Optimization and Control
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Saturday 5 September
This paper studies a dynamical system approach for solving indefinite quadratic programming problems subject to linear constraints. We investigate the convergence of the trajectory generated by the system to a Karush-Kuhn-Tucker point of the quadratic programs. In addition, we derive an estimate for the distance between the trajectory and a solution of the problem. We further prove that the objective value is decreasing along the trajectory. An illustrative example and a numerical test are presented to demonstrate the behavior and performance of the proposed method.
Optics
Construction-Constrained Bernstein Synthesis of Joint Pitch-Width RDL Spiral Inductors for High-Q mmWave FoWLP Applications
We present a construction-constrained Bernstein framework for joint pitch-width synthesis of planar spiral inductors. A positive cubic Bernstein profile is integrated to prescribe a monotonic radial trajectory, while an independent Bernstein profile controls the conductor width. The nonnegativity, partition-of-unity, and convex-hull properties of the Bernstein basis convert endpoint, winding-direction, and width requirements into algebraic constraints. Embedded in differential evolution (DE), these constraints reject inadmissible candidates before full-wave evaluation, focusing expensive solver calls on geometries that satisfy the construction rules. The method is demonstrated on a two-turn mmWave inductor implemented in the redistribution layer (RDL) of fan-out wafer-level packaging (FoWLP), with the quality factor selected as the optimization objective. Searching the constrained Bernstein design space with DE yields a simulated Q of 34.85 at 30 GHz, compared with 29.66 for a uniform Archimedean baseline under the same stack, footprint, and ports. This corresponds to a 17.5% improvement while reducing planar copper area by 1.53%. The optimized response maintains Q >= 30 from 16 to 49 GHz.
Read the full abstract
Distributed Computing
Poseidon: DAG-Guided Parallelism Search for LLM Pre-Training on Heterogeneous Clusters
With the rapid advancement of accelerator technologies, pre-training large language models (LLMs) on heterogeneous accelerator clusters has become increasingly crucial for maximizing hardware utilization. Existing systems, however, suffer from inaccurate training time modeling, which undermines the parallelization optimizations built upon it. Moreover, for current approaches, the vast configuration search space makes exhaustive exploration infeasible, forcing a trade-off between search time and training efficiency. To overcome these limitations, we introduce Poseidon, an efficient and scalable LLM training framework designed with heterogeneity awareness. Its core is an explicit training time model based on a directed acyclic graph. Building on this graph, Poseidon employs two efficient, theoretically grounded strategies: stage-level pruning via early stopping with partial estimation, and layer-to-stage mapping exploiting a ridge-like distribution pattern. These strategies reduce the search space without sacrificing optimal training efficiency. Experiments on heterogeneous clusters show that Poseidon improves training throughput by up to $2.76\times$ over state-of-the-art systems.
Read the full abstract
Trading and Market Microstructure
Explainable Deep Learning for Price-Trade Dynamics: From Black-Box Forecasts to Effective Parametric Models
Understanding the joint dynamics of prices and trades is central to market microstructure, where returns and order flow interact through nonlinear and state-dependent mechanisms. Linear models are interpretable but may miss these effects, while deep neural networks improve forecasting at the cost of transparency. We use neural networks as tools for structural discovery rather than only for prediction. A deep feed-forward network is trained on high-frequency returns and signed volumes for large- and small-tick stocks and compared with a linear VAR benchmark. The neural network improves predictive performance, especially for returns, revealing nonlinear dependencies beyond the linear specification. Using Shapley-based explainability, we show that the dominant contributions are concentrated at the most recent lags. Model-implied responses are consistent with conditional averages reconstructed from the data. Unlike empirical averages, however, the neural-network decomposition isolates individual regressor contributions to the aggregate dependence. Lagged signed volume generates sign-preserving and saturating effects, consistent with nonlinear price impact and order-flow persistence. Lagged returns act as state variables: when the previous trade does not move the price, the model predicts continuation in the direction of past order flow, whereas non-zero returns generate attenuation or reversal. Building on these findings, we introduce a parsimonious SHAP-inspired nonlinear parametric model. It reproduces the main return-volume dependencies, outperforms the linear VAR benchmark, and achieves performance comparable to the neural network. A multi-lag extension captures residual longer-memory effects while preserving interpretability. Overall, explainability offers a route from black-box prediction to economically meaningful parametric models of price and trade dynamics.
Read the full abstract
General Relativity and Quantum Cosmology
A principal-symbol obstruction in metric-affine Gauss–Bonnet gravity
The four-dimensional Riemannian Gauss–Bonnet combination admits more than one continuation to a geometry with an independent affine connection. Two natural candidates, the metric-contracted representative and the double-epsilon representative, agree in their Levi-Civita limit, and we ask whether they also give the independent connection the same leading differential structure. In a metric-affine model with a scalar coupling $f(\phi)$ to the Gauss–Bonnet term, we compare their quadratic two-derivative connection principal forms about the Levi-Civita locus. For the metric-contracted continuation, the principal form, evaluated on a twenty-four-dimensional hook-free torsion-free witness subspace at a timelike principal covector, has rank $18$ and nullity $6$ and is indefinite; the full torsion-free form at the same covector is therefore indefinite with rank at least $18$. Locally where $f(\phi)\neq0$, the two-derivative principal symbol of the linearized connection equations is nonzero, which obstructs purely algebraic local elimination of the connection in the activated directions. For the double-epsilon continuation, the corresponding quadratic two-derivative form vanishes identically for an arbitrary perturbation of the connection about that locus, so its principal symbol vanishes with it, and that particular obstruction is absent, though its absence does not by itself establish algebraic eliminability. Thus the leading differential structure of the independent connection depends on the affine continuation of the Riemannian Gauss–Bonnet combination.
Read the full abstract
Machine Learning
Learning to Price and Stock Under Contextual and Censored Demand
To make optimal joint pricing and inventory control decisions is a critical challenge for modern retailers. In practice, retailers face changing market conditions where demands are influenced by various contextual factors, while simultaneously dealing with the difficulty of lost sales that obscure true demand information. However, existing approaches often fail to account for both contextual information and censored demand observations. We address this gap by presenting a framework where we model demand as a linear combination of basis functions with unknown coefficients, allowing for adaptive pricing and inventory decisions that respond to changing contexts. We propose an efficient algorithm to achieve regret bound $\mathcal{O}(K\sqrt{T}\log T)$ under concave revenue conditions and $\mathcal{O}(K^{2/3}T^{2/3}(\log T)^{1/2})$ for the general case, with matching lower bounds confirming optimality. Extensive numerical experiments across diverse scenarios demonstrate our algorithm's effectiveness.
Read the full abstract
Numerical Analysis
Polytopic symmetric interior-penalty approximation of elliptic SPDEs driven by spatial white noise
We analyse symmetric interior-penalty discontinuous Galerkin approximations of elliptic stochastic partial differential equations driven by spatial Gaussian white noise on quasi-uniform polytopic meshes. The low regularity of the exact random field precludes the usual SIP consistency argument. We instead restrict the same continuum white-noise functional canonically to the discontinuous polynomial space and separate the error into stochastic forcing-restriction and deterministic discretisation components. For every fixed polynomial degree $k\geq1$, we prove the sharp two-sided estimate \[ \|Y-Y_h\|_{L^2(\Xi;L^2(\Omega))} \simeq h^{2-d/2}, \] giving rates $h$ in two dimensions and $h^{1/2}$ in three dimensions. A finite-dimensional lower bound establishes optimality. The analysis allows polytopic elements with arbitrarily many and arbitrarily small faces, including admissible non-convex elements. In two dimensions we also treat polygons with a single reentrant corner. Although the deterministic approximation exhibits an angle-dependent loss of regularity, the singular resolvent contribution is finite rank, and the sharp first-order stochastic rate is retained. The discontinuous structure yields an exact element-local sampler for the restricted white noise. Numerical experiments on irregular polygonal, non-convex polytopic and reentrant domains support the predicted convergence behaviour.
Read the full abstract
Other Condensed Matter
Multiplicity of surface polariton configurations in bi-isotropic media with anomalous Hall current
In this work, we investigate surface polaritons (SPs) arising at the interface between a simple dielectric and a bi-isotropic medium, in the lossless regime and in the presence of small losses. For the former, non-Dyakonov surface polaritons can arise for positive permittivities and magnetoelectric parameters. General restrictions on the permittivities and bi-isotropic parameter (existence conditions) are derived. For the latter, the effect of small losses is considered, and the characteristic surface plasmon-polariton lengths are discussed. The propagation length can exhibit a local maximum for specific values of the magnetoelectric parameter, an unusual new behavior that can be regarded as a signature of these chiral surface polaritons. Surface waves in an interface involving a bi- isotropic medium with an anomalous Hall effect (AHE) are also examined. Multiple configurations for their occurrence are obtained, revealing a richer scenario with distinct regimes of propagation and coexistence in the parameter space. The presence of a non-null AHE contribution creates a magnetoelectric range for which the surface wave propagation is forbidden, providing an additional signature of surface-polariton propagation in bi-isotropic media with an AHE contribution.
Read the full abstract
Machine Learning
PhenoBench: Mapping What a Deeply Phenotyped Human Cohort Can Tell Us
Deeply phenotyped cohorts combine clinical, imaging, molecular, and wearable observations across timescales from seconds to years, but heterogeneous analyses are not directly comparable. We present PhenoBench, an executable benchmark that turns deep-phenotyping measurements into explicit questions and controlled comparisons of information sources and predictive models. It is built around the Human Phenotype Project, with more than 13,000 participants at the initial visit. Each question fixes the target, population, timing, and allowed information; its evaluation contract specifies the split, metric, baseline, and claim boundary. PhenoBench defines 90 clinically grounded tasks across 15 domains and 26 input modalities. Across 160 matched regression comparisons spanning 52 tasks, six pretrained tabular models ranked above the evaluated task-specific baselines, including XGBoost and CatBoost, under a fixed single-estimator protocol with bounded tuning. Giving each task equal weight, their mean advantage over ridge was 0.0103 $R^2$ (95% task-bootstrap interval, 0.0071-0.0136). We also evaluated 14 language models, collectively covering 40 tasks spanning phenotype recovery, classification, follow-up forecasting, and participant ordering. Without cohort-specific fitting, language models made informative predictions on some tasks but showed task-specific capability gaps, shared failures of scale, and rarely surpassed task-specific ridge or logistic regression models fitted on the same input fields. PhenoBench provides a versioned, auditable evaluation system where new questions, measurements, and models can be added without redefining existing comparisons.
Read the full abstract
Artificial Intelligence
LayerRoute: Action-Conditioned Mixture-of-Layers Routing for Vision-Language-Action Policies
Vision-Language-Action (VLA) policies leverage pretrained vision-language models (VLMs) to guide action generation for robot control. VLMs provide hierarchical visual-semantic representations that evolve across layers, from local visual geometry to abstract, language-aligned semantics; different manipulation tasks may therefore require different mixtures of layer representations. Meanwhile, the action module maintains intermediate representations that evolve throughout action computation and may provide useful information for subsequent decisions. However, existing VLA interfaces offer limited flexibility in representation access: VLM information is exposed through fixed layer assignments for each action layer, while intermediate action states are only propagated implicitly through residual streams without explicit reuse. We introduce LayerRoute, an action-conditioned representation routing interface that enables adaptive access to VLM layers and action representations. The Layer Mixture Router dynamically forms mixtures of cached VLM representations, while Action-State Reread reuses earlier action representations. Across diverse simulation and real-world benchmarks, LayerRoute consistently improves StarVLA-$\pi$ and $\pi_{0.5}$, achieving up to 7.2 gains on LIBERO Long with only 0.31% / 3.87% additional parameters. Ablation studies validate the benefit of action-conditioned layer routing, while routing analyses reveal structured allocation patterns across action layers and task settings.
Read the full abstract
Computer Vision
Report of the 8th LSVOS Challenge: Complex and Multimodal Video Object Segmentation
This report summarizes the 8th Large-scale Video Object Segmentation (LSVOS) Challenge, held in conjunction with ECCV 2026. The challenge evaluates video segmentation in three complementary settings: complex semi-supervised video object segmentation on MOSEv2, text-guided referring video object segmentation on MeViSv2-Text, and audio-guided referring video object segmentation on MeViSv2-Audio. We describe the tasks and evaluation protocols and review the methods of the top three teams in each track. Across the nine leading solutions, foundation segmentation models are combined with target-aware memory, multimodal reasoning, explicit target-existence verification, agentic interaction, and corrective tracking. These systems illustrate a broader transition from single-model mask propagation toward modular pipelines that reason about object identity, query validity, and temporal reliability.
Read the full abstract
Signal Processing
SPARC: Sparse Path-Aware Residual Calibrator for Wireless Ray Tracing at Upper Mid-Band
Accurate site-specific ray tracing (RT) is essential for upper mid-band network planning, yet raw RT can produce per-path multipath component (MPC) power errors on the order of 19–24~dB in cluttered indoor environments. A fixed-geometry material-sensitivity bound shows that a 30% relative-permittivity perturbation changes each surface interaction by at most 6.28~dB across the considered indoor materials. However, even MPCs with only one surface interaction exhibit a 19.2~dB mean RT–measurement bias, suggesting that missing clutter, displaced surfaces, and simplified 3D geometry dominate the per-path RT error. We propose SPARC (Sparse Path-Aware Residual Calibrator), a lightweight per-path calibration method that learns a sparse linear residual model from one completed RT simulation. SPARC uses standard RT features selected per fold by nested cross-validation, with ridge regularization and power-gated path matching; four features recur in both environments. Using measured indoor factory (InF) and indoor hotspot (InH) datasets at 6.75 and 16.95~GHz, SPARC reduces per-path power RMSE from 18.74 to 4.74~dB in InF and from 23.12 to 5.39~dB in InH. A jointly trained InF+InH model achieves 5.73~dB RMSE. When all links from one transmitter location are held out for testing, SPARC achieves 4.99~dB RMSE in InF and 5.85~dB RMSE in InH. SPARC therefore provides a practical post-processing calibration layer for site-specific per-path power prediction without ray-tracer modification or additional RT runs.
Read the full abstract
Machine Learning
Beyond Retraining-Free MoE Compression: A Cost-Normalized Study of Post-Compression Adjustment
Retraining-free MoE compression reduces deployment memory by pruning or merging experts, but often treats the compressed checkpoint as the final artifact. We argue that this view is incomplete: compressed MoE checkpoints are better understood as compressed initializations that benefit from a tiny post-compression adjustment stage. Across two MoE LLM backbones, four pruning/merging methods, three expert-retention ratios, and 28 benchmarks, we compare LM fine-tuning and teacher-based KD under matched small-data budgets and measured GPU costs. Using only 3,000 C4 examples and a single epoch of adjustment, Full FT recovers 37.3% of the original-to-compressed performance gap on average. Moreover, LM fine-tuning is more cost-effective than standard token-level KD, and full-parameter adjustment gives the strongest cost–recovery trade-off among the tested scopes. These results suggest that retraining-free compression should be paired with small post-compression adjustment to recover a substantial portion of the performance lost during compression.
Read the full abstract
High Energy Physics (Theory)
Non-critical M-Theory and the Resolved Conifold: Refinement and Nonperturbative Completion
We extend the Hořava–Keeler correspondence between finite-temperature non-critical M-theory and the resolved-conifold A-model to the refined theory. The grand potential of the stationary rotating non-critical M-theory vacuum reproduces the refined Gopakumar–Vafa expansion, with the angular chemical potential deforming the $\Omega$-background away from the self-dual locus $\eps_1=-\eps_2$. At finite thermal radius, we establish an exact match with the nonperturbative conifold completions proposed by Hattab–Palti and by Chuang in the unrefined and refined cases, respectively. Both the one-particle Schwinger integrand and its integration cycle arise from the non-critical M-theory spectrum and resolvent, providing a microscopic spectral realization of these completions. The correspondence fixes the nonconstant BPS sector and the cubic local contribution, while the remaining polynomial ambiguity and the modulus-independent constant-map sector require separate normalization. We explain why the single primitive spinless multiplet of the resolved conifold makes this identification possible, and discuss what additional charge- and spin-dependent microscopic data would be needed for an extension to more general local Calabi–Yau geometries.
Read the full abstract
Computer Vision
NSFlow: End-to-End Differentiable Neuro-Symbolic Optical Flow for Visual Odometry
Sparse optical flow provides stable inter-frame correspondence, playing a key role in Visual Odometry (VO) and Visual-Inertial Odometry (VIO). Classical optimization-based methods, such as Lucas-Kanade (LK), perform well under small displacements but are sensitive to large motions and illumination changes. Modern regression-based learning methods, while more robust in complex scenes, are often computationally heavy and lack explicit geometric consistency, making them less suitable for efficient VO/VIO front-ends. To bridge this gap, we propose a hybrid neuro-symbolic framework that combines the strengths of both paradigms. Our method uses a Convolutional Neural Network (CNN) to extract robust feature representations, which is fed into a differentiable LK optimizer to estimate optical flow in an end-to-end trainable manner. Through implicit differentiation, gradients are propagated across the iterative solver, enabling joint optimization of feature extraction and flow estimation. The resulting system integrates seamlessly into existing VO/VIO pipelines and runs in real-time on embedded platforms. Experiments show that our method outperforms conventional optimization-based flow in challenging conditions such as dynamic lighting and low texture, while also achieving higher accuracy and lower latency than purely regression-based alternatives. When deployed in a VIO system, our method demonstrates significant performance improvement, achieving an average error reduction of 42% on challenging datasets while enhancing tracking stability. The code is publicly available.
Read the full abstract
Machine Learning
FedSubMuon: Communication-Efficient Federated LLM Fine-Tuning via Structured Subspace Muon
Federated fine-tuning adapts large language models (LLMs) to decentralized client data, but its scalability in cross-device training is often limited by the high communication cost. Muon is an optimizer that improves optimization performance by orthogonalizing momentum for matrix-valued parameters. Existing federated Muon methods demonstrate the benefit of matrix-aware optimization in federated learning, but still require transmitting full layer-size updates and optimizer state. A natural way to reduce communication is to directly apply Muon to LoRA factors, but this changes the optimized object and weakens Muon's matrix-aware update geometry. We propose FedSubMuon, a communication-efficient federated Muon fine-tuning method that optimizes compact coefficient matrices within shared structured subspaces. This design keeps Muon on a single matrix-valued trainable object, while reducing the client upload to compact coefficient matrices. We further introduce FedSubMuon-GT, an accuracy-oriented extension that uses projected gradients to adapt tracked subspace bases toward task-relevant gradient directions. Experiments on instruction tuning and mathematical reasoning show that FedSubMuon-GT achieves the best overall accuracy on four of five dataset-model pairs, while FedSubMuon performs best under all matched communication budgets. On Dolly-15K, the closest communication baseline requires 5.5 times and 1.4 times more total communication on Llama-1B and Qwen-4B, respectively.
Read the full abstract
Machine Learning
ACE: Adapter Consolidation across Experts for Parameter-Efficient Fine-Tuning of MoE LLMs
Parameter-efficient fine-tuning (PEFT) of mixture-of-experts (MoE) models commonly attaches a separate low-rank adapter to each expert. This expert-wise design fragments adaptation in three ways: capacity is split across narrow low-rank updates, gradient supervision becomes sparse and imbalanced under sparse routing, and execution is decomposed into many small GEMMs. We find that such expert-wise separation is often unnecessary, as subsets of LoRA adapters become functionally similar during fine-tuning, revealing redundancy among expert-specific adapters. Based on this redundancy, we propose ACE (Adapter Consolidation across Experts), which groups redundant experts and replaces their expert-specific adapters with group-shared higher-rank LoRA modules under the same PEFT budget. ACE further introduces grouped adapter execution, which consolidates fragmented expert-wise adapter computations into fewer, larger group-level GEMMs. Across evaluations covering 12 datasets and four MoE backbones, ACE achieves the highest observed mean accuracy among the parameter-matched PEFT methods on the three backbones with complete baseline coverage, while providing $1.31\times$ to $1.48\times$ wall-clock training speedup over expert-wise LoRA without increasing peak memory. Our code is available at https://github.com/UbiquitousAILab/ACE.
Read the full abstract
Artificial Intelligence
Generating Instance Generators in PDDL Planning
PDDL, the de-facto standard language in the AI Planning community, is designed to specify planning domains: sets of instances that share the same predicates and action schemas. Yet it does not provide any means to specify the actual instance set, i.e., legality constraints on initial states and goal conditions, as well as possibly domain subset constraints specifying an instance subset we are interested in. One consequence of this is that instance generation has always been ad-hoc, with manually written domain- and subset-specific instance generators. Recent work has started to address this, through reasoning and learning methods that however suffer from scalability limitations. Here we introduce an alternative approach, leveraging LLMs to generate instance-generation programs, with built-in soundness guarantees through prescribed checks. We show that these automatically generated instance generators return large numbers of sound and diverse instances efficiently.
Read the full abstract
Computation and Language
DPH Parser: A Bottom-Up Grammar-Driven Parser for Joint Constituency and Dependency Analysis
This paper presents Dependency-Phrase Hierarchy Parser (DPH Parser), a grammar-driven bottom-up unsupervized parsing framework inspired by Generalized Phrase Structure Grammar (GPSG) and Head-driven Phrase Structure Grammar (HPSG). The parser incrementally constructs constituency structures using a compact inventory of feature-based syntactic rules while deriving dependency relations through explicit head annotations. The system combines probabilistic POS tagging, recursive phrase projection, and weighted parse hypotheses to process realistic and partially noisy text input. Unlike purely neural and data-driven parsers, the resulting syntactic derivations remain explicitly interpretable. We evaluated parser performance on English corpora from the Universal Dependencies (UD) project using Unlabeled Attachment Score (UAS) as the main parsing metric, comparing the outcomes against Stanza and spaCy parsers. For a small inventory of syntactic rules, DPH parser achieved UAS values of 53.32% & 52.58% (UD Devset/Testset respectively). For the same data, Stanza achieved 89.12% & 88.67% while spaCy achieved 56.91% and 58.59%. Although the current system does not yet approach the accuracy of modern neural parsers, the results demonstrate the feasibility of applying transparent rule-based bottom-up parsing to realistic treebank data while jointly producing constituency and dependency structures.
Read the full abstract
Instrumentation and Detectors
Long-term performance of non-ALD MCP-PMTs in the high-radiation environment of ALICE
The Fast Interaction Trigger detector of the ALICE apparatus at the LHC includes a Cherenkov subsystem utilizing 52 Planacon XP85002/FIT-Q microchannel plate-based photomultiplier tubes. It is a precise trigger, multiplicity and luminosity detector located in the forward region, where the flux of collision products is the largest. The photon load reaches up to 2*10^8 photoelectrons / cm^2 / s for the innermost photosensors, corresponding to 0.4 {\mu}A / cm^2 average anode current density. The lifetime-extended version of the Planacon MCP-PMT with atomic layer deposition cannot handle such a large anode current. We therefore use the non-ALD version of the Planacons that, on the other hand, suffer from a limited lifetime in terms of the integrated anode charge. Here we present the long-term performance trends of the non-ALD Planacons up to the integrated anode charge of 2 C / cm^2. We demonstrate that the low-gain operation strategy allows to maintain excellent time resolution in multiphoton mode unchanged up to that large integrated charge. While the majority of the equally illuminated MCP-PMTs feature almost identical ageing speed, two outliers suffering from a faster ageing coincide with the outliers in the noise characteristics measured prior to the irradiation. Moreover, we report that the effective ageing speed is partially suppressed by the self-recovery of the response of aged MCP-PMTs newly observed by our group during months-long no-beam periods.
Read the full abstract
Number Theory
Infinite-order $p$-adic differential equations for Hurwitz-type Euler zeta functions: Uniqueness and higher-dimensional extensions
We generalize the one-dimensional infinite-order linear differential equation satisfied by the $p$-adic Hurwitz-type Euler zeta function (Abh. Math. Semin. Univ. Hambg. 91: 117–135, 2021) to $n$ variables. By encoding the differential operator as a convolution with a distribution kernel, we introduce partial zeta functions and partial operators indexed by subsets of $\{1,\dots,n\}$. A tensor product expansion of the distribution kernel is established, whose Möbius inversion via the binomial theorem yields the higher-dimensional analogue of the original equation in the form of an alternating-sum identity, reducing to the one-dimensional case when $n=1$. Convergence of the resulting series is confirmed by non-Archimedean estimates. As a second main contribution, we prove that, under the condition $|a|_p > 2p^{1/(p-1)}$, within the Banach space of bounded analytic functions on $\mathbb Z_p$, the shifted $p$-adic Hurwitz-type Euler zeta function is the unique solution to the one-dimensional equation. This uniqueness result establishes that the infinite-order $p$-adic differential equation admits at most one bounded analytic solution, and together with the existence theorem, it characterizes the shifted $p$-adic Hurwitz-type Euler zeta function uniquely. This sharply distinguishes it from the complex setting, where the operator series diverges on the Hurwitz zeta function, and the formal kernel fails to converge in the usual spaces of analytic test functions.
Read the full abstract
Spectral Theory
Weinstock Inequality on Regular Trees
Let $T_n$ be the infinite $n$-regular tree, $n\ge3$. We prove that every finite connected vertex set $\Omega\subset T_n$ satisfies the sharp inequality \[ \sigma_1(\Omega)\le \frac{n}{(n-1)|\Omega|+1}. \] Equality holds if and only if $\Omega$ is a ball. Since \[ |\delta\Omega|=(n-2)|\Omega|+2, \] the result is equivalently a sharp upper bound at fixed external boundary cardinality, and hence a discrete Weinstock inequality on $T_n$.
Read the full abstract
Signal Processing
Scalability of Graph Neural Network Policies in Wireless Communication Networks
Graph Neural Networks (GNNs) offer scalable solutions for wireless resource allocation, yet existing formal performance guarantees across varying scales for spatial and sparse graphs do not directly apply to these settings. Scalability theories rely on dense graphon limits or continuous manifold approximations, both of which fail under the sparse, bounded-degree regimes and Euclidean metric constraints of physical wireless environments. This paper establishes a theoretical framework for GNN transferability over sparse Random Geometric Graphs (RGGs), capturing distance-dependent channel decay and spatial interference. We model sparse RGGs as spatial perturbations of regular Deterministic Grid Graphs (DGGs) and employ spatial windowing operators to compare networks across differing scales. Assuming Lipschitz continuity of GNN architectures and signal stationarity, we prove scalability over DGGs and bound same-scale transferability between DGGs and RGGs. Combining these results establishes formal scalability bounds across sparse RGG topologies. Finally, we extend this framework to conflict graph models, deriving equivalent scalability guarantees for link-level resource allocation policies. We verify our results for scalability with two experiment settings: power allocation and wireless link scheduling. The simulations show GNNs exhibit the expected scalable behavior, and analyze the relevance of our theoretical assumptions in practical deployment.
Read the full abstract
Computation and Language
Don't Lose Entities from Retrieval to Generation: Dual Entity Recovery RAG for multi-hop QA
Retrieval-augmented multi-hop question answering (QA) decomposes a query into sub-questions and decomposes the corpus into smaller retrieval units such as sentences. Both forms of decomposition improve the pipeline, but we show that both share the same vulnerability, the loss of entity information, and that this loss breaks the pipeline at two separate points. The first point is retrieval, where a sub-question loses the entity resolved at the previous hop, leaving the retriever with nothing to match against. The second point is harder to see, because retrieval still appears to succeed. Once a passage is split into sentences, an isolated sentence loses the context that grounds its pronouns, so even with the correct sentence in hand the LLM cannot tell which entity the sentence is about. We isolate this second point as a distinct failure mode that we call lost-in-generation, and a retrieval-controlled experiment shows that it degrades answers even when the gold evidence is fixed in the context. We then propose Dual Entity Recovery RAG (DER-RAG), which keeps the grounding entity explicit from retrieval through to generation with two lightweight components, a two-way query decomposition that carries the resolved entity across sub-questions and a subject entity prefix attached to each sentence at generation time. DER-RAG needs no graph construction, no corpus modification, and no fine-tuning, yet on three multi-hop QA benchmarks it matches or exceeds strong baselines, including graph-based methods that depend on costly offline structures.
Read the full abstract
Optimization and Control
The Role of Gradient Modification in Heavy-Tailed Nonconvex Stochastic Min-Max Optimization
Stochastic min-max optimization has attracted increasing attention due to its applications in modern machine learning, while existing theoretical studies mainly rely on the bounded variance assumption for stochastic gradients. Under heavy-tailed noise, where stochastic gradients only possess a finite $p$-th moment for $p\in(1,2]$, gradient clipping or normalization is commonly believed to be necessary to guarantee convergence. In this work, we revisit stochastic min-max optimization under heavy-tailed noise and provide a comprehensive theoretical study of stochastic gradient descent ascent (SGDA). We first show that vanilla SGDA, without any modification to its update rule, can converge under heavy-tailed noise in both nonconvex-strongly-concave (NC-SC) and nonconvex-concave (NC-C) settings, establishing the first convergence guarantees for SGDA in these regimes. Beyond unregularized problems, we further investigate regularized stochastic min-max optimization, where directly incorporating gradient normalization into proximal updates is nontrivial due to the incompatibility between normalization and proximal structures. We overcome this difficulty by developing new clipping-free algorithms, i.e., Stoc-TRGDAM and Stoc-TRGDmax, and they both can achieve the optimal dependence on the target accuracy without using gradient clipping.
Read the full abstract
Artificial Intelligence
Explaining AI Agents Through Execution Traces
AI Agents are increasingly deployed in real-world settings, where they interact with external tools and make sequential decisions with limited human oversight. This creates a pressing need for reliable and auditable explanations of what an agent did and why. However, traditional Explainable AI (XAI) methods fall short of providing the process-level transparency required for such interactive, multi-step systems, motivating a paradigm shift toward approaches specifically designed for AI Agents. To address this gap, we present a post-hoc XAI framework that transforms a lengthy agent's execution trace into a structured report and a faithful natural-language explanation explicitly grounded in its observable behavior. Because it relies solely on execution traces, the framework applies across different agent architectures, environments, and tasks. Human and automated evaluations across multiple benchmarks and architectures show that our framework produces high-quality, trace-faithful explanations while reliably identifying unsupported claims, unjustified actions, and evidence gaps, outperforming naive LLM-generated explanations.
Read the full abstract
Atomic Physics
Floquet Interpretation of Avoided Crossings in AC Stark-Shifted Rydberg-EIT Spectra
We present a combined experimental and Floquet-theoretical study of avoided crossings in the AC Stark-shifted electromagnetically induced transparency (EIT) spectra of Rydberg atoms. Observed avoided-crossing structures cannot be fully explained by conventional AC Stark maps alone, which provide the energy shifts but do not reveal the underlying state composition and coupling pathways. Using the Shirley method, we analyze the Floquet eigenstates to establish a direct connection between the measured spectra and the underlying dressed-state dynamics. By tracking the evolution of state mixing, bare-state composition, and dominant higher-order Floquet coupling pathways, we reveal the physical mechanisms responsible for the formation of avoided crossings across different principal quantum numbers and RF frequencies. Our results demonstrate that Floquet-eigenstate analysis provides a powerful framework for interpreting complex AC Stark-shifted Rydberg-EIT spectra and the underlying higher-order Floquet coupling interactions.
Read the full abstract
Robotics
SCoCaT: Success Conditioned Constrained Reinforcement Learning for Spacecraft Docking
Termination-based constrained reinforcement learning is attractive for safety-critical robotic deployments: it avoids online optimization at inference, scales easily to many constraints via a single scalar per constraint, and is simpler to implement than commonly used Lagrangian methods. Instead of pricing violations through summed cost penalties, this approach makes violations structurally unprofitable by shortening the effective horizon for each violation. We identify a structural failure mode of this method class on terminal-navigation tasks: reaching a precise goal configuration while satisfying safety constraints that tighten along the final approach. When the goal sits inside the region close to where the constraints become active, the survival-weighted objective makes dwelling outside the goal region strictly preferable to entering, producing high constraint compliance with low task completion. We formalize this pathology and show that a minimal augmentation to off-the-shelf RL algorithms like PPO resolves this “feasibility collapse”. We empirically demonstrate that adding a dense per-step success signal via an auxiliary value critic improves the task completion rate while maintaining safety-critical constraint compliance. Validation across two representative spacecraft platforms: a 6U-CubeSat spanning the mass and degree-of-freedom envelope of operational proximity operations, and a floating platform testbed for zero-shot sim-to-real transfer in our laboratory, supports the generality of these findings.
Read the full abstract
Machine Learning
Calendar-Structured Sparse Principal Component Analysis for Interpretable Multi-Periodic Electricity Consumption Profiles
Long-term electricity-consumption profiles exhibit several simultaneous periodic structures, including daily, weekly, and annual cycles. This work introduces Calendar-Structured Sparse Principal Component Analysis (Calendar-SPCA), a structured representation-learning method that incorporates this known multi-periodic geometry directly into a low-dimensional factorization. The method represents the feature domain as the Cartesian product of cyclic calendar axes and combines an L1 loading penalty with graph total variation, producing sparse, locally coherent, and directly interpretable latent factors. In this study, Calendar-SPCA is applied to the interpretable analysis of long-term electricity-consumption profiles and evaluated on two independent public smart-meter datasets, GoiEner and Low Carbon London, with different population sizes and temporal resolutions. A factorial experiment characterizes the effects of sparsity and calendar coherence and examines robustness across sample size, latent dimensionality, and repeated fits. At rank 15, Calendar-SPCA retains 96.92% and 82.90% of the explained variance of rank-matched principal component analysis (PCA) in GoiEner and Low Carbon London, respectively, with mean loading sparsities of 61.95% and 81.50%. Comparisons with classical sparse PCA and sparse PCA with total variation (SPCA-TV) show that Calendar-SPCA organizes latent factors into interpretable structures over the daily, weekly, and annual calendar axes while preserving substantial low-rank information.
Read the full abstract
Artificial Intelligence
OmniStepBench: Deployment-Aware Agent Evaluation across Modalities and Task Complexity
As large language models evolve from question-answering systems into general-purpose agents, evaluation must move beyond static answer correctness to assess multimodal perception, multi-step execution, tool use, and artifact delivery. However, existing benchmarks are often tied to specific task types, execution environments, or scoring protocols, limiting their comparability, interpretability, and reliability for deployment decisions. We introduce OmniStepBench, a benchmark for deployment-aware agent evaluation across modalities and task complexity. Built on a shared OpenClaw execution environment, OmniStepBench organizes 233 tasks selected and adapted from 22 source benchmarks into a $2\times3$ workload matrix defined by input modality and execution form, and evaluates them under a unified contract-based protocol with evidence-based score auditing. We evaluate 30 commercial API models and 12 locally deployed open-weight models (42 models in total) over 9{,}218 applicable model–task runs, reporting accuracy and token consumption alongside reference costs for API models. Results show workload-dependent rankings, distinct accuracy–cost trade-offs between text and multimodal tasks, and competitive but weaker overall performance from local models. These findings support deployment decisions based on workload profiles, deployment mode, and accuracy–cost trade-offs rather than a single aggregate score.