Search arXiv⌕ Search

arXiv subjects

Search papers

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

At least 973 records · Page 54Linked to original sources
arXiv2026

Testing Connectedness and Edge Connectedness

We revisit the classical problem of testing connectedness of graphs represented in the adjacency lists model and obtain optimal bounds. Specifically, we show that $\varepsilon$-testing connectedness of graphs with maximum degree at most $d$ requires $Ω(\frac{1}{\varepsilon} \log \frac{1}{\varepsilon d})$ queries (when $\varepsilon \le \frac{1}{4d^5}$). Our result implies that the connectedness tester of Levi, Pallavoor, Raskhodnikova, and Varma ({\em ACM Trans.\ Comput.\ Theory, 2022}) is optimal. We also study the problem of testing $k$-edge-connectedness of both directed and undirected graphs of average degree $\overline{d}$ in the adjacency lists model and design a tester with query complexity $O(\frac{k}{\varepsilon} \log \frac{1}{\varepsilon \overline{d}})$, improving upon the bounds by Forster, Nanongkai, Saranurak, Yang, and Yingchareonthawornchai (SODA 2020). Furthermore, we complement our upper bound by showing that every one-sided error $k$-edge-connectedness tester must make $Ω(k/\varepsilon)$ queries.

cs.DS↗
arXiv2026

Making every stage of medical device document assessment with large language models deterministic

To show that a medical device complies with regulation, experts judge whether documents and records meet each requirement and record the judgment for others to verify. For each assessment item that experts define, large language models could prepare a preliminary assessment: a label such as Satisfied or Not satisfied, a reason and quoted evidence. We sent fourteen commercial deployments of such models (seven models from three providers, through six routes) the same request repeatedly, with the deterministic sampling settings each accepted; every deployment returned differing assessments for at least one request, and in twelve the label changed. Reducing reasoning where the interface allows it did not reduce the variation, and prompts of retrieved passages removed it only for the smallest models. The same open-weight model was reproducible or not depending only on how it was served. A pipeline we built to be deterministic at every stage -- text extraction and optical character recognition (OCR), chunking, embedding, indexing, retrieval and inference -- returned byte-identical output at each stage across reruns, interruptions, edits, machines of one graphics processing unit (GPU) model and days, for blood glucose meter and infusion pump documents. OCR, embedding and approximate retrieval varied when their own serving was not made invariant, and outputs changed with the GPU model and with the inference engine's scheduling settings, which therefore belong in the recorded configuration. End-to-end determinism is achievable and gives a testable basis for audit and change control.

cs.CL↗
arXiv2026

Reconstructing constant point heat sources from few temperature sensors: sparse selection and spatial resolution

Thermal diffusion and measurement noise can obscure the number and individual parameters of localized heat sources. We study the reconstruction of constant positive point sources from temperature histories at a few sensors, without prescribing the source count. Orthogonal matching pursuit or least angle regression selects candidate locations from a unit-response dictionary. These candidates initialize continuous constrained fits, and a noise-based discrepancy rule selects the smallest accepted model among the attempted fits. The response analysis distinguishes rejection of a merged one-source model from accurate recovery of the individual sources. It gives a projected quadratic residual for a close pair, a weak direction in the two-source parameters, and a conditional bound for recovering the total strength and strength-weighted center of an unresolved cluster. A local Gaussian approximation relates one-source rejection to noise and separation. Paired numerical tests compare sensor and time-sampling budgets, noise levels, and source separations. Finite-plate data generated independently of the inverse response and a three-dimensional example assess the same fitting procedure. In the tested configurations, source-count selection and individual parameter recovery differ substantially, while unresolved pairs can retain accurate aggregate estimates. The framework combines reconstruction with a local interpretation of which heat-source information the observations support.

stat.ME↗
arXiv2026

Dusting the Phantom: Gaussian Allocation for Fading Dynamic Content

Dynamic Gaussian representations have made high-quality, real-time volumetric video increasingly practical, yet challenging motion can still break their reconstruction: small fast-moving objects may disappear, and fine textures in highly dynamic regions are often lost. In particular, simply training longer or using more Gaussians may still fail to recover small, fast-moving objects. We argue that the key bottleneck is not representational capacity itself, but where and when Gaussian primitives are allocated. Existing allocation strategies rely on current Gaussian locations and often overlook the spatiotemporal destinations of relocated Gaussians. To address these limitations, we introduce two complementary strategies. First, Seed Injection identifies under-reconstructed dynamic regions using temporal appearance changes and reconstruction errors, and initializes new Gaussians at geometrically plausible 3D locations through multi-view observations, independent of existing Gaussian locations. Second, Temporal Biased Relocation uses time-resolved optimization signals to guide the spatiotemporal placement of relocated Gaussians. Both strategies maintain a fixed Gaussian budget without modifying the underlying representation. Experiments across multiple dynamic-scene datasets demonstrate improved reconstruction quality, particularly in highly dynamic regions, recovering small, rapidly moving objects and fine-grained appearance details.

cs.CV↗
arXiv2026

Asymptotics of the Moduli Spaces of Parabolic $\mathrm{SL}(3,\mathbb C)$-Higgs Bundles

We study the large-scale geometry of the moduli spaces of stable strongly parabolic $\mathrm{SL}(3,\C)$-Higgs bundles on the three-punctured sphere for a family of parabolic weights. We glue the local Toda model metric to a decoupled metric obtained from the spectral line bundle, and perturb the resulting approximate metrics to harmonic metrics adapted to the parabolic structure. We also prove that Hitchin's $L^2$ metric approaches the semi-flat metric exponentially. The metric comparison and the spectral period calculation show that each moduli space is a complete ALG gravitational instanton of type $E_6$. The asymptotic estimates are uniform on compact subintervals of the weight chamber.

math.DG↗
arXiv2026

VALCON: Benchmarking and Enhancing Value Consistency in Large Language Models via Dual-Role Variation

Current value alignment evaluations for large language models (LLMs) heavily rely on output compliance, overlooking whether LLMs maintain value consistency when alternating between generator and evaluator roles. Such consistency is increasingly important because modern LLM agents and alignment pipelines often rely on the same LLM to both generate actions and evaluate their acceptability. To mitigate this limitation, we propose VALCON, a role-driven data construction framework that operationalizes value consistency into three benchmark suites (VAB-Consistency, VAB-Boundary, VAB-Challenge) via dual-role variation. By systematically pairing an LLM's generative behavior with its evaluative judgments under identical principles, VALCON synthesizes target evaluation data that exposes hidden value discrepancies. Experiments on 28 LLMs reveal pervasive cross-role inconsistency: models with highly compliant outputs frequently fail to align their judgments with their own generations. To understand and address this gap, we perform mechanistic probing that uncovers a role-dependent representational decoupling associated with this behavioral asymmetry. Guided by these insights, we propose a representation-directed intervention that effectively mitigates consistency failures. Our work establishes value consistency as a distinct alignment dimension and provides a complete evaluation-analysis-mitigation pipeline. Code and data are available at https://github.com/tjunlp-lab/VALCON.

cs.CL↗
arXiv2026

Unique Charge-density-wave Order and Fluctuation-induced Phonon Softening in the Optimally-doped Trilayer Cuprate Bi$_2$Sr$_2$Ca$_2$Cu$_3$O$_{10+δ}$

Charge-density-wave (CDW) order is a ubiquitous feature of cuprate superconductors that competes with superconductivity. However, its behavior remains poorly understood in trilayer compounds, which contain the optimal number of CuO$_2$ layers for superconductivity. Here, we report Cu $L_3$- and O $K$-edge resonant inelastic x-ray scattering (RIXS) measurements of optimally-doped Bi$_{2}$Sr$_{2}$Ca$_{2}$Cu$_{3}$O$_{10+δ}$ (Bi2223), the trilayer cuprate with the highest superconducting transition temperature ($T_c$) among Bi-based cuprates. We observe CDW order with a single in-plane wavevector of ${\bm q}_{\text{CDW}} = (0.25, 0)$, demonstrating that a coherent CDW order is formed across the trilayers. We also resolve CDW fluctuations emanating from the static order, whose profiles differ between the Cu $L_3$ and O $K$ edges. The phonon softening occurs where these CDW fluctuations interact with the phonon branch, indicating that the RIXS phonon anomaly is driven by the coupling between CDW fluctuations and the phonon branch.

cond-mat.supr-con↗
arXiv2026

VLA-$Δ$: Learning Physical State Transition for Vision-Language-Action Models

Vision-language-action (VLA) models transfer semantic knowledge acquired through vision-language pretraining to robot control, achieving strong performance across various benchmarks. However, most VLAs directly map observations to demonstrated actions without explicit supervision of intermediate physical state changes, making them prone to overfitting observation-action correspondences. Some methods introduce intermediate reasoning or future-state prediction in language, visual, or latent spaces to guide action generation, but do not explicitly supervise the physical state transition itself. We therefore propose VLA-$Δ$, a general framework that uses physical changes between current and future states as explicit learning targets. To support this objective, we construct a large-scale pretraining dataset and derive physical state-transition annotations covering task-relevant semantics and visual motion changes, providing more direct supervision for continuous action generation. During inference, the model generates actions directly from current observations without additional inference overhead. VLA-$Δ$ achieves performance competitive with the state of the art on LIBERO and RoboTwin, and outperforms $π_{0.5}$ in real-robot experiments. Our code is available at https://github.com/ZJU-OmniAI/VLA-Delta

cs.CV↗
arXiv2026

Nonleptonic decays of the double-charmed baryon $Ξ_{cc}^{+}$ with dynamically determined baryon wave functions

Motivated by the recent observation of the double-charmed baryon $Ξ_{cc}^{+}$ by the LHCb Collaboration, we systematically investigate the two-body nonleptonic decays of the $Ξ_{cc}^{+}$ into a single-charmed baryon and a light pseudoscalar meson. In our calculation, both factorizable and nonfactorizable decay amplitudes are evaluated within the framework of nonrelativistic quark model combined with the baryon wave functions dynamically determined by numerically solving the three-body Schrödinger equation with a nonrelativistic potential model. In particular, the nonfactorizable $W$-exchange amplitudes are estimated under pole model by considering $J^{P}=1/2^{\pm}$ low-lying intermediate charmed-baryon poles, and our results show that these contributions are comparable to or even dominate the factorizable amplitudes in several decay channels. Moreover, assuming a lifetime of $τ_{Ξ_{cc}^{+}}=47~\text{fs}$, we predict relatively large branching fractions for several CF decays, including $\mathcal{B}(Ξ_{cc}^{+}\toΞ_{c}^{\prime0}π^{+})=8.90\%$, $\mathcal{B}(Ξ_{cc}^{+}\toΩ_{c}^{0}K^{+})=7.09\%$, $\mathcal{B}(Ξ_{cc}^{+}\toΣ_{c}^{+}\bar{K}^{0})=6.92\%$, $\mathcal{B}(Ξ_{cc}^{+}\toΞ_{c}^{+}π^{0})=5.28\%$, and $\mathcal{B}(Ξ_{cc}^{+}\toΣ_{c}^{++}K^{-})=5.17\%$. These channels are promising candidates for future experimental searches at the LHCb.

hep-ph↗
arXiv2026

PocketTSLM: Deploying an On-Device Soft-Prompt Time-Series Language Model for Privacy-Preserving Smartphone Reasoning

Consumer phones and wearables increasingly collect physiological time-series data, yet most mobile systems treat these streams as summaries or text rather than as temporal evidence. This representation is poorly matched to personal health systems that need to retrieve, route, or summarize sensor evidence while preserving temporal structure. Cloud-hosted time-series language models avoid smartphone limits, but they repeatedly move sensitive physiology off the device and add latency and per-query compute cost. Mobile model alternatives that aggregate time-series data into statistical features or serialize raw samples as text either compress temporal structure or exhaust the context and memory budgets of small on-device models. We present PocketTSLM, a mobile-native deployment of OpenTSLM SoftPrompt, the soft-prompt variant of the OpenTSLM time-series language model. PocketTSLM runs fully on a consumer smartphone by injecting encoded time-series windows as embeddings instead of tokenizing time series as text. A numerical parity protocol verifies identical greedy-decoded token sequences between the Swift port and a Python reference using Apple's MLX machine-learning framework. Answers also match the PyTorch reference, although rationale wording can differ. An on-device context-economics benchmark compares the soft-prompt path with a text-serialized baseline on the same backbone. In our benchmark format, soft-prompting uses approximately 12 times fewer time-series tokens and reduces prompt-prefill cost. PocketTSLM processes 30-second Sleep-EDF electroencephalography clips for sleep staging and a 1,000-sample (10 s) 12-lead electrocardiogram clip from ECG-QA. In comparison, text-serialized baselines exceed the phone's context or memory budget at longer time-series windows. All inference runs entirely on the device, with no health data transmitted during inference.

cs.LG↗
arXiv2026

Combining LLM and Dynamic Analysis for Universal Testing of Static Application Security Testing Tools

Static application security testing (SAST) tools help detect potential software vulnerabilities, but they themselves contain faults undermining reliability and usability. Automated testing can address these issues, yet current techniques suffer from limited program transformation diversity and lack mechanisms to verify generated variants, leading to reduced fault detection and high manual effort. We present LLM4SAST, the first universal framework detecting faults in SAST tools using large language models and dynamic feedback. Our key insight is leveraging LLMs to understand and transform code in diverse programming languages. We design novel metamorphic relations with transformations to guide variant generation, using an LLM to produce multi-language variants that satisfy these relations. To reduce erroneous variants and identify unique faults, we propose two heuristics: dynamic-feedback-driven selection and program-differencing-based clustering. For each metamorphic relation, a report-centric test oracle detects variants that fail tests and identifies underlying faults by differentially analyzing SAST reports of the seed program and its variants. LLM4SAST identified 66 unique faults across six SAST tools, with 32 confirmed or fixed after being reported to developers, demonstrating effectiveness. The proposed heuristics eliminate 47.8% of false positives, and LLM4SAST detects 3.4 times more faults while requiring only 39.8% of the variants compared to existing approaches.

cs.CR↗
arXiv2026

Transdimensional Chern insulators in rhombohedral multilayer graphene superlattices

Recent experiments on lightly electron-doped rhombohedral multilayer graphene (RMG) in the "transdimensional" regime revealed anomalous Hall hysteresis under both perpendicular and in-plane magnetic fields, pointing to the emergence of transdimensional orbital magnetism in an interaction-driven "Fermi-lune" metal. Here we demonstrate that, by coupling to the moiré potential, this transdimensional orbital ferromagnetic metal can evolve into quantized Chern insulator phases in twisted RMG/hBN moiré superlattices. Using unrestricted Hartree-Fock calculations, we find that interaction-driven reconstruction isolates a spin-valley-polarized $|C| = 1$ Chern band, leading to a "transdimensional Chern insulator" (TDCI) state exhibiting both in-plane and out-of-plane orbital magnetizations. A minimal two-band model is introduced to describe the unique properties of TDCI, which unveils the intimate connection between non-reciprocal velocity and in-plane orbital magnetization. We further analyze the collective excitations of TDCIs, and identify soft charge and spin modes that may lead to density wave instabilities in certain cases. An in-plane magnetic field can harden the soft density-wave mode and may drive a transition from density-wave state to TDCI. Finally, when the $|C| = 1$ band is fractionally filled, the many-body ground state may be a fractional Chern insulator with broken rotation and mirror symmetries.

cond-mat.mes-hall↗
arXiv2026

First Mathematical Runtime Analyses of Evolutionary Multitask Multi-Objective Optimization

To facilitate the solving of complex problems, evolutionary multitask optimization is regarded as a promising technique due to its capacity for cross-task knowledge transfer. In contrast to empirical evidences, no mathematical runtime support exists for this technique within the domain of multi-objective optimization. In this work, we aim to fill this important gap. We analyze how the representative multiobjective multifactorial evolutionary algorithm (MO-MFEA) covers the Pareto fronts of two bi-objective problems simultaneously. For the two-task problem of maximizing the classic bi-objective OMM and the multimodal OJZJ benchmarks, the MO-MFEA requires only an expected runtime of $O(n^2 \log n)$---the same as optimizing the simple OMM alone---yet achieves a speedup by a factor of $Θ(n^{k-1}/\log n)$ compared to optimizing OJZJ alone (with gap size $k$). Similarly, for the task pair comprising the classic LOTZ and the multimodal DLTB benchmarks, a speedup of $Θ(n)$ is attained over optimizing DLTB alone, while the runtime matches that of optimizing LOTZ alone.

cs.NE↗
arXiv2026

Spectrum Is Not Enough: A Random-Matrix Account of Data-Free Preconditioning for Differentially Private Learning

Data-free preconditioning builds the Kronecker factors of KFAC from synthetic probes instead of private data. DP-KFC validates a probe by matching the eigenvalue decay of the private factors; we turn this criterion into a random-matrix account of what a probe can reconstruct. A single dichotomy organizes it: in the factorization $ A \otimes G $, the error-signal factor $ G $ and the architecture-determined bulk of the activation factor $ A $ are reconstructible without data, while only the data-induced spikes of $ A $ are not. We prove the $ G $ half is data-insensitive up to an explicit error-energy bound, with zero violations over a full training trajectory across four modalities and a natural-image benchmark. We locate the failure boundary at a Baik-Ben Arous-Péché threshold: a bulk-only probe suffers an exact damage $s^2/(1-s^2)$ from a data-induced spike, which becomes invisible once it exceeds $(d/n)^{1/4}$, confirmed on an injected-spike surrogate. Holding the probe spectrum fixed and varying only its eigenbasis, we show that spectral-shape matching is neither sufficient nor the operative object: a probe with a near-exact matched spectrum (log-ratio dispersion $0.02$) still inflates the residual $2 \times $-$90 \times $ when its eigenvectors are random. A closed-form whitened-volume law explains this and predicts the within-spectrum ordering of the critical volume, though not its cross-dataset magnitude. Downstream, per-layer tuning does not transfer to accuracy and the headroom is a small oracle-minus-white gap. Under a privacy mechanism the source ordering survives while learning does, then inverts at a noise cross-over where the flat probe overtakes the matched ones.

cs.LG↗
arXiv2026

PSA: A Pruning-Friendly Sparse-Aware Accelerator for Efficient Sparse Convolution in SNNs

Spiking neural networks (SNNs) are promising for energy-efficient artificial intelligence due to their event-driven sparse computation, and neuron pruning can further reduce redundant operations. However, in practical hardware accelerators, spike sparsity and pruning-induced sparsity produce highly irregular sparse workloads that are difficult to execute efficiently. Existing approaches do not jointly support sparse-input execution, low-overhead sparse convolution lowering and pruning-friendly execution. This paper presents PSA, a Pruning-friendly Sparse-aware Accelerator for efficient sparse convolution in SNNs. PSA features a Sparse-Aware Merge-Accumulation (SAMA) dataflow that aligns sparse spike indices and converts irregular activations into continuous computation streams for efficient sparse convolution execution. It further introduces a Sparse Implicit Group-wise im2col (SIG-im2col) convolution lowering mechanism that converts sparse convolution into SpMM on-chip with low mapping cost and memory overhead. In addition, PSA integrates a lightweight pruning-friendly gating mechanism to directly support pruned SNNs. Results on ResNet-18/CIFAR-10 show that PSA achieves a 4.45x improvement in energy efficiency and an 18.52x reduction in energy-delay product (EDP) over the dense baseline.

cs.AR↗
arXiv2026

Money-Mule Detection as a Relational Learning Problem: A Study on the RBIH NFPC Dataset

Money-mule networks move criminal funds through many small transfers, each of which looks ordinary on its own. What marks an account as a mule lies outside any single transfer: in the counterparties and addresses its network shares, and in the shape of its activity over time. Production systems remain largely rule-based and see neither, while most research classifies transactions rather than accounts, recovering account-level risk only by aggregating afterwards. We instead score the account directly, reading the bank's relational schema as a heterogeneous graph in which rows are nodes and foreign keys are edges. Each account is scored from a multi-hop neighbourhood using per-column encoders, a relative-time encoding and HeteroGraphSAGE message passing, with that neighbourhood restricted to rows predating the account's own pre-flag cutoff so that no model sees evidence recorded only after the account was flagged. On a five-year Reserve Bank Innovation Hub (RBIH) retail-banking dataset of 400 million transactions and 94{,}063 labelled accounts, 2.6\% of them confirmed mules, we compare against RelGNN, XGBoost, LightGBM and CatBoost trained on identical labels, splits and cutoffs. Our detector recovers nine in ten held-out mules at 83\% precision and a 0.5\% false-alarm rate, finding 115 more mules than the strongest tree ensemble given the same evidence. At the tuned thresholds the gap is mostly in recall: XGBoost and LightGBM lead across the first hundred alerts, but only the graph models hold their precision deep into the queue. Since the graph is the schema itself and every score is dated per account, the approach deploys on a bank's existing database without a custom feature graph build.

cs.AI↗
arXiv2026

RAT-EC: Introducing a New Open-Source 3D Volume-Integral Eddy-Current Module

We present RAT-EC, an open-source volume-integral eddy-current module for unstructured meshes, driven in the time domain by coils of arbitrary winding geometry. It is the first openly available code of this kind and part of the open-source RAT suite. Because only the conductor is meshed, neither the surrounding air nor an artificial outer boundary has to be modelled, which simplifies model preparation and brings problems with thin gaps and intricate air regions within reach of standard hardware. The formulation combines a semi-analytical evaluation of the singular and near-singular inductance terms, a multi-level fast multipole method that accelerates both the far-field inductance action and the coil field, topology-aware constraints for multiply connected conductors, and a matrix-free, GPU-accelerated time integration. The implementation is verified against closed-form solutions for a conducting sphere and a conducting ring, and is validated on the TEAM-7 benchmark. At 50 Hz, RAT-EC reproduces the measured field as accurately as finite-element solutions. At 200 Hz, it agrees with a converged finite-element model, but both deviate from published measurements, which suggests that this benchmark point should be re-examined. RAT-EC provides an open platform for volume-integral eddy-current simulation and method development.

physics.comp-ph↗
arXiv2026

When CONTRIBUTING Files Fall Short: Revising OSS Contribution Guidelines with Repository Evidence

Contribution guidelines help newcomers participate in open-source software projects, yet often contain incomplete or outdated information. Existing studies characterize newcomers' information needs and documentation gaps, but revising these guidelines requires recovering and validating project-specific knowledge from scattered sources. We present ContribAlign, a framework for automatically revising CONTRIBUTING files through category-guided evidence integration. A qualitative analysis yields a taxonomy of 30 fine-grained guidance types across six contribution categories. ContribAlign uses a classifier to diagnose coverage of the six categories in the original guidance, then uses the taxonomy to collect and organize project-specific evidence. It filters detected conflicts and uses a large language model to generate category-level revisions. The classifier achieves a weighted F1 score of 0.764, improving on the existing classifier by 0.156, while the verifier detects factual conflicts with 97.3% precision and 91.5% estimated recall. Human evaluation shows that ContribAlign achieves the highest overall quality among the original guidelines and three baselines, with significant improvements in usefulness, beginner-friendliness, and completeness. These results suggest that ContribAlign can help maintain more accurate and useful contribution guidelines.

cs.SE↗