arXiv:2601.20450v2 Announce Type: replace
Abstract: We study a production network in which firms use non-substitutable (Leontief) inputs, hold precautionary inventories and face idiosyncratic productivity shocks, with adjustment occurring through quantities rather than prices. We show analytically and numerically that a critical boundary exists in the space of shock volatility and inventory holdings: above this threshold, the economy absorbs shocks and fluctuates mildly while below it, cascading shortages make system-wide crises inevitable. Close to the threshold, aggregate output volatility diverges through network-mediated amplification of purely idiosyncratic shocks, providing a concrete mechanism for the ``small shocks, large business cycles'' puzzle. Because inventories are costly, competitive pressure drives firms toward the fragility boundary: a resilience-efficiency trade-off emerges, putting the gains from lean supply chains at risk. Finally, we show that supplier diversification shifts the threshold and, depending on its abundance, can eliminate the fragile regime entirely.
Science Journals
arXiv:2607.07846v1 Announce Type: new
Abstract: VectorizationLLM is a specialized Large Language Model based on Google open-weight LLMs. The model is designed to assist students to learn smart vectorization, time/wave vector analysis, piecewise functions, Fourier analysis, and differential equations in MATLAB. The course application is CTEC 247: Applied Computational Analysis II by the Department of Electrical & Computer Engineering Technology at New York Institute of Technology Old Westbury. The LLM model is designed to be an instructive assistant, providing detailed explanations of concepts with examples from in-class notes without providing direct answers to questions. The model is designed with a RAG (Retrieval Augmented Generation) knowledge base and system prompt architecture. Examples in both code, text, and images are provided in the LLM responses.
arXiv:2607.07979v1 Announce Type: new
Abstract: Plasma accelerators sustain gradients orders of magnitude higher than conventional radiofrequency machines, but most proposed paths to TeV energies still require tens of stages, each demanding sub-micrometer alignment, femtosecond synchronization, and precise matching of the accelerating trailing bunch. Here we introduce plasma wakefield acceleration via regenerative cascading, in which each stage self-injects a fresh trailing electron bunch and the accelerated trailing bunch becomes the driver for the next stage. This approach has several advantages: energy multiplication instead of addition; automatic alignment, synchronization, and matching of the trailing bunch to the wake; and trailing bunch brightness reset in each stage. Particle-in-cell simulations show the generation of a 1.1 TeV electron beam with ~0.3% rms energy spread and 0.12 nC charge from a two-stage, sub-kilometer plasma accelerator driven by a 45 GeV, 100 nC beam. The low energy spread is achieved via dynamic beam loading in the evolving wake of the post-depletion driver that acts as a built-in energy dechirper.
arXiv:2607.07995v1 Announce Type: new
Abstract: Multi-robot localization that is accurate and consistent is imperative for downstream tasks such as planning and control. Centralized filtering approaches optimally fuse all available sensor measurements of the team. However, a centralized solution is rarely implementable due to hardware, communication, and computational constraints. Distributed approaches deploy a filter on each robot to estimate their own state and neighbours' states using inter-robot communication. This paper proposes a consistent, communication-efficient, and consensus-based distributed filtering framework that shares both preintegrated odometry and relevant shared states among communicating robots. The proposed method is validated in simulated and experimental scenarios, showing near centralized performance in accuracy, and especially in consistency, compared to the current state-of-the-art decentralized approach.
arXiv:2607.08116v1 Announce Type: new
Abstract: Inference-time scaling has emerged as an effective approach for enhancing the capabilities of Large Language Models (LLMs), addressing the growing demand for stronger reasoning without increasing model size. This novel form of LLM scaling comprises two representative approaches: explicit reasoning, which generates intermediate chain-of-thought tokens during an explicit thinking phase, and implicit reasoning, which iteratively updates hidden states in the latent space without producing explicit outputs. Despite their effectiveness, both paradigms incur substantial computational and memory overhead, raising challenges for deployment on resource-constrained edge devices. To address these issues, we propose a Mobile Reasoning-as-aService (MORES) framework that treats reasoning as a computational service accessible to edge devices over wireless networks. Focusing on implicit reasoning, we leverage its recursive structure to partition hiddenstate updates between edge devices and servers, enabling cooperative inference that allows devices to access additional cloud computation on demand. To optimize long-term performance, we formulate a joint computation and communication scheduling problem and solve it using a semantic Mixture-of-Experts (MoE)-based Deep Reinforcement Learning (DRL) algorithm to address heterogeneity in wireless conditions and task demands. The agent adaptively allocates resources by adjusting the number of recurrent steps and the transmission pruning rate, while a semantic router enables high-speed gating for real-time expert selection. Experimental results show that the proposed method achieves an approximately 18% improvement in system throughput over the baseline Soft Actor-Critic (SAC) algorithm. Our code is available at https://github.com/NICE-HKU/MORES.
arXiv:2607.08144v1 Announce Type: new
Abstract: Through-the-wall radar (TWR) human activity recognition (HAR) is important for non-line-of-sight indoor sensing, security monitoring, and emergency rescue. However, structured distribution shifts caused by person variation, observation-view variation, and wall-condition variation severely degrade recognition generalization, while the origin of the target-domain error still lacks a rigorous theoretical explanation. To address this issue, a generalization-analysis framework for TWR HAR is proposed in this paper. First, models for indoor human kinematics, TWR echo generation, radar image formation, feature representation, and bounded-weight neural networks are established within a unified source-to-target learning formulation. Then, the source risk, target risk, empirical risk, and admissible physical domain descriptor are defined, and a unified target-domain generalization bound is derived. Next, the structured shift term is decomposed into cross-person, cross-view, and cross-wall components, and the bound-tightening effects of physical low-dimensional representations, multi-source training, and parameter-space coverage are analyzed. Simulated and measured experiments jointly support the resulting theoretical analysis and illustrate its application value.
arXiv:2607.08151v1 Announce Type: new
Abstract: Optimal selections in a decision process are often constrained by prerequisites. However, such prerequisites can encode functional rather than literal dependencies, so a required dependency may be supplied by one or several interacting alternatives. We introduce matroidal prerequisite systems (MPS), a constraint structure where a poset specifies prerequisites while a matroid determines when those prerequisites have been satisfied by its span. This creates an order-sensitive notion of feasibility over words, where feasible words are associated with independent sets, while dependencies may be fulfilled through substitutable functionality.
Our main contribution is approximation algorithms for additive maximization and submodular maximization over the feasible words of an MPS. The guarantees are determined by two structural parameters: the maximum matroid rank $\Delta$ of a principal ideal in the poset and the maximum matroid connectivity $\lambda_\mathrm{max}$. These measure the distance an MPS is from encoding a matroid or a poset antimatroid, respectively, both of which are generalized by an MPS. For additive maximization, we obtain efficient deterministic $\Delta$- and $(1+\lambda_\mathrm{max})$-approximation algorithms. By extending these techniques, we obtain efficient deterministic $(2+\lambda_\mathrm{max})$-approximation and randomized $(\Delta^2\cdot(1 - 1/e - \delta)^{-1})$-approximation algorithms for all $\delta >0$ for submodular maximization. The algorithm design and analysis use the theory of polymatroid greedoids, via cryptomorphism we prove between an MPS and a strong polymatroid greedoid. Finally, an approximation-preserving reduction from densest $k$-subgraph shows it is not possible to efficiently compute a $\min\{\Delta,\lambda_\mathrm{max}\}^{o(1)}$-approximation to additive maximization over the feasible words of an MPS under the Gap Exponential Time Hypothesis.
arXiv:2607.08198v1 Announce Type: new
Abstract: This paper studies the problem of learning a joint distribution from marginal observations, which is inherently ill-posed due to the ambiguity of feasible couplings. We propose LUD-MSR, a latent-variable probabilistic framework that models the joint distribution via auxiliary representations and optimizes evidence lower bounds using only marginal data. Under mild assumptions, we establish an upper bound on the distribution approximation error. This analysis reveals a trade-off in representation learning between domain consistency and information preservation. To address this trade-off, we introduce a Multi-Scale image Representation (MSR) mapping that exploits structural similarity at coarse scales while suppressing domain-specific variations. We show that MSR achieves a more favorable balance of this trade-off compared to existing approaches. Experiments on real-world denoising benchmarks, including cryo-electron microscopy (cryo-EM), demonstrate the effectiveness of the proposed framework.
arXiv:2607.08204v1 Announce Type: new
Abstract: For continuous-time linear quadratic regulation with unknown system matrices, data-driven off-policy policy iteration typically estimates the value matrix and the improved feedback gain through a joint critic--actor regression. We show that the critic is not needed in the policy-improvement step. The key is to anchor the Riccati equation at a known stabilizing gain and express optimality as a policy-space residual. An endpoint null-space projection then removes the value-matrix term from the integral data equation. This yields a critic-free, actor-only least-squares update computed directly from input-state data. Under a verifiable projected rank condition, the resulting data equation is equivalent to the policy-space residual equation, and each update coincides with the Kleinman iteration. Thus, the stabilizing and convergence properties of Kleinman iteration are retained without a critic regression. We further show that the conventional off-policy full-rank condition decomposes into an endpoint critic rank condition and a projected actor rank condition. The proposed method removes the rank requirement needed for critic identification while retaining the one needed for policy improvement. The repeated least-squares dimension is reduced from $n(n+1)/2+mn$ to $mn$. Finally, comparative simulations validate the effectiveness of the proposed algorithm.
arXiv:2607.08037v1 Announce Type: cross
Abstract: Quantum learning provides a versatile paradigm for information processing by exploiting the intrinsic representational capacity of high-dimensional Hilbert spaces. Here, we investigate a Hamiltonian-encoding framework for quantum reservoir computing that simultaneously addresses three key challenges in quantum learning: trainability, hardware efficiency, and information stability. In this framework, input data are directly mapped onto a fixed Hamiltonian and transformed into expressive nonlinear features through quantum dynamical evolution. By employing the reservoir-computing paradigm, the approach naturally circumvents the barren plateau problem in quantum learning landscapes. We validate the framework across two complementary platforms: an analog superconducting array processor and a digital gate-based quantum circuit implementation. Despite their fundamentally different realizations, both platforms exhibit comparable representational power and achieve competitive learning performance, establishing a unified framework for cross-platform quantum learning. While both implementations achieve comparable performance, the analog processor may offer a more hardware-efficient realization by bypassing the temporal overhead of gate-based decomposition and thereby making more effective use of finite coherence times, albeit at the expense of universality. Furthermore, we find that finite dissipation suppresses quantum-scrambling-induced instabilities at long evolution times and can enhance learning performance, revealing a constructive role for environmental coupling in stabilizing quantum learning dynamics. Collectively, these results establish Hamiltonian-encoded reservoir computing as a compact, expressive, and hardware-efficient paradigm for quantum learning on current-generation quantum platforms.
arXiv:2607.07905v1 Announce Type: new
Abstract: Tree growth determines how much CO2 is sequestered from the atmosphere and temporarily stored in woody biomass. At the same time tree growth is affected by increasing temperatures, more frequent drought periods, late frosts and other extreme events associated with climate change. While continuous measurements of radial (secondary) tree growth using dendrometers are well established, monitoring of shoot elongation (primary growth) has largely been neglected because suitable measurement techniques are lacking. As a result, the effects of climate change on primary tree growth remain insufficiently understood. This work aims at reconstructing native deciduous trees in 3D as a basis for measuring and monitoring shoot elongation over entire tree canopies. Here we explored the use of low-cost UAV photogrammetry and of a multi-camera CraneCam system under real-world conditions. Data were collected in two study areas over an entire growing season. We present sensor evaluations, photogrammetric data acquisition and processing strategies. A special focus is placed on the analysis of the resulting photogrammetric 3D point clouds in terms of accuracy, resolution and completeness. Results demonstrate 3D point accuracies of 5-6 mm for entire trees using consumer-grade UAVs weighing less than 250 g and a 3D reconstruction completeness between 92% and 98% depending on the UAV type. The paper introduces a novel 3Dprinted ground-truth branch to evaluate the capability to reconstructing fine-detail structures such as thin tree shoots. Finally, we discuss operational challenges and initial experiments towards a skeletonization of entire trees based on photogrammetric point clouds.
arXiv:2607.08237v1 Announce Type: new
Abstract: In this paper, we propose and analyze a Continuous Interior Penalty (CIP) stabilized Virtual Element Method (VEM) for three-dimensional advection-diffusion-reaction equations on general polyhedral meshes. While CIP-VEM schemes have been recently explored in a two-dimensional setting, their analysis heavily relies on global mesh quasi-uniformity and constant physical parameters. We overcome these limitations by introducing a novel three-dimensional variant of the Oswald-type quasi-interpolant. This allows us to establish robust, uniform error estimates in the hyperbolic limit under a realistic local quasi-uniformity assumption and variable coefficients. Finally, we provide a comprehensive set of three-dimensional numerical experiments to validate the theoretical convergence rates and demonstrate the absence of non-physical oscillations.
arXiv:2607.08243v1 Announce Type: new
Abstract: The k-means++ algorithm is commonly restarted multiple times to avoid poor local optima, yet the number of restarts is almost always chosen arbitrarily and applied uniformly regardless of data set difficulty. This undermines any comparison relying on such a choice and wastes computation on easy data sets while potentially under-serving hard ones. We introduce GTRC, a restart criterion combining a Good-Turing estimate, a proven unconditional bound, and a confidence-based bound on the probability that a further restart would improve on the current result, stopping once this probability falls below a user-specified tolerance $\varepsilon$. Across 36 data sets, GTRC reached clustering quality competitive with well-chosen fixed restart counts, while the number of restarts used varied considerably and appropriately with data set difficulty, governed by an interpretable, data-dependent signal rather than a fixed rule. GTRC offers a principled and reportable alternative to fixing the number of $k$-means++ restarts in advance. Software:https://github.com/RCdeAmorim/Good-Turing-Restart-Criterion.
arXiv:2607.08231v1 Announce Type: new
Abstract: Smart homes have emerged as an important domain for HCI research, including work on usable security and privacy. Ideally, studies in these areas draw on datasets collected in real homes with real residents, capturing authentic device interactions, network traffic, and daily routines. However, creating such datasets is slow, expensive, and raises significant privacy concerns, as it requires long-term observation of people in their most private spaces. We propose using LLMs to generate diverse resident personas that interact with a simulated smart home, producing behaviorally grounded interaction schedules that can be executed on physical testbeds. We present (1) a design framework configuring simulated households across five socio-technical dimensions, (2) a multi-stage LLM pipeline that produces structured, executable device interaction schedules, and (3) a proof of concept demonstrating feasibility. As a work in progress, we aim to support scalable, privacy-conscious smart-home experimentation without relying on intrusive real-world data collection.
arXiv:2606.29031v2 Announce Type: replace
Abstract: In regulated domains such as banking and healthcare, where privacy constraints make real speech costly to collect and retain, synthetic speech from modern text-to-speech (TTS) is an appealing alternative for training automatic speech recognition (ASR) without exposing sensitive customer recordings. Yet a persistent distributional gap between synthetic and real data limits how far it can replace genuine recordings. Prior work largely treats this gap as a black box to be engineered around, but in our work, we instead examine its origin directly by probing a SLAM-ASR architecture. Then, we localise where its LLM backbone separates real from synthetic speech and find the discriminative signal concentrated in the early-to-middle layers, where temporal and prosodic perturbations disrupt it most. We further show that representation-level separability, help, but does not directly predict downstream ASR gains. On the other hand, convolving synthetic audio with room impulse responses (RIRs) narrows the gap not by making synthetic speech sound cleaner or more natural, but by reproducing the acoustic irregularities of real recordings. Translating these findings into the training procedure, by adding a layer-selection module combined with RIR augmentation matches a fully real-data baseline using only 25% of the real speech (13.6h) and surpasses it at all higher proportions.
arXiv:2607.03663v3 Announce Type: replace
Abstract: The accurate estimation of Above-Ground Biomass (AGB) in mature tropical forests remains a critical challenge in remote sensing, primarily due to the saturation of Synthetic Aperture Radar (SAR) signals in high-density areas and persistent cloud cover affecting optical imagery. To overcome these physical limitations, we propose the Trimodal Coherent Co-attention Transformer (TCCT), a physics-informed deep learning architecture. The TCCT natively fuses optical surface reflectance (Landsat-5) with complex-valued Polarimetric SAR Interferometry (PolInSAR) data from both P and L bands. Unlike traditional fusion methods, our architecture employs complex-valued encoders to preserve spatial phase coherence, coupled with a dynamic co-attention mechanism that acts as an adaptive gating module, reducing the weight of cloud-corrupted optical pixels and shifting reliance to microwave phase data. We also derived a localized spatial allometric calibration model via Levenberg-Marquardt optimization, tailored to the specific wood density of the Paracou region in the Amazon basin. Evaluated using a two-stage protocol, the TCCT first underwent a rigorous 5-fold cross-validation to establish robust global weights (achieving a global RMSE of 4.19 m). Subsequently, following a localized spatial fine-tuning phase over 200 epochs, the model attained an absolute RMSE of 3.78 m and an $R^2$ of 0.33 for Canopy Height Models (CHM), outperforming standard Random Forest, CNN, and Vision Transformer baselines. Our ablation study confirms that preserving phase coherence mitigates deep-canopy signal saturation. When converted to AGB, the fine-tuned TCCT map yielded a Relative RMSE (rRMSE) of 4.51% in dense forest areas above 50 Mg/ha. By meeting the European Space Agency (ESA) BIOMASS mission requirement of less than 20% error, the TCCT provides a robust framework for continuous carbon stock mapping in tropical biomes.
arXiv:2607.08261v1 Announce Type: new
Abstract: We prove that for every Cayley graph $\mathcal{G}$ over any finite abelian group $G$, there is a weighted Cayley graph with $O(\log |G|)$ generators that is a spectral sparsifier for $\mathcal{G}$. This bound is optimal. Applying our bound to the group $G = \mathbb{F}_2^n$, yields, as a corollary, $O(n/\varepsilon^2)$-sized code sparsifiers for $\mathbb{F}_2$-linear codes, improving on the work of Khanna, Putterman and Sudan (SODA'24) who obtained a similar result with an additional $\mathrm{polylog}(n)$ loss.
Our proof is strongly inspired by a recent work of Reis and Rothvoss for the construction of $\ell_1$-sparsifiers. Following their work, the abelian Cayley sparsification problem can be reduced to establishing a lower bound for the volume of a certain natural convex body. This volume bound follows from a short, elementary argument that relies on character symmetry.
arXiv:2607.08268v1 Announce Type: new
Abstract: High-volume structured extraction pays a large model's latency on every item, so distilling the task into a small on-device model is attractive: comparable output at a fraction of the time and cost. We measure what that distillation actually delivers, per sub-task. Each news article is mapped to one JSON object with a short summary and five categorical labels. We distill an 8B reasoning teacher (deepseek-r1:8b) into a 0.6B student (Qwen3-0.6B; QLoRA, three seeds), and add two teacher controls: a same-size non-reasoning teacher and a larger managed pipeline. A blinded, reference-free, three-judge panel scores every arm against the full article, alongside two non-distillation baselines, few-shot prompting and constrained decoding. The student runs at about 0.8 s per article against the teacher's 39 s, and recovers 58% of the base-to-teacher gap on summary quality, beating its primary baseline (constrained decoding) by +16.8 points and few-shot prompting by a secondary +4.9. A same-size non-reasoning teacher trains a student no better than the untuned base, so the summary gain follows from the teacher's reasoning nature rather than its scale. Capabilities then split by teacher: the reasoning teacher transfers writing quality and the managed pipeline transfers label diversity, while a same-size instruction teacher's students stay more grounded on the 22 short, thin-source articles in the 93-item test set (74 versus 55 faithful), where the reasoning-lineage student fabricates. That grounding difference is a consistent ordering rather than a significant aggregate effect, and the subgroup is small, so we report it as a direction. Because no single engine wins every field, the deliverable is a per-field routing map for on-device enrichment.
arXiv:2310.07895v2 Announce Type: replace
Abstract: This paper presents a method to efficiently classify the gastroenterologic section of images derived from Video Capsule Endoscopy (VCE) studies by exploring the combination of a Convolutional Neural Network (CNN) for classification with the time-series analysis properties of a Hidden Markov Model (HMM). It is demonstrated that successive time-series analysis identifies and corrects errors in the CNN output. Our approach achieves an accuracy of $98.04\%$ on the Rhode Island (RI) Gastroenterology dataset. This allows for precise localization within the gastrointestinal (GI) tract while requiring only approximately 1M parameters and thus, provides a method suitable for low power devices
arXiv:2607.08437v1 Announce Type: new
Abstract: Authorities increasingly rely on social media to advance sustainability transitions, infrastructure investment, and service reform. Yet how citizens respond to these digital communications remains poorly understood. Existing approaches rely on aggregate engagement metrics (e.g., likes), providing limited insight into discourse structure and quality. We developed a data-driven, multidimensional framework to analyse how social media communication shapes the content of discourse, focusing on sustainability-related engagement in Dutch public housing. We analysed 792 posts and 3,197 tenant comments from the Facebook pages of 92 housing providers (2018-2023). A machine-learning pipeline classified comments into recurring discourse configurations across three dimensions - communicative intent, sentiment, and semantic relatedness. Multinomial logistic regression estimated the effects of post-design and organisational characteristics on discourse. Tenant comments were significantly more semantically aligned with their corresponding posts than with randomly paired content, indicating that organisational communication structures responses to topics. Six discourse types emerged, with critical and inquiry-driven engagement increasing over time. Post-level features did not significantly explain variation; organisational characteristics dominated. Larger housing associations attracted more substantive responses, while lower-rent organisations received fewer evaluative comments. While applied to housing associations, our methodology provides a scalable approach to analyse online discourse dynamics, quality, and content across organisations and contexts.
arXiv:2607.08297v1 Announce Type: new
Abstract: Background and Objective: Quantitative analysis of cell dynamics is central to modern biological research, providing critical insights into immune cell interactions, disease progression, and drug mechanisms. Automated cell tracking in time-lapse microscopy remains challenging due to noise, morphological variations, overlapping cells, and dynamic events such as divisions and fusions.
Methods: We present ARGUS, a framework for Accelerated, Robust, General, and Unsupervised Cell Tracking Solutions. ARGUS combines adaptive cell detection, dense Farneback optical-flow prediction, frame-to-frame linear assignment, and a sequence-level tracklet-refinement step that reconnects trajectory fragments across short temporal gaps.
Results: On publicly available Cell Tracking Challenge datasets, ARGUS achieved detection accuracy of 0.905-0.971 and tracking accuracy of 0.897-0.964, with runtimes within 1 minute (5-6 seconds for 3 frames).
Conclusions: ARGUS is a modular, interpretable framework that can be adapted to different imaging modalities and biological applications without training data or GPU infrastructure. The implementation is publicly available at https://github.com/Gitinc/argus
arXiv:2607.08104v1 Announce Type: new
Abstract: Stochastic gradient descent (SGD) is a cornerstone of modern optimization. While its performance under heavy-tailed noise is often addressed through specialized modifications such as gradient clipping or normalization, we investigate a more fundamental question: how does vanilla SGD, particularly with momentum, perform in the presence of heavy-tailed noise? In this paper, we refine existing convergence results for vanilla SGD and, more importantly, provide the first comprehensive convergence analysis of vanilla SGD with momentum for strongly convex, convex, and nonconvex objectives, without employing any gradient control mechanisms. Our results demonstrate that the obtained convergence rates are inferior to the optimal rates achieved by clipped or normalized variants of SGD, thereby revealing inherent limitations of vanilla methods under heavy-tailed noise. The theoretical findings are supported by experiments on synthetic functions.
arXiv:2607.07719v1 Announce Type: new
Abstract: Parameter-efficient fine-tuning adapts a large language model to one task cheaply, but across a task sequence LoRA-style methods keep stacking low-rank updates on the same frozen weight, so each new task tends to overwrite the previous ones. We present ReCoLoRA (Recursive Consolidation of Low-Rank Adapters), a spectrum-aware framework for continual fine-tuning: adapters are initialized from a randomized SVD of the pretrained weight, per-layer effective ranks are selected by an elbow criterion, and the principal subspace is adapted before residual capacity is opened. Before each new task, ReCoLoRA re-decomposes the current effective weight, rather than the original one, into a frozen residual, a slowly updated principal component, and a fresh adapter (recursive consolidation), so every task starts from the model that has already absorbed its predecessors. On a six-task continual GLUE sequence over four 7-8B backbones, ReCoLoRA attains the best final average score on three of the four backbones against rank-swept LoRA, PiSSA, AdaLoRA, and DoRA baselines while training fewer parameters; an oracle-routed task-bank variant serves as an upper bound under full task isolation. Code: https://github.com/bhqy666/ReCoLoRA.
arXiv:2607.07724v1 Announce Type: new
Abstract: Block-sparse attention scales long-context language models by replacing the O(N^2) softmax with a per-query top-k selection over key blocks. This cutoff is myopic: when the k-th and (k+1)-th blocks are nearly tied in score, the selector commits without spending extra budget, and a dropped block carrying answer evidence is unrecoverable downstream. We propose a value-of-information router that measures, for each query, how decisively the top-k cut was made, and doubles the kept set for the queries where that gap is smallest; the rule is backbone-agnostic and stacks with existing block-scoring methods such as Quest. On LongBench-v2 medium at n=215 (the entire dataset subset), router-on-Quest reaches paired recall 0.75 vs. top-k 0.47 -- +28 pp over the SSA-style baseline (McNemar p<0.01) -- and lands within 2 pp of dense on RULER NIAH multikey at the same context. The lift reproduces on four models from three architectures (Qwen2.5, Mistral-Nemo, Qwen3.6). At 128K, the router preserves 0.81 and 0.89 of dense accuracy on Qwen2.5-7B-1M and Qwen3.6 (vs. SSA-style top-k at 0.09 on the former) while the fused selection-plus-kernel pipeline runs at 0.62x and 0.80x dense wall time.
arXiv:2607.07737v1 Announce Type: new
Abstract: GNSS-denied unmanned aerial vehicles require occasional absolute position fixes to bound the drift of visual-inertial odometry. Cross-view image retrieval can provide such fixes, but raw appearance is sensitive to season, illumination, viewpoint, map age, and sensor modality. We propose \sas, a semantic map-localization framework that represents the environment through persistent structures such as roads, buildings, waterways, railways, intersections, and field boundaries. The method combines semantic raster alignment, relational graph evidence, feature stability and geographic distinctiveness, explicit positive/contradictory/unknown observations, and integrity-aware rejection of ambiguous fixes. Unlike a broad architecture-only proposal, this paper specifies concrete weighting and decision models and reports a reproducible synthetic proof of concept. In 220 randomized retrieval trials with rotation, scale changes, partial crops, occlusion, simulated map changes, and hard semantic decoys, a global semantic descriptor achieved 58.6\% Recall@1, while spatial semantic matching variants achieved 94.5-95.5%. Wilson 95\% intervals separate the global descriptor from the spatial variants but overlap among the spatial variants, so the experiment supports semantic geometry rather than a definitive benefit from each proposed module. The preliminary experiment does not validate real-flight navigation; rather, it demonstrates that structured semantic geometry can discriminate locations under controlled cross-view perturbations and identifies the harder aliasing, map-aging, and rejection tests required next.