arXiv:2607.08785v1 Announce Type: new
Abstract: High-quality data drives machine learning advances across industries. Recognizing the value of data, data transactions are increasingly common, giving rise to many data marketplaces, e.g., AWS Marketplace, Databricks, and Datarade. However, determining the appropriate prices for data products remains a significant challenge due to the unique properties of data products. Traditional pricing methods in economics can be categorized into the cost approach, the income approach, and the sales comparison approach. The cost approach fails in data pricing due to near-zero marginal cost from data replication, and the income approach fails due to inherently unpredictable data revenue. The sales comparison approach remains viable, yet its application is hindered by the absence of standardized pricing benchmarks for data products across marketplaces. To address this challenge, we introduce \texttt{DaDaDa}, the first dataset for data product pricing, containing metadata for 16,147 data products from 9 major data marketplaces worldwide. \texttt{DaDaDa} enables the training of pricing models, thereby establishing price benchmarks for new data products. In addition, \texttt{DaDaDa} can be utilized for other important tasks in data markets, such as data product classification and retrieval. Experiments and a retrieval prototype demonstrate the effectiveness of \texttt{DaDaDa} for pricing, classification, and retrieval of data products. The dataset and code are available at https://github.com/ZJU-DIVER/DaDaDa.
Science Journals
arXiv:2607.08946v1 Announce Type: new
Abstract: A transformer can be built from operators that are legible by construction -- bounded, named units that read as fuzzy set operations rather than dense activations -- but legibility must be pressed for during training, and the pressure has a failure mode. A crispness penalty meant to sharpen a bounded operator into a decisive detector instead collapses it into a dead constant. An identity, E[v(1-v)] = mu(1-mu) - var, shows why -- the penalty is a variance-minimizer blind to the difference between a live detector and a constant -- and names the fix: a per-channel variance floor, the target legibility metric written as a loss, which recovers both legibility and quality. A learned per-unit fraction then retires the hand-set reserved-GELU partition of prior work: given the choice the model keeps no unit as pure GELU and routes 87% of its load-bearing computation through crisp operators. The result is the most legible transformer we have built -- 78% of its feed-forward operands and 50% of its attention value channels are crisp-and-contextual detectors, and per-head legibility rises from 18% in shallow layers to 78% in deep ones. Read in the correct rotated per-layer frame, these units separate a clean detection (what a unit responds to) from a harder naming (what its output decodes to); and because the objective makes each unit crisp and sparse, edits to them are far more local -- 50-184x in the deep layers where the edit sites concentrate -- and can target explicit conjunctions a single neuron cannot express. Finally, a between-unit decorrelation pressure exposes a legibility dial: it trades a circuit's reuse for independence at no quality cost, turning concepts into single, surgically editable units and a prediction into a short explanation read off a handful of named operations. Quality holds at parity with a conventional baseline throughout.
arXiv:2607.09071v1 Announce Type: new
Abstract: Multiscale fluid flows often contain localized flow structures, such as viscous shock layers, wet-dry fronts, steady viscous wakes, decaying vortical structures, and vortex-shedding patterns, whose accurate prediction requires the simultaneous preservation of global conservation trends and small-scale gradients. This study examines these flow-physics requirements through a physics-informed wavelet-Fourier (PIWF) representation for multiscale fluid dynamics. Instead of relying on a single monolithic neural approximator, the formulation separates two complementary components of the flow field within a physics-informed neural representation: long-range coherent modes through a Fourier-basis branch and localized steep-gradient or vortical features through a compactly supported wavelet branch. The outputs are fused with a residual multilayer perceptron using channel attention, and the governing equations, initial conditions, and boundary conditions are imposed directly through the physics-informed loss. The model is assessed on five canonical fluid-dynamics problems: Burgers' equation, the shallow water equations, Kovasznay flow, Taylor--Green vortex flow, and two-dimensional cylinder wake flow. The results show that PIWF improves the resolution of shock-like gradients, wet--dry interfaces, steady wake fields, decaying vortical structures, vorticity extrema, and broadband wake spectra relative to standard physics-informed neural networks and physics-informed Kolmogorov--Arnold networks. These findings indicate that a wavelet-Fourier physics-informed representation can provide a useful route for analyzing multiscale flow phenomena when high-fidelity interior reference data are limited or unavailable.
arXiv:2607.09375v1 Announce Type: new
Abstract: We present Mach-Mind-4-Flash, a 35B-parameter Mixture-of-Experts (MoE) agentic model with 3B activated parameters. Through post-training optimization alone without scaling pre-training compute, the model achieves performance on par with or surpassing that of 100B-parameter-class models. By introducing scalable agentic interaction environments for large-scale reinforcement learning, the model attains significant performance gains on real-world application tasks. Our pipeline comprises three stages: (1) a unified RL/OPD training infrastructure with dynamic multi-teacher scheduling and operator-level acceleration, delivering 17\% end-to-end training speedup; (2) multiple domain-specific RL experts trained in parallel across Reasoning, General, and Agent tracks, then fused into a single generalist via Multi-Teacher On-Policy Distillation (MOPD) -- a routed reverse-KL objective that eliminates the see-saw degradation of mixed-reward RL; (3) Hybrid Median-length Policy Optimization (HMPO), a single-stage token-efficiency method that compresses reasoning chains by 19--46\% with $\le$0.7 percentage-point accuracy loss. Mach-Mind-4-Flash scores 92.70 on AIME'26, 82.82 on IFBench, 80.74 on Behavioral-SafetyBench, 75.80 on BFCL-v4, 72.31 on BrowseComp-zh, and 84.20 on ClawBench -- leading or matching models with 10--30$\times$ its activated size at a fraction of the inference cost.
arXiv:2607.09377v1 Announce Type: new
Abstract: The graph homomorphism problem HOM is: given an $n$-vertex source graph $G$ and an $h$-vertex target graph $H$, is there a mapping from $V(G)$ to $V(H)$ that preserves edges? A straightforward brute-force algorithm for HOM has running time $O(2^{n \log h})$ and it is known that, under ETH, there are no $2^{o(n \log h)}$ algorithms. In recent years, less restrictive graph parameters $p$ have been identified that allow one to solve HOM in time $p(H)^{O(n)}$. Examples include treewidth, maximum degree, and track number. On the other hand, it is known that the chromatic number parameter is too small: under ETH, HOM cannot be solved in time $\chi(H)^{O(n)}$.
We study the complexity of HOM in terms of the degeneracy of $H$. This is perhaps the most natural unresolved graph parameter between the known algorithmic and hardness regimes: on the one hand, each of bounded treewidth, bounded maximum degree, and bounded track number implies bounded degeneracy; on the other hand, bounded degeneracy implies bounded chromatic number. Our results show that, at the same time, the influence of degeneracy of $H$ on the complexity of HOM differs significantly from that of the previously studied parameters. We show that, under ETH, there is no $2^{o(degen(H) n)}$ algorithm for any value of $degen(H)$ as a function of $n$. We also show that bounded degeneracy alone does not make target size benign: even targets with $degen(H)$ at most $2$ and quasi-polynomial size force $n^{\Omega(n)}$-scale hardness. Finally, we introduce a no-compression barrier that explains why the known fine-grained lower bounds for sparse $2$-CSP are not tight under ETH. Moreover, it shows that substantially stronger lower bounds for polynomial-target degeneracy are unlikely to follow from standard reductions from sparse $3$-SAT.
arXiv:2607.09220v1 Announce Type: new
Abstract: A non-intrusive optical diagnostic based on Laser-Induced Fluorescence (LIF) is developed for measuring salt mass fraction in highly concentrated aqueous solutions relevant to absorption heat pump applications. The method relies on a systematic redshift of the fluorescence emission spectra of several dyes, likely induced by an increase of ionic strength with salt mass fraction. A two-color ratiometric approach, based on the ratio of fluorescence intensities in two spectral bands, enables robust mass fraction measurements. This study demonstrates that the techniques is not limited to LiBr but also NaCl, relevant for desalination applications. In parallel, fluorescence lifetime measurements using Time-Correlated Single Photon Counting (TCSPC) reveals that fluorescence lifetime is also affected by the LiBr mass fraction, allowing an alternative measurement technique. For both techniques, many dyes are tested and are compared in terms of salt mass fraction sensitivity. Both techniques are validated independently through comparison with shadowgraphbased measurements during evaporation of droplets. Good agreement is obtained in liquid-phase conditions over a wide range of salt mass fractions, confirming the reliability of the proposed diagnostics. Key sources of uncertainty including temperature, reabsorption effects and phase changes has been identified and analyzed. Finally, preliminary results demonstrate the potential extension of the approach toward planar imaging of evaporating droplets and more complex multiphase flows. The methodology provides a basis for future quantitative diagnostics of heat and mass transfer in desorption systems such as falling-films.
arXiv:2607.09225v1 Announce Type: new
Abstract: Recent 3D geometric foundation models, such as VGGT, provide robust feed-forward 3D reconstruction by directly predicting camera poses and 3D scene points from input images. However, their results remain inaccurate, and scaling them to long sequences or large unordered image sets typically requires chunk-wise processing, which can introduce drift and inconsistency. We present Glob3R, a global SfM-style reconstruction built on 3D foundation models. Our key idea is to explicitly optimize feed-forward geometric predictions. To this end, we augment a frozen Pi3X backbone with a lightweight dense matching head that predicts image warps between selected reference frames and neighboring views. These dense warps are converted into sparse but reliable multi-view feature tracks, which provide correspondence constraints for global optimization. We further introduce a keyframe-based sliding-window association strategy that propagates tracks and relative poses across overlapping windows, enabling scalable reconstruction. Finally, we perform global motion averaging and bundle adjustment to refine camera poses, reduce scale inconsistencies, and recover dense scene geometry. Extensive experiments on indoor, outdoor, large-scale driving, and unordered SfM benchmarks demonstrate that Glob3R achieves robust and accurate reconstruction. It consistently improves over feed-forward foundation-model baselines and recent scalable reconstruction methods, while being more robust than classical SfM pipelines. The refined poses also lead to higher-quality neural rendering, validating the benefit of combining foundation-model priors with global geometric optimization. Project page: https://junyuandeng.github.io/Glob3r
arXiv:2607.09453v1 Announce Type: new
Abstract: Intermittent magnetic turbulence, namely the presence of non-ordered and clusterized fields, is a ubiquitous phenomenon in space and astrophysical plasmas. It is currently understood that it plays a crucial role in the dynamics of astrophysical systems at all scales, from influencing the evolution of the cosmos as a whole to governing local particle acceleration. While there is direct evidence of turbulence in the solar wind, and despite progress obtained through multi-wavelength observations, most of our knowledge of it outside the solar system derives from indirect evidence, through modeling. Here we show that magnetic turbulence, that quantitatively matches that measured in space, can be reproduced in the laboratory. Starting from a homogeneous magnetized plasma, we randomly perturb it using a speckled laser beam. Using proton radiography, we can follow the development and quantitatively characterize the produced intermittent turbulence from its inception.
arXiv:2607.09275v1 Announce Type: new
Abstract: Electrical railway harmonic instability issues are common in the high-frequency range. The effective frequency of the traditional converter's small-signal averaging model is below 1/2 switching frequency since the pulse width modulation (PWM) sideband harmonic components are ignored. In this article, the dynamic propagations of perturbation frequency and the generated PWM sideband components are constructed first. Then the locomotive rectifier's multi-frequency input-admittance model is derived appropriately. Afterward, an admittance conversion approach is used to convert the multi-frequency model into the single-input-single-output (SISO) model whereas retaining the sideband frequency couplings. The proposed SISO model is more accurate than the traditional small-signal averaging model in the frequency range higher than 1 / 2 switching frequency. It is found that PWM sideband harmonics dominate the locomotive rectifier's input-admittance characteristic higher than 1 / 2 switching frequency. Finally, based on the proposed model, the influence of different switching frequencies, control bandwidths, and traction network impedance on system harmonic stability is revealed by the hardware-in-the-loop (HIL) results.
arXiv:2607.09480v1 Announce Type: new
Abstract: The human visual system (HVS) employs foveated sampling and eye movements to achieve efficient perception, conserving both metabolic energy and computational resources. Drawing inspiration from this robustness and adaptability, we introduce the Foveated Dynamic Transformer (FDT), a foveation-guided dynamic token-selection architecture that integrates these mechanisms into a vision transformer framework. The FDT exhibits strong resilience to various types of noise and adversarial attacks, despite not being explicitly trained for such challenges. This inherent robustness is achieved through the use of fixation and foveation modules: the fixation module identifies fixation points to filter out irrelevant information, while the foveation module generates foveated embeddings with multi-scale information. At the 50% fixation-budget setting, FDT achieves higher accuracy than DeiT-S (81.9% vs. 80.9%) while reducing multiply-accumulate operations by 34.57%, highlighting one operating point on its accuracy-efficiency trade-off. These attributes position FDT as an HVS-inspired step toward artificial neural networks that combine adaptive computation with improved resilience.
arXiv:2607.09487v1 Announce Type: new
Abstract: Within-class variance in language-model representations is commonly read as incomplete neural collapse. We argue it is allocated information storage, and that the allocation obeys a law. A one-line centering identity voids a family of simplex equiangular-tight-frame claims, including our own earlier ones; in dimensionless variance shares across 14 models, macro-category structure carries only 4-12% of representational variance and within-token context carries 79-91%, stable across a 100x parameter range. On the theory side, token-level weight decay penalizes a category in proportion to its type count, not its occurrence mass, reducing next-token prediction to an imbalanced K-class problem whose optimum orders category norms by type count. A converse floor, proved for binary categories, forces within-category dispersion to be at least proportional to the conditional mutual information I(token; context | category). The law holds: identity dispersion, not total variance, tracks this information across every tested model and partition, under a model-free estimate and even across models, where one model's information predicts another's dispersion; and over pretraining the category share overshoots, decays, and partially recovers, because the information it must carry never left.
arXiv:2607.09515v1 Announce Type: new
Abstract: Robotic manipulation tasks often require simultaneous reasoning over motion and contact forces, yet most Learning from Demonstration (LfD) methods model only spatial trajectories and neglect force interactions with the environment. This limitation reduces robustness and can lead to unsafe or inconsistent task reproduction in force-constrained settings. We propose a novel one-shot multimodal LfD framework for the segmentation, encoding, and reproduction of force-inclusive demonstrations. First, we introduce a multimodal probabilistic segmentation method that adaptively weighs spatial and force modalities over time, enabling the automatic extraction of force-aware motion primitives. Second, we extend the elastic maps representation to incorporate external force constraints during skill encoding and formulate a convex optimization procedure for learning force-consistent trajectory models. The resulting skills reproduce both motion and contact characteristics from a single demonstration while promoting safer execution by accounting for demonstrated force profiles. We validate our approach on five real-world manipulation tasks across two distinct force-sensing configurations: wrist force sensing on a UR5e with a Robotiq 2f-85 gripper and finger force sensing on a Kinova Gen3 with an Openhand Model O gripper. Experimental results demonstrate robust multimodal segmentation, accurate force-aware reproduction, and cross-platform generality.
arXiv:2607.09529v1 Announce Type: new
Abstract: Food formulation requires balancing taste, nutrition, sustainability, and cost. Traditionally, new foods have emerged through empirical experimentation, expert intuition, and iterative refinement. Artificial intelligence is advancing rapidly across food science, yet most applications remain isolated prediction and optimization tasks rather than parts of a broader scientific framework. Here we define a unified framework for the generative science of food formulation, in which digital food representations enable artificial intelligence to predict, discover, generate, organize, simulate, and optimize. We illustrate this framework through sustainability and nutrition, where generative artificial intelligence transforms environmental and nutritional metrics from post hoc evaluation criteria into explicit design objectives. Finally, we identify the data, models, benchmarks, and automation that will establish computational food design as a rigorous scientific discipline. Together, these advances are transforming food formulation into a generative science.
arXiv:2607.09534v1 Announce Type: new
Abstract: Processes with indefinite causal orders (ICOs), such as the quantum switch, are higher-order quantum processes that superpose the order in which quantum operations are performed. Such coherent control yields computational advantages but is not faithfully captured by existing quantum programming languages: either they are restricted to the unitary case, and thus cannot combine ICOs with measurement, or they treat coherent control nonlinearly. In both cases, they do not realize the full computational power of ICOs. We introduce a higher-order quantum functional language that supports general quantum computation, not merely the permutation of channels, and whose linear type system allows quantum control to be well-defined beyond the unitary case, on arbitrary quantum channels. We equip this language with a small-step operational semantics that synchronizes measurement outcomes across superposed branches, using device references and a memory function. We also give a denotational semantics by means of completely positive maps. With linearity as the only constraint, some well-typed terms would denote unphysical maps. We therefore impose a typing discipline that goes beyond linearity, and interpret programs in the causal category Caus[CPM], under which every well-typed program is physically meaningful, a property that can be checked statically and efficiently. We prove soundness, and study the language's expressive power: it can express every quantum channel at first order, and at second order a large subclass of the so-called quantum circuits with quantum control (QC-QCs), containing the quantum switch. Last but not least, we show that this language is well-designed enough to be extended to the nonlinear setting with recursion.
arXiv:2607.09545v1 Announce Type: new
Abstract: This paper investigates the Gaussian-quadratic lossy compression with arbitrary source length under individual distortion constraints. The rate-distortion function (RDF) is lower-bounded by a Hadamard inequality-based rate, which is tight if and only if the semidefinite condition (SDC) holds. Otherwise, this bound becomes loose, and analytical results are lacking. Moreover, the fundamental quantitative relationship between source correlations and the RDF remains incomplete. In this paper, we provide new theoretical results under different source covariance matrices and distortion constraints. First, under arbitrary covariance and distortion constraints, we obtain the spectral properties of the optimal source reconstruction achieving the RDF, and a stronger scalar inequality version of the SDC. We propose a class of source covariance matrices based on hierarchical correlations and show that studying the two-type correlation (2-TC) model is sufficient to establish the analytical foundation for the broader class. Under this covariance, we obtain the RDF with source correlations explicitly incorporated when the SDC holds, and analyze the SDC from the perspectives of distortion constraints and source correlations. Next, under the 2-TC covariance and two-type distortion (2-TD) constraints, we establish the complete RDFs over seven regions on a distortion plane, with the optimal distortion (rate) allocations determined in each region. It is revealed that the essence of pursuing the complete RDF lies in thoroughly analyzing the correlations between the optimal distortions. Finally, under isotropic correlation and identical constraints, we provide the per-component compression rate and show that exploiting correlations can significantly reduce compression costs.
arXiv:2607.09552v1 Announce Type: new
Abstract: In this Letter, local time stepping (LTS) is incorporated into the unified gas-kinetic wave-particle (UGKWP) method for steady multiscale flow simulation. It accelerates convergence step by a factor of $3.8\times$--$20\times$ and reduces wall-clock time by up to $21\times$ relative to global time stepping (GTS). A rigorous analysis of the particle flux under LTS identifies that fixed per-cell as $\Delta t_i$ is a sufficient condition for the time-averaged flux balance. This condition has not been stated in prior particle-based LTS work, where $\Delta t_i$ varies in time and the flux balance is therefore not guaranteed. Together with proportional rescaling of particle mass and free transport time at cell interfaces, the fixed-$\Delta t_i$ condition yields a conservative framework with no free parameters. The UGKWP-LTS method is validated on cylinder and flat-plate benchmarks that possess multiscale flow features.
arXiv:2607.06766v2 Announce Type: replace
Abstract: At Amazon Prime Video, we face the critical operational challenge of managing code deployments during live events and rapid feature releases without causing service outages. Current change control approaches use blanket deployment freezes that block all changes regardless of risk, creating significant developer toil. While prior research has explored risky change predictors, these rely on developer-specific metadata or extensive historical data, raising privacy concerns and limiting applicability to new projects. We introduce a framework centered on diff-aware features, characteristics derived directly from code modifications. Our key contribution is the systematic identification of which quantitative metrics (code-level and change-level metrics) and qualitative indicators (coding style violations, change type classification) are necessary for risk prediction. We employ LLMs as multi-language feature extractors, demonstrating their effectiveness for code analysis beyond generation tasks and eliminating the need for language-specific tooling. We evaluated our framework on two datasets: Prime Video's production environment and the public ApacheJIT dataset. Our best-performing model achieves an average recall of 0.83 and F1 score of 0.81 across both datasets for detecting risky code changes. Notably, ablation analysis reveals that change-level volume metrics (e.g., lines added/deleted) are noisy predictors, while structural code complexity provides a substantially stronger risk signal. These results demonstrate that thoughtful feature curation enables effective change risk assessment across different programming languages and organizational contexts while avoiding privacy concerns.
arXiv:2607.09559v1 Announce Type: new
Abstract: Molecular polaritons are hybrid light--matter quasiparticles whose collective character is often associated with molecular excitations extending over many emitters. However, molecular ensembles are intrinsically disordered and dissipative, and spectrally visible polariton peaks do not necessarily imply delocalized molecular character. Here, we theoretically examine how static energetic disorder and finite cavity and molecular linewidths affect the delocalization of electronic polaritons in cavity--coupled molecular ensembles. Using a disordered Tavis--Cummings model, we show that energetic disorder mixes polariton states with the dark-state manifold, causing a rapid loss of collective molecular character even when polaritonic spectral features remain visible. We quantify this crossover using the molecular participation ratio, a density--matrix--based coherence measure, and an energy--resolved autocorrelation function. In the lossless electronic model, preserving an extended polaritonic molecular component requires the collective Rabi splitting to exceed the disorder width by more than a factor of five, providing a stricter condition than conventional spectroscopic strong coupling. Extending the analysis to a non--Hermitian Hamiltonian shows that cavity--molecule linewidth imbalance further reduces disorder tolerance. The resulting delocalization boundary indicates that preserving an extended molecular polariton component requires a collective Rabi splitting larger than roughly eight times the disorder width plus approximately twice the cavity--molecule linewidth mismatch. These results provide a quantitative criterion for polariton delocalization under disorder and loss and show that disorder, dissipation, and collective coupling must be considered together when assessing whether molecular polaritons remain collectively extended in realistic optical cavities.
arXiv:2607.09095v1 Announce Type: new
Abstract: Procedural generation of music game levels is an exciting yet challenging problem, as levels must translate musical structure into interactive sequences of timed gameplay events. Most existing approaches formulate this task by frame-based representations, dividing audio into uniform time grids and predicting events at each frame. This makes gameplay events implicit across many frames. As a result, it is hard to describe event-level timing relations and longer-range structure found in human-authored levels. We use procedural generation as a practical setting to study how musical cues map to interactive event sequences. Inspired by event-based symbolic music modeling, we propose a token-level sequence formulation that casts level generation as a multimodal sequence-to-sequence problem. Conditioned on an audio excerpt and level metadata, the model generates a token sequence alternating gameplay-event and beat-shift tokens. This explicitly represents actions and their relative timing in beat space. Based on this formulation, we build a Transformer model. It outperforms representative frame-level baselines under event-level evaluation. It also enables systematic analysis of how audio supports rhythm-aligned event prediction beyond metadata conditioning.
TrustX Agent Risk Classification Framework (ARC): Risk-Tiering Internally Created Agentic AI Systems
arXiv:2607.09586v1 Announce Type: new
Abstract: The proliferation of agentic AI systems across enterprise and public-sector contexts has outpaced the capacity of general-purpose AI risk frameworks to classify and govern them. In this paper, we introduce the TrustX Agent Risk Classification Framework, a structured, repeatable instrument that can be applied to seven types of agentic AI systems and is grounded in foundational pre-existing AI governance frameworks. At the core of the framework is a twelve-dimension scoring rubric that robustly quantifies the risk. This rubric is combined with other components, such as the GPA + IAT classification model and the five-level autonomy framework derived from existing literature. These inputs produce a three-tier governance output with mapped control recommendations. A specialised Coding Assistant extension is also included to account for nuances specific to this type of agentic AI system. We then use an illustrative example to show our framework in practice. ARC is intended for AI governance practitioners, risk officers, developers, and regulators, and it will regularly undergo iteration as we continue to expand it and make it more robust. The community can access the interactive framework here: https://arc.responsible.ai/
arXiv:2607.09598v1 Announce Type: new
Abstract: Lightweight speech recognition models are critical for edge deployment, yet highly optimized architectures like Moonshine often fail on morphologically rich, non-Latin languages such as Bengali. This study identifies the root cause of this failure as the model's English-centric byte-level tokenizer, which fragments Bengali words into high-fertility byte chains and triggers catastrophic autoregressive collapse during inference. To resolve this, a novel vocabulary transplantation pipeline is proposed to replace the decoder vocabulary with the native-script BanglaBERT WordPiece vocabulary and resize the corresponding token embedding matrix. Experimental results demonstrate a reduction in token fertility from 9.16 to 1.30. By decreasing autoregressive sequence length by 85.8%, decoding instability is entirely mitigated. When evaluated on the 882-hour Lipi-Ghor dataset, the modified architecture achieves a competitive 21.54% Word Error Rate (WER) and a Real-Time Factor (RTF) of 0.0053. Ultimately, this research provides a scalable, reproducible blueprint for cross-script adaptation of compact ASR models without the need for resource-intensive pre-training.
arXiv:2607.09342v1 Announce Type: new
Abstract: The Irrelevant Vertex Technique is one of the cornerstones of algorithmic graph theory, underlying Robertson and Seymour's algorithm for \textsc{Disjoint Paths} and much of the algorithmic Graph Minors theory. We show that, in the setting of spanning routing, this technique exhibits an exact combinatorial limitation. Unlike classical routing problems, spanning routing is not governed by the number of distinguished vertices but by the way they are distributed throughout the graph. The input is a triple $(G,R,\mathcal{T})$ where $(G,R)$ is an annotated graph and $\mathcal{T}$ is a set of terminal pairs. The goal is to determine if $G$ contains a family of internally disjoint paths connecting the pairs in $\mathcal{T}$ such that the union of the paths spans the set $R$. We identify a new structural parameter of annotated graphs, called $\mathsf{depth}_2$, that measures precisely this phenomenon. Our main result is a complete combinatorial dichotomy: for every red-minor-closed class of annotated graphs, the Irrelevant Vertex Technique applies to \textsc{Spanning Disjoint Paths} \textsl{if and only if} the class has bounded $\mathsf{depth}_2$. Thus $\mathsf{depth}_2$ forms the exact structural boundary between classes where the Robertson-Seymour paradigm survives and those where it breaks down. Our proof combines a new local structure theorem for annotated graphs of bounded $\mathsf{depth}_2$ with a spanning analogue of the celebrated Vital Linkage Theorem. The resulting algorithm solves \textsc{Spanning Disjoint Paths} in time $2^{2^{\mathbf{poly}(k+d)}}\cdot n^2$ where $d$ is the $\mathsf{depth}_2$ of the input instance. We provide matching lower bounds showing that beyond bounded $\mathsf{depth}_2$ no irrelevant-vertex rule can exist, even on planar graphs. In particular, $\mathsf{depth}_2$ is the exact combinatorial barrier for the Irrelevant Vertex Technique under spanning constraints.
arXiv:2607.09366v1 Announce Type: new
Abstract: Program verification is crucial for software correctness, but producing fully verified programs remains difficult in practice. This paper studies whether implementation structure affects automated verifiability when multiple generated programs are intended to satisfy the same task-level semantics.
We present Diversify2Verify, a staged LLM-based pipeline for Why3 that infers representation-specific contracts, generates and tests diverse recursive and imperative array/list implementations, and attempts verification with bounded verifier-guided annotation repair. We also construct a verification-oriented benchmark of 73 tasks over integers, arrays, and lists, yielding 292 implementation variants. Diversify2Verify verifies 96 artifacts initially and 154 after two repair passes, improving artifact-level verification from 32.9% to 52.7%. At the task level, at least one variant verifies for 49 of 73 tasks, a 67.1% success rate. These results show that task-equivalent implementations can differ substantially in verifiability and that implementation diversity helps find verification-friendly artifacts.
arXiv:2607.09602v1 Announce Type: new
Abstract: Reed-Muller (RM) codes have been shown to achieve capacity over a range of channels, and recently proposed projection-aggregation (PA) decoding has been experimentally shown to achieve near-maximum-likelihood decoding performance. These recent achievements motivate theoretical research on PA decoding. In this work, we analyze the density function of the soft output from collapsed projection-aggregation (CPA) decoding for RM codes over the binary-input additive white Gaussian noise (BIAWGN) channel. We prove that soft-decision CPA decoding returns an exact marginal probability and is symmetric. Based on the analysis, we build a density evolution model for CPA decoding. To simplify the density evolution, we approximate the projection and the fast Hadamard transform decoding using hard-decision decoding. Simulation results over the BIAWGN channel show that our proposed density evolution model captures the fast reduction in the mean and the variance of the soft information returned from the CPA decoding, which qualitatively explains the decoding mechanism and the fast convergence speed of the CPA decoding. We perform an asymptotic analysis based on the proposed density evolution, and we show that CPA decoding can achieve a vanishing error probability for RM codes with a vanishing code rate.
arXiv:2607.09130v1 Announce Type: new
Abstract: Mechanical thrombectomy (MT) is a time-critical intervention for acute ischemic stroke; however, access remains limited due to a shortage of neuroradiologists and specialized centers. Reinforcement learning (RL) offers potential to automate endovascular navigation and improve accessibility, yet current models lack standardized frameworks to assess navigation difficulty for model training and evaluation. This study aims to identify vascular metrics associated with navigation difficulty and to develop an automated pipeline for quantitative vascular feature extraction, enabling future complexity grading. Vascular trees were segmented from computed tomography angiograms from 61 patients, and vascular metrics including aortic arch type, presence of bovine arch, vessel length, tortuosity, take-off angle, number of reverse curves, were measured using a custom pipeline. A Soft Actor-Critic RL algorithm was used for 120 s autonomous navigation. Outcomes were analyzed using both mixed effects linear and logistic regression. On the left side, the presence of a bovine arch and aortic arch type II/III increased navigation time by 30.19 s and 37.92 s, respectively, while greater tortuosity (\b{eta} = 118.20) further prolonged the procedure and reduced success probability. On the right side, type II/III arches extended procedure time by 45.94 s, while each additional reverse curve was associated with 3.96 s longer navigation time and lower probability of success. These findings demonstrate for the first time that MT agent navigation difficulty is strongly influenced by vascular geometry. The proposed automated pipeline enables objective and quantitative characterization of vascular features, providing a foundation for future development of standardized complexity grading and RL model evaluation, without aiming to demonstrate clinically generalizable autonomous navigation.