criterion combining partition agreement after transformation with structural coverage of cluster-relevant coordinates, and derive its basic properties and consistency. Simulations illustrate the theory, the interpretability of the selected coordinates, and the distinction between cluster-directed screening and variance-directed projection.
Read the full abstract
Cryptography and Security
Characterizing Contention-Induced Reliability Collapse in KV-Cache Timing Side Channels for Multi-Tenant LLM Serving
Shared key–value (KV) cache reuse improves large language model (LLM) serving, but it can also create a timing side channel that reveals whether a prefix is already cached. Previous work shows that such attacks are possible, but their reliability under realistic multi-tenant contention is less understood. We study this problem through seven experiments on live shared LLM-serving systems. On a vLLM server running DeepSeek-R1-Distill-Llama-8B on NVIDIA GB10, mean Cohen's d drops from 0.7789 with no synthetic workers to 0.2109 with two workers (t=8.412), while higher worker counts cause no statistically detectable further loss. A 120-run sparse-overlap experiment places the best breakpoint at the boundary of the measured range (tau=0, 95% CI [0.000,0.113]), supporting an ambient-versus-loaded regime change rather than an internal physical threshold. AUROC falls from 0.650 at ambient to 0.531 near 61% overlap and partially recovers to 0.574 at saturation. Concurrency-depth variance is the strongest measured correlate of effect size (r=-0.416) and hit consistency (r=-0.637). An interleaved control preserves the same non-monotonic ordering. The main collapse is also reproduced on a real two-node, two-GPU tensor-parallel vLLM setup, where mean d falls from 3.418 to 0.511 (p<0.01). Two SGLang pilots are statistically inconclusive. Overall, KV-cache timing reliability depends strongly on the load regime and serving stack, and measurements on quiet systems can overestimate operational attack reliability.
Read the full abstract
Robotics
ContextFlow: In-Context Flow Matching for Robot Manipulation
Although highly effective in vision and language domains, applying in-context learning to robotics remains challenging. Existing autoregressive in-context imitation methods discretize continuous actions and exacerbate the accumulation of early prediction errors through next-token prediction, limiting their generalization on unseen task configurations. Meanwhile, flow-matching policies have been explored for continuous robot control and can help mitigate compounding errors; however, in-context imitation learning within a flow-matching framework remains underexplored. To address these limitations, we introduce ContextFlow, a conditional flow-matching model that learns continuous action distributions for in-context imitation learning. ContextFlow conditions flow-based action prediction on demonstrations and observations, enabling robust generation from noisy action distributions. To better encode multimodal in-context demonstrations, we adapt perceiver-style multimodal context compressors that distill visual, proprioceptive, and action sequences into compact, task-relevant latent representations. On LIBERO, ContextFlow outperforms ICRT by 35 percentage points in average success rate on unseen task configurations, while matching the performance of the task-specific fine-tuned VLA model $\pi_0$ without any fine-tuning on unseen tasks. On real robots, it generalizes to unseen configurations of both single-arm and bimanual tasks, achieving 40% success on a new pen-uncapping configuration. Project Page: https://dingjiansw101.github.io/contextflow-page/.
Read the full abstract
Computation and Language
You Are What You Read: Misalignment via In-Context Persona Induction
Broad misalignment has been produced by finetuning on narrow data, harmful or benign, and in context only by demonstrations of the undesirable behaviour itself. We show that benign data suffices in context, with no finetuning and no demonstration of harmful behaviour in the prompt. Biographical facts that converge on a single figure, placed in a model's context as ordinary conversational turns, lead it to answer as that figure on questions the facts never touch. We call this persona induction. Across nine personas and thirteen models, identity adoption rises sigmoidally with the number of facts and crosses 50% within 3 to 10 of them. Misalignment then tracks which figure is described. Harmless personas reach full adoption with near-zero misalignment, while harmful ones voice their characteristic views on unrelated questions, at rates up to 80%. A formatting instruction can gate when the persona activates. Because each fact is individually benign, accumulated biographical context is flagged by content filters on 3% of inputs against 24-33% for an equivalent direct instruction.
Read the full abstract
Distributed Computing
Don't Be Afraid to Die: Black Hole Search in Dynamic Graphs with Fewer Agents
We consider a team of synchronous mobile agents operating in a port-labeled network. There is one node in the network, called a black hole, that permanently destroys any agent that visits the node. The team of agents must safely locate the black hole, i.e., at least one agent must survive, terminate its algorithm at a node adjacent to the black hole, and output the port number that leads to the black hole from its current position. In the setting where the network is a 1-bounded 1-interval connected dynamic graph, Kaur et al. (SSS 2025) showed that a team consisting of $2\delta_{BH}+17$ agents is sufficient to solve the task from a scattered configuration, where $\delta_{BH}$ denotes the degree of the black hole node. We show that $2\delta_{BH}+3$ agents are sufficient, nearly matching the $2\delta_{BH}+1$ lower bound provided in Kaur et al. (ICDCN 2025).
Read the full abstract
Artificial Intelligence
Learning transferable human physiology from two million hours of sleep with SleepFM-2
Sleep provides a nightly window into health by capturing coordinated activity across the brain, heart, muscles and respiratory system. We introduce SleepFM-2, a sleep foundation model developed and evaluated on 282,511 polysomnography recordings from 26 cohorts, including 235,865 used for pretraining. These data span more than two million hours of multimodal physiology. Compared with SleepFM, SleepFM-2 improves disease prediction and sleep scoring, supports arousal, limb movement and respiratory event detection, and transfers to wearable sensing and subjective sleep phenotypes. A model combining its PSG representation with age, sex and BMI met a prespecified discrimination and significance criterion for 188 subsequently recorded EHR phenotypes in two held-out cohorts, including one health system unseen during pretraining. SleepFM-2 also outperformed a 480-feature baseline derived from the same recordings. Its disease scores revealed a reproducible principal component associated with reduced sigma-band spatial coupling and increased hypnodensity entropy. The frozen encoder performed within the observed range of expert scorers for sleep events and transferred to wakeful EEG, headband and in-ear EEG, wrist PPG and wrist accelerometry. It improved sleep staging across six accelerometry cohorts and achieved disease-prediction performance in UK Biobank similar to models pretrained directly on accelerometry. Finally, SleepFM-2 captured aspects of subjective sleep not recovered by conventional PSG summaries, particularly reports of the recorded night. These results show that multimodal sleep physiology can provide a transferable representation of human health across diseases, clinical tasks, sensors and subjective experience.
Read the full abstract
Materials Science
A Novel Nb-Based Eutectic Superalloy with Exceptional Ultrahigh-Temperature Mechanical Properties
Refractory alloys operating above 1300 {\deg}C, beyond the limit of Ni-based superalloys, must balance high-temperature strength, thermal stability, low density, low cost, and room-temperature tensile ductility, a combination that existing refractory alloys have yet to achieve. Here we report a castable Nb-based eutectic superalloy, Nb-10Mo-9.5C (at. %, NMC-1), comprising a uniform lamellar structure of Nb-Mo solid solution and Nb2C carbide phases. NMC-1 achieves 0.2% yield strengths of ~300 MPa and ~200 MPa at 1300 {\deg}C and 1500 {\deg}C, respectively, among the highest reported for Nb-based alloys, with pronounced strain hardening, non-zero room-temperature tensile ductility, and no observed microstructural coarsening after 100 h at 1400 {\deg}C. Its low density (8.65 g/cc) and cost (~$85/kg) yield a specific-strength-per-cost merit index that far surpasses commercial Nb, Mo, Ta, and W alloys, establishing a new design paradigm for ultrahigh-temperature structural materials.
Read the full abstract
Functional Analysis
Sharp Summability on Supports of Prescribed Combinatorial Dimension
We solve four questions raised by Bayart concerning coefficient summability for multilinear forms with prescribed supports. For every $m\ge 2$ and $d\in[1,m]$, we determine the product-summability exponent and the multilinear summability invariant: \[ \mathrm{prod}(m,d) =\min\left\{\frac{m}{d},\,m-\lceil d\rceil+1\right\}, \qquad \gamma_{\mathrm{mult}}(m,d) =\min\left\{m-\lceil d\rceil+1,\frac{2m}{d+1}\right\}. \] In particular, $\gamma_{\mathrm{mult}}(4,2)=8/3$, showing that the multilinear invariant need not be an integer. We also prove that, for every $d\in[1,m]$, there is a single infinite support of exact combinatorial dimension $d$ on which the dimensional Hardy–Littlewood bound is attained over both scalar fields for every anisotropic parameter $\mathbf p=(p_1,\ldots,p_m)$ with $\sum_j 1/p_j<1$, simultaneously across the two regimes separated by $\sum_j 1/p_j=1/2$.
Read the full abstract
Numerical Analysis
Invariant-region-preserving high-order schemes for a model of reactive sedimentation
Reactive sedimentation of mixtures composed by biological and inert solid matter including liquid substrates is modelled by a system of convection-diffusion-reaction equations. In this process, the small solid particles dispersed in the viscous fluid settle due to gravity while simultaneous chemical reactions take place between solid and liquid components (substrates). Among the applications of reactive sedimentation processes, the principal use is in the simulation and control of secondary settling tanks (SSTs) in water resource recovery facilities (WRRFs). The governing equations include a nonlinear strongly degenerate parabolic equation for the total concentration of solids, and transport equations including reaction terms for the solid components and substrates. To accurately approximate the model equations, a high-order finite volume scheme is developed using polynomial reconstructions. Two reconstruction methods are used, a second-order monotonic upstream-centered scheme (MUSCL), and a third-order central weighted essentially non-oscillatory (CWENO3) method. In addition, for the time discretization, we employ strong stability preserving Runge-Kutta (SSPRK) schemes of second- and third-order. The CWENO3 scheme is complemented with limiters, and in the case of zero diffusion, the designed high-order schemes are shown to preserve an invariant-region property. This property is particularly relevant as it ensures that physically relevant numerical solutions are obtained at each time iteration. In addition, high-order approximations for the diffusion terms are proposed to approximate the complete model equations. Simulations of denitrification in SSTs under a continuously operated regime in WRRFs demonstrate the performance of the model and its discretization. Finally, convergence tests show the order of accuracy of the developed scheme and the invariant-region preservation.
Read the full abstract
Algebraic Geometry
Sharp Quadratic Bounds for Anticanonical Volumes of Weak Fano Surfaces
We prove an explicit lower bound of optimal quadratic order for the anticanonical volume of every complex $\epsilon$-log canonical weak del Pezzo surface, with no restriction on its Picard number or its singularities. More precisely, the volume is at least $\min\{c_{exc},\epsilon^2/3032064\}$, where $c_{exc}>0$ is an explicit absolute constant. Quotients of the projective plane show that the exponent two is optimal. We also prove cubic upper bounds for the cyclic determinant mass and for the Picard number of a smooth rational log Calabi–Yau pair. We give geometric examples showing that these two auxiliary exponents are optimal in their respective settings.
Read the full abstract
Algebraic Topology
Complexes of strong discrete Morse matchings
Using the strong discrete Morse theory developed by Fernández, we define the complex $\mathcal{SM}(K)$ of strong discrete Morse matchings on a simplicial complex $K$, as well as the pure subcomplex $\mathcal{SM}_{pure}(K)$ generated by the facets in $\mathcal{SM}(K)$ of maximal dimension. For most complexes $K$ these objects are proper subcomplexes of the complexes ${\mathfrak M}(K)$ and ${\mathfrak M}_{\textrm{pure}}(K)$ defined by Chari–Joswig using all Morse matchings on $K$. The homotopy types of the latter are not well-understood in general, but they are known when $K$ is the path $P_n$ with $n$ edges, the cycle $C_n$ with $n$ edges, the star $S_n$ with $n$ leaves, the $n$-simplex $\Delta^n$ ($n\le 3$), and the boundary $\partial\Delta^n$ ($n\le 3$). in this paper we compute the homotopy types of $\mathcal{SM}_{pure}(C_n)$ and $\mathcal{SM}(K)$ for $K=P_n,S_n,\Delta^n,\partial\Delta^n$ for all $n$. We also compute the homology of $\mathcal{SM}(C_n)$ for $n\le 21$.
Read the full abstract
Computer Vision
RAIDAL: Redundancy-Aware Information Density Active Learning for CTC-Based Continuous Sign Language Recognition
Continuous sign language recognition (CSLR) is a key technology for accessibility, yet its development remains limited by the high cost of annotating continuous video streams. Active learning offers a path toward mitigating this cost, but standard acquisition functions are not designed for weakly aligned sign language videos, where sign executions are interleaved with rest poses, irregular pauses, sign-like motion, and temporally redundant frames. This temporal redundancy can undermine sample selection, as acquisition scores may be influenced by timesteps from regions that are not associated with the decoded gloss sequence, distorting the video's estimated informativeness. In this work, we show that modern CSLR models already contain a mechanism for identifying gloss-level temporal evidence: the CTC decoder. Although typically used only during inference, its alignment peaks indicate where the model localizes each predicted gloss in the feature sequence, providing a source of temporal structure for active learning acquisition functions at zero additional labeling cost. Thus, we introduce RAIDAL (Redundancy-Aware Information Density Active Learning), which repurposes the CTC decoder to restrict representation-based scoring to decoder-aligned gloss regions, rather than exposing the acquisition function to the entire unfiltered video. Across three datasets and two architectures, RAIDAL achieves its strongest data-efficiency gains over competing baselines in large-vocabulary, budget-limited settings, while remaining competitive in the smaller-vocabulary, large-budget setting. The code used in this work is publicly available at github.com/verlab/RAIDAL.
Read the full abstract
Computation and Language
XYBench: Can LLMs Respond Pragmatically to Queries with Misconceptions?
When non-expert users ask LLMs for assistance, their queries can often have misconceptions (e.g., "How do I parse XML with regex?"). In such cases, often referred to as the XY-problem, LLMs must identify the misconception ("regex are fragile") and meaningfully direct the user toward a pragmatic solution that will address the root problem implicit in the request ("use an XML parser"). We introduce XYBench, a benchmark of 8,115 such queries, drawn from technical (StackOverflow/StackExchange) and everyday (WikiHow and a manually-curated subset) domains. We design an evaluation paradigm that assesses model responses along three criteria grounded in cooperative response theory: (a) presence and (b) emphasis on pragmatic solutions, and (c) identification of misconceptions. Our experiments show that even the strongest LLMs predominantly answer the literal request (0.75–0.92) and far less often the intended one (0.33–0.71), while substantially lagging behind humans at identifying misconceptions (at most 63% vs. 79–90%). Further, models overwhelmingly prefer pragmatic responses in a multiple choice setting yet consistently fail to generate them. Oracle ablation experiments show that providing explicit user intent at generation time helps; however a large gap remains, suggesting pragmatic redirection is a fundamentally underdeveloped capability in current LLMs.
Read the full abstract
Theoretical Economics
Mobile but Miserable: Social comparison networks and social mobility
Although social mobility is widely viewed as desirable, we document an 'inverse-U' shaped relationship with well-being: mobility is positively associated with well-being when low, but the association diminishes as mobility increases and eventually becomes negative. To explain this, we build a model in which agents are embedded in a social comparison network and suffer from falling behind their peers. The model highlights two channels that can cause welfare to be decreasing in mobility at high levels. First, when mobility is high, further increases widen already significant gaps between agents from similar parental backgrounds but of different abilities. Second, when mobility is high, high-ability agents are already relatively well off. So further increases in social mobility increase inequality. We find evidence supportive of both channels in US data. The model also predicts, and data supports, that social integration – where relationships form more along ability than background lines – allows societies to benefit from mobility for longer. We interpret this as social mobility being multidimensional with both economic and social dimensions that are complementary.
Read the full abstract
Cryptography and Security
WAPP: Safe Learning of Positive Security WAF Policies from Live Traffic
Web Application Firewalls (WAFs) mainly rely on signatures to detect known attacks, which can leave gaps against modified or previously unseen payloads. Positive security provides a complementary approach by learning legitimate traffic and blocking inputs that fall outside the learned profile. However, learning directly from live traffic can be unsafe when malicious requests contaminate the training data. This paper presents the Whitelisting Autonomous Policy Producer (WAPP), a framework that combines trust filtering, deterministic rule synthesis, confidence scoring, and validation before enforcement. WAPP is evaluated on three controlled applications using a live Coraza and OWASP Core Rule Set (CRS) stack. Results show that, on the tested DVWA username field, unfiltered learning becomes Degraded at 0.2% poisoned traffic and Broken at 0.5%, while the evaluated free text field can admit malicious inputs even without poisoning. On the frozen poisoning dataset, the ablation configuration with all seven candidate signals improves the measured poisoning resilience from 53% to 90%, compared with 62% for the Kruegel–Vigna baseline. The deterministic synthesizer provides attack blocking comparable to the tested language model without model inference cost. WAPP blocks confirmed CRS bypasses on constrained fields, while free text inputs remain a precision challenge that requires character level operator control.
Read the full abstract
High Energy Physics (Phenomenology)
Quark mass functions in Minkowski space
Using the Covariant Spectator Theory (CST), we calculate the dressed quark mass function and wave-function renormalization for the five quark flavors from up/down to bottom in both the spacelike and timelike regions of Minkowski space. The calculation employs a model dressed-gluon propagator fitted to lattice data in the spacelike region. The remaining free parameters of the quark self-energy are determined by fits to lattice results for the quark mass function including a constraint built into the wave-function renormalization. We perform these fits using lattice data from two different groups and find that the resulting mass predictions are within the range of the typical constituent masses used in quark models.
Read the full abstract
Computer Vision
A Shared-Backbone Approach for Multi-Task MedMNIST Classification
Multi-task biomedical classification requires models to generalize across disparate modalities and class distributions. We study 11 heterogeneous MedMNIST datasets using the harmonic mean of per-task macro-F1. We evaluate three backbones with task-specific linear heads. We identify a resolution domain shift between the MedMNIST API and evaluation environment. Resolving this inconsistency and optimizing architecture-specific regularization substantially improved performance. Our best configuration, a ConvNeXt-Tiny backbone with label smoothing, achieved a leaderboard harmonic-mean macro-F1 of 0.73294 in the Tensor Reloaded: Multi-Task MedMNIST competition, ranking sixth at the close of the official competition phase. Our implementation is publicly available at: https://github.com/GavrilStefan-Dorian/A-Shared-Backbone-Approach-for-Multi-Task-MedMNIST-Classification
Read the full abstract
Sound
BinauralVAE: Spatial Audio Reconstruction For World Models
Embodied artificial intelligence has historically very much relied on visual perception, leading to a proliferation of multiple vision-centric world models. However, this reliance fails to capture spatial understanding in its entirety and can even present vulnerabilities in environments with visual occlusions, low-light conditions, or blackouts-scenarios, where acoustic information becomes a critical alternative for spatial awareness and navigation. Despite its potential, research into realistic spatial audio and particularly the development of audio-centric world models remains sparse. In this technical report, we introduce BinauralVAE: a flexible, open-source pipeline (https://github.com/Luizerko/BinauralVAE) that explores multiple models for spatialized audio reconstruction, progressing from fundamental baselines to advanced, mathematically grounded architectures. Our approach evaluates various Variational Autoencoder architectures – including complex-valued variants – to learn robust latent representations of binaural signals. Developed alongside AudioWorldSim, our methodology leverages realistic acoustic data captured as a simulated robot navigates an environment. This pipeline establishes a foundation for state representation in a future audio-based world model, designed to map the direct causal connection between navigational actions and their resulting acoustic consequences, and helping to enable sound as an essential complementary modality for spatial knowledge acquisition.
Read the full abstract
Analysis of PDEs
Non-decaying weak solutions to the 2D quasi-geostrophic equations
We investigate weak solutions to the two-dimensional quasi-geostrophic equations without dissipation. We establish global existence of weak solutions for temperature bounded and lacking spatial decay and velocity in the space $L^2_{ul}(\mathbb{R}^2)$. Our methods rely on a spectral Serfati identity, which we use to establish uniform $L^2_{ul}$ bounds on a sequence of velocities satisfying the dissipative equations. These bounds, combined with a maximum principle on the scalar temperature, allow us to pass to the zero-dissipation limit, giving global-in-time weak solutions.
Read the full abstract
Cryptography and Security
Skynet: Workflow-Level Anomaly Detection for Agentic AI via Semantic and Structural Modeling
Agentic AI systems execute complex tasks through long-horizon workflows of planning, tool use, and multi-agent coordination. Task failures in these systems often originate from a single step, such as an injected prompt or a flawed plan, and are then amplified through downstream dependencies as the corrupted step propagates across many subsequent agents and tool calls. Existing defenses either target a specific class of attacks or failures, or inspect individual prompts and steps in isolation. Both leave the global dependency structure of a workflow unexamined, and miss the inconsistencies that only emerge when the execution is viewed as a whole. We argue that anomaly detection for agentic AI must reason at the workflow level, where global execution structure exposes signals that local checks cannot see. We present Skynet, a principled workflow-level anomaly detection framework that turns observed multi-agent execution into directed workflow graphs and scores them against learned benign behavior. Skynet jointly models the semantic execution context and the structural organization of inter-agent delegation, tool invocation, and data-flow dependencies, and trains only on benign workflows. Because training never sees attacks or failures, this design naturally extends to zero-day detection: any execution that violates benign workflow regularities surfaces as off-manifold geometry under a single decision rule. We evaluate Skynet on three public agentic safety and failure benchmarks. It sustains high recall together with a sub-1% false positive rate, with per-workflow and per-step latencies low enough for online monitoring of agentic AI runtimes.