General Economics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
We review the evidence on subsidized job training programs in industrial countries. We provide critical, narrative summaries of randomized studies of major job training in the US; of two "judge randomization" studies in the US and Denmark; and of previous meta-analyses. We then perform a meta-analysis of 56 randomized trials of job training in the US since 1973, the most exhaustive meta-analysis in this niche. Here, average impacts are positive but modest: employment rises 1.7 percentage points in years 3-5, and annual pre-tax earnings by roughly \$800 per person offered training. At an average cost of \$13,598 per participant, programs roughly break even under (debatable) assumptions about how long benefits persist. They recoup about three-quarters of their cost for government through higher tax revenue and lower benefit use. Evaluations of the largest federal efforts–JTPA, the Workforce Investment Act, and Job Corps–return small or null effects. One cluster of programs stands out: "sector programs," which screen applicants intensively, design curricula with employers, and track local skill demand. They raise earnings roughly ten times as much. They have proven difficult to replicate.
Computer Vision
Fine-Grained Visual Preprocessing and Dual-Stream Temporal Modeling for Multimodal Sentiment Analysis on Social Media
Multimodal sentiment analysis often remains text-dominant due to raw-video noise and insufficient temporal modeling. Using CH-SIMS v2.0S, this study proposes three improvements: the NAPS pipeline—a seven-stage system integrating face tracking,identity embedding, and normalized lip-motion analysis to reduce visual noise;DS-TANet, combining an EfficientNetB2 static stream, RAFT optical-flow motion stream, motion-guided attention, and Bi-GRU temporal modeling; and DS-TAFNet, fusing visual and MacBERT-Base textual representations via concatenation fusion. With NAPS, the static visual baseline achieves 80.98% Macro F1, comparable to the text baseline of 80.55%; DS-TANet improves visual Macro F1 to 82.58%;and DS-TAFNet achieves 87.49% accuracy and 87.48% Macro F1. These results demonstrate that improving visual input quality and temporal representation is more effective than increasing fusion complexity under limited-data conditions.
Read the full abstract
Machine Learning
NeuCME: Toward Dynamic Multimodal Continual Learning via Neural Combinatorics of Multiple Experts
Multimodal continual learning has recently shown great potential for developing agents with human-like intelligence by continuously learning new tasks across multiple modalities. However, existing methods typically assume that the set of modalities per task is predefined and fixed. In this paper, we investigate a more realistic learning setting, referred to as dynamic multimodal continual learning, in which the set of modalities may vary across tasks rather than remaining fixed. This setting involves two primary challenges: (i) spatio-temporal catastrophic forgetting and (ii) adaptive multimodal fusion. To address these challenges, we propose NeuCME (as shorthand for Neural Combinatorics of Multiple Experts), a novel framework designed to effectively learn and integrate knowledge across tasks with varying modalities. The proposed NeuCME model comprises three key components, namely modality-combinational rehearsal, multi-gated mixture-of-experts, and task relevance-guided distillation. Furthermore, we formulate an evaluation metric to quantify the dynamism of task sequences and then set up a comprehensive benchmark with different degrees of dynamism. Extensive experiments using four real-world datasets demonstrate that the proposed NeuCME outperforms state-of-the-art methods markedly.
Read the full abstract
Artificial Intelligence
RedKnot-MLA: Multi-Head Offline-Online Reuse for DeepSeek-V4 Long-Context Serving
Multi-head latent attention (MLA) exposes many logical query heads through one packed latent KV stream. This representation is memory efficient, but it removes the physical per-head cache boundary assumed by conventional head-wise reuse. We present our system, a DeepSeek-V4 realization of RedKnot's head-aware reuse principle. Each immutable document is processed offline at canonical position zero; certified Local-head contributions are retained as MLA-Off. At serving time, query-side RoPE relocation restores the document's request position, a small Global-head set and protected Local token rows are recomputed as MLA-Online, and the two paths are merged before a single shared output projection. The packed MLA latent is never split. DeepSeek-V4-Flash uses 37 reusable layers and a 56/8 Local/Global partition, giving a 75.29% analytic logical head-row ceiling; the Pro-0813 profile uses 55 layers and 112/16 heads, giving 78.89%. Frozen Flash operating points show hot-artifact TTFT speedups of 2.02-3.84x. At 256K, the archived three-dataset study reports an aggregate F1 change of +3.24 percentage points, an EM change of +4.16 points, and a 78.7-79.5% analytic major-operator arithmetic saving, while one dataset decreases by 2.81 F1 points. A separate author-reported 256K hot-artifact QPS measurement is approximately 2.0x; because its raw concurrency trace is not included in this bundle, we mark it as preliminary rather than archived evidence. We describe the factorization, position repair, token-row closure, sparse-MoE support, TP8 integration, and the measurement boundaries needed to interpret these results.
Read the full abstract
Differential Geometry
Bott connection in the mapping cone case
We generalize the concept of Bott connection to the de Rham mapping cone situation. As the main results, when the given closed smooth manifold is foliated, we have the mapping cone version of the Bott vanishing theorem and an adiabatic limit of a family of mapping cone connections on the orthogonal complement of the foliation. The second result shows the necessity of a refined metric if we want to see more interactions between geometry and topology in the mapping cone situation.
Read the full abstract
Computer Vision
GIFT: Goal-Injected Fine-Tuning for Efficient Manipulation Policy Adaptation
Compared with relying solely on initial observations and language instructions, predicting goal images with generative models as high-level visual guidance can significantly enhance the robustness of Vision-Language-Action (VLA) models. However, most existing foundation models have not systematically incorporated goal image conditioning due to the high computational training cost. To this end, we propose Goal-Injected Fine-Tuning (GIFT), a lightweight and efficient fine-tuning framework that seamlessly integrates generated goal images into multiple representative pretrained VLA models. Our approach introduces goal image features into observations via a zero-initialized convolution which progressively grows parameters from zero and prevents harmful noise from disrupting the pretrained policy during fine-tuning. As training proceeds, goal information is gradually incorporated, enabling efficient goal understanding without disrupting model stability. We further introduce a refined image editing method to generate semantically and visually consistent goal images from initial observations and task instructions. Experiments show that goal-aware VLA models achieve substantial performance gains across tasks: with only a single epoch of fine-tuning, GIFT outperforms the base model by 6.0% and 13.4% on two SIMPLER settings, and by 4.7% on LIBERO, demonstrating both efficiency and effectiveness.
Read the full abstract
Numerical Analysis
Limiter-based fully-discrete entropy stable explicit DG schemes for ideal MHD equations
We propose a class of high-order fully-discrete entropy stable (ES) explicit discontinuous Galerkin (DG) solvers for the compressible ideal magnetohydrodynamics (MHD) equations. Our main theoretical contribution is the introduction of a novel generalized-path-decomposition framework for MHD equations in Godunov's symmetric form. By innovatively interpreting the interior volume integral of the non-conservative source term as a path integral along a generalized path constructed by the solution polynomial, we establish the weak cell entropy inequality for the fully-discrete DG schemes. This overarching framework also accommodates other existing DG solvers based on the symmetric form. Combined with a carefully designed ES limiter, the proposed scheme satisfies the genuine fully-discrete cell entropy inequality. With this property, a Lax–Wendroff-type theorem can be obtained to show that the solution limit satisfies the entropy condition. Finally, the scheme is naturally compatible with the locally divergence-free space. Extensive numerical experiments demonstrate the scheme's low numerical dissipation and strong robustness.
Read the full abstract
Computer Vision
CGSM: Concept-Guided Segmentation Model for Precise Pulmonary Lesion Delineation
Accurate segmentation of pulmonary lesions is essential for effective clinical diagnosis and treatment strategies. Existing segmentation approaches often lack task-specific semantic guidance, as text-based annotations typically offer coarse localization of lesions, leading to inadequate delineation of lesion boundaries and poor performance on small-scale lesions. To address this, we propose CGSM, a Concept-Guided Segmentation Model that integrates LLM-generated and clinically reviewed concepts into the segmentation process. Specifically, we design a Concept-Visual Alignment Module (CVAM) to activate relevant tokens within the concepts that align with visual features, enhancing the interaction between textual and visual information. In addition, we introduce a Concept Modulated Decoder (CM-Decoder), which uses concepts from CVAM as modulation signals to facilitate the adaptive fusion of image and text features, improving the segmentation accuracy. Extensive experiments on two public datasets show that CGSM achieves state-of-the-art performance, with results of 91.59% Dice and 84.49% mIoU on the QaTa-COV19 dataset, demonstrating its effectiveness in pulmonary lesion segmentation.
Read the full abstract
Statistical Methodology
Optimizer as Detector: Stochastic Gradient Descent for Latent Mixture Models
Pooling latent subpopulations can obscure relationships and yield misleading regression conclusions, including Simpson's paradox (SP). We propose a detector based on the steady-state dynamics of constant-step stochastic gradient descent (SGD). Unlike likelihood-based mixture tests and confounder-search methods, it requires neither a normal-mixture specification nor observed candidate confounders. Two linear pieces compete under a winner-take-all squared-error loss, and their normalized terminal separation forms the test statistic. Using diffusion approximations, we derive its asymptotic null distribution for general centered scalar covariates and for Gaussian multivariate covariates under symmetric noise. The distribution-dependent null center reduces to the dimension-free constant $4/\pi$ when both covariates and noise are Gaussian. We establish asymptotic size control and consistency against fixed mixture alternatives under regularity conditions. We extend the method to intercept and partial-mixture heterogeneity and study endogeneity, heteroskedasticity, and nonlinear misspecification. Simulations examine calibration, power, and robustness. Finally, a three-stage Detect–Screen–Verify toolkit separates evidence of heterogeneity from its substantive explanation. Across eight public datasets, it recovers four established SP benchmarks and identifies four cases that, to our knowledge, have not been documented previously. The detector requires neither latent-group labels nor a prespecified number of mixture components.
Read the full abstract
Robotics
WM-Craftnet: World Synesthesia Model for Generalizable and Robust Dexterous In-Hand Manipulation
Generalizable and robust dexterous in-hand manipulation requires a policy to infer object pose, geometry, contact, and potential slip from partial and noisy observations. Although recent tactile and visuotactile RL methods achieve strong in-hand rotation in controlled settings, their robustness often degrades under pose shifts, force disturbances, and object variation. We propose WM-Craftnet, a world-model-conditioned framework that learns compact action-conditioned latent dynamics from proprioception, depth, tactile sensing, and actions, supervised by multimodal reconstruction and reward prediction. Rather than using the world model for latent imagination or policy optimization, WM-Craftnet uses the learned World Synesthesia Model (WSM) as recurrent task context for an asymmetric actor–critic policy. Importantly, WSM is trained to reconstruct clean depth targets from noisy depth inputs, providing a denoised geometric state for real-robot deployment. Ablations over recurrent baselines, auxiliary heads, tactile masking, and WSM modality heads show that predictive world modeling, clean-depth supervision, and tactile contact cues all shape the learned state. A WSM pretrained on nine \(z\)-axis objects serves as a reusable prior for \(49\)-object downstream policy learning. This context improves multi-object rotation, with quantitative and qualitative evidence for unseen-object, perturbation-recovery, and sim-to-real transfer.
Read the full abstract
Human-Computer Interaction
Adaptive Complementarity in Human-AI Systems: Architecture as a State-Shaping Choice
Human-AI interaction can improve current performance while changing the capabilities and relationships on which future performance depends. We develop adaptive complementarity, a framework for choosing interaction architecture with these state consequences in view. Access, information exposure, task allocation, timing, and communication can alter which arrangement will be valuable later; their settings can often be reset faster than the capabilities, search patterns, or conventions they create. Three mechanisms organize the argument: information exposure and collective search, delegation and capability evolution, and strategic interdependence and information governance. Their integration yields cross-mechanism implications, including conditions under which a loss of expertise heterogeneity increases the information differentiation required to preserve independent search. We distinguish strong human-AI complementarity from advantage over another workflow and from advantage over an evolving reference policy. A computational illustration examines scarce human review in a workflow whose success requires several specialized stages. Review develops human expertise and AI capabilities, changing where subsequent review is most valuable. Adaptive allocation improves net output over untailored procedures and the optimal predetermined calendar. An understandable priority rule derived from the adaptive solution retains essentially all of its gain: the procedure stays fixed while assignments respond to the capabilities that interaction creates. The framework directs evaluation toward the states present interaction creates, their consequences for later architectural fit, and the conditions under which observing and responding to them is worthwhile.
Read the full abstract
Functional Analysis
An Improved Lower Bound for the Complex Grothendieck Constant
We prove $K_G^{\mathbb{C}}>1.35584631827168$ for the classical complex Grothendieck constant, closing more than one quarter of the gap between Davie's lower bound and Haagerup's upper bound. The numerical part of the proof is rigorously verified by interval arithmetic. The lower bound and the proof were discovered by the Odin Automatic AI Research Agent.
Read the full abstract
Numerical Analysis
Uniform Inf-Sup Norm Equivalence and Robust Operator Preconditioning for Stokes Flow in tight domains with Periodic Pillars
Many microfluidic and porous-media computations reduce to the same core task: solving a Stokes saddle-point system on a domain perforated by a dense periodic array of pillars, as in deterministic lateral displacement (DLD) particle sorters. After rescaling the device to unit size, the geometry is controlled by a single dimensionless parameter $m$—the number of pillars across the device, equal to the inverse period. In realistic devices $m$ reaches the hundreds or thousands, and as it grows the Stokes inf-sup constant decays like $m^{-1}$, the pressure Schur complement becomes severely ill-conditioned, and standard block solvers slow down in proportion to the pillar density. We remove this bottleneck by identifying the pressure norm that the divergence operator induces on such geometries. For periodic pillar arrays in the proportional-hole regime, we prove that this inf-sup norm is uniformly equivalent to the $L^2+\sigma_\epsilon H^1$ $K$-functional norm at the pore scale $\sigma_\epsilon\asymp\epsilon$, with constants independent of the period, the pillar count $m$, and the mesh size $h$. The equivalence identifies the perforated Stokes problem with a Brinkman problem at a homogenized permeability, and its Riesz map reduces to a pressure-mass inverse plus a scaled stiffness inverse. Combined with operator preconditioning, this yields a block preconditioner built from standard algebraic-multigrid solves whose iteration count is essentially independent of both mesh size and pillar density. Two-dimensional Taylor–Hood experiments confirm the predicted robustness in mesh refinement, pillar density, geometric scale, close packing, and time step.
Read the full abstract
Statistical Mechanics
Relaxation of Sliding Friction from a Statistical Model of Aging Contacts
Frictional relaxation after a sudden velocity change underlies the phenomenological laws that describe the stability of steady sliding and of seismic faults. To connect this relaxation with measurable statistics of microscopic contacts, we introduce a minimal model incorporating logarithmic aging of contact forces and a power-law contact-size distribution reflecting fractal interface roughness, and derive the relaxation function analytically. Depending on the power-law exponent $\alpha$, the intermediate-time response exhibits a plateau, logarithmic decay, or power-law decay, in contrast to the exponential decay assumed in the phenomenological laws. The theory provides experimentally testable links between microscopic contact statistics and macroscopic frictional relaxation.
Read the full abstract
Quantum Algebra
Operator Michelson Contrast: Logarithmic Parametrisation, Subadditivity, and a Sharp Noncommutativity Threshold
We study the operator Michelson contrast Delta(A) = (kappa(A)-1)/(kappa(A)+1), with kappa(A)=||A|| ||A^{-1}||, for positive invertible operators, and introduce its logarithmic parametrization via arctanh(Delta(A)) = (1/2) ln kappa(A). This defines the Operator Logarithmic Contrast (OLC), extended to all invertible operators via polar decomposition. We prove OLC is subadditive under multiplication: OLC(AB) <= OLC(A)+OLC(B). For positive invertible A, we identify OLC(A) exactly as the projective Thompson distance from the identity ray to the ray of A. We establish a sharp dimensional threshold for equality in subadditivity: for dimensions 2 and 3, equality forces commutativity, while n=4 is the smallest dimension admitting noncommuting equality examples. Moreover, equality still forces commutativity in any total dimension if the operators factor as pure Kronecker products X = tensor X_i and Y = tensor Y_i with each factor dimension <= 3. We derive contraction inequalities, sum bounds, trace estimates for density operators, Lipschitz continuity (including singular limits), and limiting behavior for Cesaro means and infinite products. For density operators, we obtain an exact closed-form bijection between OLC and von Neumann entropy for qubits; prove this bijection does not exist for d >= 3; and establish upper/lower entropy envelopes at fixed condition number, showing the admissible range [S_min(kappa), S_max(kappa)] satisfies ln 2 <= S_max(kappa) < ln 3 for every kappa > 1. The qubit case recovers, as an operator lift, the classical Michelson-Jensen-Shannon equivalence of Bruni, Rossi, and Vitulano, while a counterexample shows this equivalence fails for d > 2 as a direct consequence of the non-bijectivity.
Read the full abstract
Signal Processing
Joint Communication Sensing Beamforming and Uplink Power Control for Network-Assisted Full-Duplex Cell-Free ISAC Systems
In this paper, we investigate the problem of designing joint downlink (DL) communication beamformer and uplink (UL) power as well as sensing beamformer in a multi static integrated sensing and communication (ISAC) enabled cell free massive multiple input multiple output (CF mMIMO) system. To this goal, we first consider two priority based beamforming strategies. In the communication prioritized sensing design, we first design the communication beamformers and then the sensing beamformer is projected onto the nullspace of the effective communication channels to suppress its additional interference to both the DL UEs and UL receivers. In the sensing prioritized communication design, we select the sensing beamformer first according to the target direction, while the communication beamformers are subsequently optimized using a max min SINR formulation. In addition to these priority based designs, we formulate a sensing centric joint optimization problem to determine the optimal DL communication beamformers, sensing beamformer, and UL UE transmit powers. Our objective is to maximize the sensing SINR while guaranteeing individual DL and UL SINR requirements and satisfying the per AP and per UE transmitpower constraints. Since the resulting problem is non convex because of the coupled beamforming and UL power variables, the beamforming vectors are lifted into positive semidefinite transmit covariance matrices and semidefinite relaxation (SDR) is applied.
Read the full abstract
High Energy Physics (Theory)
Dual N=1 Lagrangians for Argyres-Douglas theories
We find a new set of four-dimensional $\mathcal{N}=1$ Lagrangian gauge theories that flow in the infrared to $\mathcal{N}=2$ Argyres-Douglas theories of type $(A_1, A_{2n})$, $(A_1, A_{2n+1})$, $(A_1, D_{2n+1})$, and $(A_1, D_{2n+2})$. The new dual theories are given by $SU(2)^n$ quiver gauge theories with a number of gauge-singlet flip fields. This description provides dual theories for the previous $\mathcal{N}=1$ Lagrangian realizations given by $Sp(n)$ or $SU(n+1)$ gauge theories, exchanging the rank and the number of nodes in a quiver diagram.
Read the full abstract
Group Theory
The Left-Nilpotent Residual and Its Supplements in Finite Skew Braces
Let $X$ be a finite skew brace, and let $L_\infty(X)$ be the final term of its left series. We introduce the left-nilpotent residual $\RL(X)$ and prove that $ \RL(X)=\Id_X\bigl(L_\infty(X)\bigr). $ Thus $X/\RL(X)$ is the largest left-nilpotent quotient of $X$. For skew braces of nilpotent type, we relate $\RL(X)$ to $\gamma_\infty(X,\cdot)$ and obtain conditions ensuring that $L_\infty(X)$ is an ideal, including the case where $|X|$ is cube-free. Our main results concern supplements to $\RL(X)$. We prove existence results using coprime action and Sylow and Hall theory, and give an example of order $18$ whose residual has no proper supplement. Finally, iterating $\RL$ yields the largest perfect subskew brace of $X$, and $X$ is KSV-solvable if and only if this subskew brace is zero.
Read the full abstract
Software Engineering
LatentMD: Benchmarking Markdown Boundary Failures in LLM-Generated Text
Large language models (LLMs) increasingly generate Markdown that is consumed by renderers, agents, code extractors, and structured downstream pipelines. Yet existing evaluations often conflate content quality with format adherence, leaving Markdown boundary failures under-measured. We introduce LatentMD, a benchmark and evaluation protocol for diagnosing CommonMark-level fence-boundary failures in LLM-generated Markdown. LatentMD separates content correctness from boundary correctness, enabling detection of outputs that are content-correct but boundary-broken. The benchmark contains 4,179 prompts and a CLI for scoring arbitrary model outputs. Across 9 LLMs and roughly 37,600 generations, we find that Markdown boundary failures are widespread: 38.0% of valid main-grid outputs are content-correct but boundary-broken, with substantial boundary breakage under unspecified prompts and in a small human-authored validation set. Ablations show that failures are driven primarily by same-family symmetric-delimiter collisions rather than nesting alone, are only partially mitigated by prompt hints, and generalize to Python triple-quote docstrings while JSON remains robust as an asymmetric-delimiter control. LatentMD provides a reproducible diagnostic target for parser-sensitive LLM evaluation.
Read the full abstract
Fluid Dynamics
A Hybrid POD-Autoencoder Framework for Reduced Order Modeling of Turbulent Flow via Strategic Field Decomposition
This study proposes a hybrid reduced-order modeling (ROM) framework for the simulation of turbulent flow. The central idea is to decompose flow dynamics according to their temporal characteristics and predict the resulting components individually. The full field is first divided into a sub-field represented by a limited number of proper orthogonal decomposition (POD) modes (named as POD-retained field) and the corresponding residual sub-field (named as POD-truncated field). A frequency-informed POD strategy identifies the retained modes by considering both modal energy and dominant frequency. The evolution of retained POD coefficients, which feature similar temporal scales, is described using a vector autoregressive (VAR) model. In parallel, the POD-truncated field is compressed into a low-dimensional latent space using a Fourier-neural-operator-based Koopman $\beta$-variational autoencoder (FK-$\beta$-VAE), with the latent variables subsequently predicted by a switching-VAR model. Turbulent statistics of the full field are recovered by combining the contributions from the two components. The framework is assessed using turbulent channel flow at a friction Reynolds number of $110$. The predicted Reynolds-stress components, turbulent kinetic energy (TKE), and dominant wavenumber spectra show good agreement with the reference. Moreover, in comparison with an alternative framework of full-field modeling (i.e., without field decomposition), the proposed framework yields more accurate and robust long-term statistical predictions.
Read the full abstract
Computer Vision
AV-SafetyBench: A Safety Benchmark for Text-to-Audio-Video Generation
Recent text-to-audio-video (T2AV) models jointly generate video, speech, sound effects, and ambience from a single text prompt. This capability poses new challenges for safety evaluation, as unsafe content may be conveyed through the audio track or arise only when the visual and audio tracks are interpreted jointly. Existing safety benchmarks largely focus on either generated video or generated audio in isolation and are therefore not designed to capture these risks. To close this gap, we introduce AV-SafetyBench, the first safety benchmark developed specifically for T2AV generation. AV-SafetyBench comprises a four-axis, 13-category taxonomy and 5,200 manually reviewed prompts that specify visual scenes, speech, and non-speech audio. Our evaluation protocol assesses each output under three views: Full-AV, Video-Only, and Audio-Only. It then uses the Video-Only and Audio-Only judgments to assign Full-AV unsafe outputs to one of four risk sources: Video-Only, Audio-Only, AV-Both, or AV-Joint. We evaluate five open-source T2AV models and validate the automated Full-AV judgments against human annotations. Across the five models, Full-AV Unsafe Rates range from 25.1% to 49.4%. Beyond these aggregate rates, risk-source analysis reveals that, for four of the five models, Audio-Only and AV-Joint cases-unsafe outputs missed by video-only evaluation-account for 41.6-48.3% of Full-AV unsafe outputs for which a risk source could be assigned. In the Cross-Modal Harm Emergence category, AV-Joint accounts for 87.5% of unsafe outputs withan assigned risk source. Together, these findings demonstrate the value of AV-SafetyBench for evaluating T2AV safety across the visual and audio modalities and their interaction.
Read the full abstract
General Topology
An overlooked weakening of perfect normality: Perfect regularity in spaces and locales
We introduce the notion of perfect regularity as an appropriate weakening of perfect normality, both for spaces and locales. Various characterizations are given, using Dedekind-MacNeille completions, injective hulls, and sublocales. We place the new class of perfectly regular frames among various well-studied classes of frames. We also introduce the construction of perfect regularization of a completely regular frame, compare it to Isbell's well-known booleanization construction, and argue that it is at least as important as the latter.
Read the full abstract
Signal Processing
Lightweight Generative Image Semantic Communication over Packet Erasure Channels
This paper addresses packet loss in semantic communication caused by network congestion or channel fluctuations. We propose LGSemCom, a lightweight generative packet-level joint source-channel coding (JSCC) framework for efficient and robust image transmission over packet erasure channels. Unlike conventional distortion-oriented recovery methods that yield blurry averages over erased regions, LGSemCom reformulates image recovery under packet erasures as a conditional generative task. By leveraging adversarial learning, the proposed decoder synthesizes plausible details without increasing complexity during inference. A key innovation of our framework is an erasure-aware weighting (EAW) strategy, which prioritizes generation in erased regions while preserving pixel fidelity in correctly received areas. To ensure computational efficiency, LGSemCom employs a fully convolutional codec based on efficient long-range attention blocks (ELABs) that capture global semantic dependencies with low complexity. Extensive experiments show that LGSemCom achieves superior perceptual reconstruction quality compared with existing benchmarks under severe packet loss, while achieving an order of magnitude faster inference. These attributes make LGSemCom highly suitable for latency-sensitive applications on resource-constrained edge devices.
Read the full abstract
Statistics Theory
Almost Sharp Equivalence between Approximate Message Passing and Low-Degree Polynomials
We prove a sharp lower bound for growing-degree polynomial estimation in the symmetric rank-one matrix estimation problem. The observation is \[ \boldsymbol{Y}=\frac{1}{\sqrt{n}} \boldsymbol{\theta} \boldsymbol{\theta}^{\top}+\boldsymbol{W}, \] where the coordinates of $\boldsymbol{\theta}$ are independent with a common, fixed sub-Gaussian distribution $\pi_\Theta$ satisfying $\mathbb E[\Theta]\neq0$, and $\boldsymbol{W}$ is symmetric with independent standard Gaussian upper-triangular entries. For every such prior, we give an explicit finite-dimensional bound implying that every sequence of polynomial estimators of degree $D(n)=o(n^{1/60})$ has normalized mean-square error with limit inferior at least $\mathbb E[\Theta^2]-q_{\mathsf{amp}}$, the limiting error of Bayes approximate message passing (AMP). This extends the constant-degree equivalence of Montanari and Wein~\cite{montanari2025equivalence} to growing degree under the same prior assumptions. Combined with their polynomial approximation of fixed-iteration AMP, the bound identifies the exact limiting low-degree MMSE whenever $D(n)\to\infty$ within this range. It thereby answers the growing-degree AMP-equivalence question discussed in~\cite{wein2025computational, maleki2026high} for nonzero-mean sub-Gaussian priors in the stated degree range. We also obtain a moment-dependent finite-dimensional lower bound for any nonzero-mean prior with all moments finite. The proof bounds low-degree MMSE using conditional joint cumulants and a refined application of Jensen's inequality. The conditional cumulants are defined using an auxiliary Gaussian channel $\boldsymbol{R}$ calibrated to the AMP fixed point. This retains signal dependence that is lost in unconditional cumulant bounds and produces the cancellations needed for quantitative control as the degree grows.
Read the full abstract
Mathematical Physics
A Note on Extended States on the Bethe Lattice
We give a new proof of the existence of absolutely continuous spectrum for the weakly disordered Anderson model on the Bethe lattice. The argument follows a general cyclicity criterion for Anderson-type Hamiltonians and reduces the problem to showing that the spectral measures of two independent copies of the rooted tree are not mutually singular. This is detected by their Hellinger affinity, and the required weak-disorder estimate follows from a short harmonic and compactness argument. We also formulate an analogous quantitative cyclicity criterion for the Anderson model on $\mathbb{Z}^d$, expressed through the Schur complement of the Poisson transform of a $2\times2$ matrix spectral measure.
Read the full abstract
Machine Learning
Continual Learning Mechanisms Compose for Long-Horizon Memorization
Language models may need to internalize information that arrives over time and retain it through many subsequent updates. To study this challenge, we introduce long-horizon memorization, a setting in which a model learns 100 query-answer tasks through continual supervised fine-tuning without retaining earlier training examples or receiving task identifiers at inference. Sequential updates cause catastrophic forgetting, and no single continual learning mechanism we evaluate maintains strong retention at this horizon. We hypothesize that mechanisms addressing complementary sources of forgetting will be more effective when composed. We organize these compositions along two design dimensions. Data, function, and weight anchors specify what prior information each update should preserve, while low-rank allocation rules determine where successive updates are retained. To test this hypothesis systematically, we construct three distinct 100-task memorization datasets. We introduce task-level successive halving to search the combinatorial design space and use a factorial experiment to measure individual and interaction effects. Our best method combines all three anchors with merged LoRA, ranks among the top 3 methods in all datasets, and raises average final retention from 1.2% under naive sequential fine-tuning to 34.9%, a 28-fold improvement. The data anchor and merged LoRA provide the largest average gains and interact super-additively on all three datasets. Together, these results show that composing complementary mechanisms substantially improves long-horizon memorization beyond what any individual mechanism achieves.
Read the full abstract
High Energy Astrophysical Phenomena
NEXUS: Transient Searches and First Results from Year One Observations
We describe ongoing efforts of high-redshift transient searches using multi-epoch NIRCam imaging (F200W+F444W) and NIRSpec MSA/PRISM spectroscopy from the NEXUS JWST multi-cycle Treasury program targeting the north ecliptic pole region. The transient search area covers $\sim 61.5\,{\rm arcmin^2}$ between the reference epoch and each subsequent NEXUS-Deep epoch at a cadence of $\sim2$~months. In the first year of observations from NEXUS, we detect 68 robust transients, with the host photometric redshift distribution declining rapidly at $z>2$ but extending to $z_{\rm phot}\approx 6$. In addition, we obtained secure spectroscopic redshifts for 37 transients ($\sim54\%$) from NIRSpec/PRISM and NIRCam/WFSS, with seventeen at $1 < z < 2$, eight at $2 < z < 3$, two at $3 < z < 4$, and one tentative host association at z = 6.151. Overall, the NEXUS program recovers observed supernova (SN) rates broadly consistent with other SN search programs with JWST. While NEXUS achieves the highest transient detection efficiency, 6.6 SNe per imaging hr ($2.6\times$ COSMOS-SN and $26\times$ JADES-SN), the limited filter coverage (F200W+F444W only) limits robust identification and classification of high-$z$ SNe for deep spectroscopic follow-up. We describe the details of the data reduction and transient detection pipeline, and report the Year-1 transient sample along with their light curves, available MSA spectroscopy, host association, and the raw detection rate. We also describe and release a new PSF photometry package that properly accounts for correlated pixel noise from combining drizzled images.
Read the full abstract
Machine Learning
AF-Mamba: Efficient Long-Term Signal Modeling for Early Prediction of Atrial Fibrillation Onset
Atrial fibrillation (AF) is the most common cardiac arrhythmia and is associated with increased risks of stroke and heart failure. The growing availability of wearable and portable ECG monitoring enables continuous assessment of cardiac rhythm outside clinical settings. Predicting AF before its onset could provide additional lead time for timely clinical assessment and potentially improve the management of patients at risk of AF-related complications. This study focuses on predicting AF onset one hour in advance using long-term RR intervals (RRIs). To address this challenge, we propose a deep learning architecture that integrates temporal convolutional networks (TCNs) for local features encoding with Mamba, a selective state-space model capable of long-range sequence modeling. This hybrid TCN-Mamba design enables efficient training and inference on one-hour input windows, overcoming limitations of Transformers' quadratic scaling and recurrent networks' vanishing gradients. In subject-wise 5-fold testing, the proposed model achieved a sensitivity of 0.889, specificity of 0.943, F1-score of 0.813, AUROC of 0.974, and AUPRC of 0.933. In paired cross-dataset holdout evaluation, AF-Mamba maintained discriminative performance across unseen AF and NSR datasets, achieving a mean AUROC of 0.897. Compared against state-of-the-art AF prediction models and general time-series models, AF-Mamba achieved competitive predictive performance while providing a favorable performance-efficiency trade-off for long RRI sequences. These findings demonstrate the potential of AF-Mamba for accurate AF prediction one hour in advance and real-time continuous ambulatory monitoring.
Read the full abstract
Optimization and Control
Input-to-State Stability Framework for Fully Distributed Primal-Dual Dynamics for Quadratic GNEPs Without Multiplier Consensus
Generalized Nash Equilibrium Problems (GNEPs) often arise in multi-agent engineering applications that require distributed algorithms. Unlike traditional approaches that enforce consensus on multipliers, our method removes the need to share multipliers, reducing communication and improving privacy. As a result, different initializations can lead to different GNEs, including non-variational ones. We establish convergence under sufficient conditions using an input-to-state stability (ISS) framework.
Read the full abstract
Distributed Computing
TreeRedux: Separating Concerns in Spark's Distributed Tree Aggregation
By default, Apache Spark's tree aggregation primitives place the tree root on the driver, requiring the driver to participate in the same aggregation computation over intermediate aggregation state as executor nodes. For large aggregates, this can expose the single coordinator to substantial computation and memory requirements. Recent Spark versions optionally move the root to an executor, but the completed aggregate must still be returned to and materialized on the driver. We demonstrate this limitation using exact quantile computation and heavy-hitter identification, where the intermediate aggregation state can be substantially larger than the desired final result. We propose TreeRedux, a minimal extension that adds a terminal finalize operation, executed on an executor, that maps the aggregation state U to a compact result V, so that V rather than the potentially large U is materialized on the driver. For exact quantile computation, applying TreeRedux to GK Select removes the driver's epsilon-n memory term, reducing driver memory requirements to the same asymptotic order as Spark's GK Sketch. In our experiments, the default GK Select implementation encountered a driver out-of-memory error at 2.5 billion elements. Spark's executor-side final aggregation option extended this limit to approximately 16-18 billion elements but still required the final aggregation state to be materialized on the driver. Redux Select completed through 28 billion elements without a driver out-of-memory error. TreeRedux allowed Space-Saving sketches with up to 32x the capacity of the largest configuration that materializes a full sketch on the driver.