Quantitative Methods
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
Single-cell studies are typically initiated from predefined research questions, leaving much of the biological information encoded within existing data unexplored. We formalize open-ended discovery as an analytical paradigm, in which data-derived signals are identified before biological context is interrogated and subsequently evaluated according to their potential to justify prospective experimental investment. Here we develop PROSPECTor, an end-to-end framework that searches for reproducible biological structures across conventional expression representations and diverse foundation-model embeddings, translating robust signals into quantitatively testable candidate hypotheses. Projection into unseen datasets then evaluates their generalizability and phenotype association, providing a scalable screen for candidates that warrant prospective validation. Supported signals emerged from different representation spaces and search strategies. PROSPECTor-nominated hypotheses were then examined in independent biological settings: fibroblast extracellular-matrix programmes demonstrated transferability to an independent mouse cohort with an intervention context, while a patient-resolved gastric-cancer T-cell programme recurred across single-cell, bulk and spatial cohorts. PROSPECTor establishes an auditable framework for systematically revisiting single-cell datasets across expanding representation spaces, turning retrospective collections into prospective resources for biological discovery that can motivate new research questions.
Functional Analysis
Sharp embeddings between quasi-Banach Besov spaces and shallow ReLU variation spaces
Let $\mathbb{D}$ be the ridge dictionary generated by $\operatorname{ReLU}^k$ on a bounded Lipschitz domain $\Omega\subset\mathbb{R}^d$. We establish sharp embeddings between Besov spaces and the associated variation space $\mathcal L_1(\mathbb{D})$ in the quasi-Banach range $0<p\leq 1$. Specifically, \[ B^s_{p,q}(\Omega)\hookrightarrow \mathcal L_1(\mathbb{D}) \] when $s\geq k+d/p$ for $0<q\leq 1$, and when $s>k+d/p$ for $1<q\le\infty$. A rescaled-bump construction shows that this smoothness threshold is sharp. Conversely, for $0<p<1$, \[ \mathcal L_1(\mathbb{D})\hookrightarrow B^{k+1}_{p,2}(\Omega), \] and both the smoothness $k+1$ and the fine index $2$ are optimal. The forward embedding converts Besov regularity of functions and solutions of partial differential equations into quantitative finite-width approximation bounds and greedy-algorithm convergence guarantees for shallow $\operatorname{ReLU}^k$ networks. The proofs use Littlewood-Paley localization, Fourier-Radon representations, measure-valued derivatives, and vector-valued singular integrals.
Read the full abstract
Machine Learning
HarmoCore: Functional Latent Diffusion for Sparse Reconstruction of Oscillatory Wave Fields
Reconstructing oscillatory wave fields from scattered sensors is a severely underdetermined inverse problem. Beyond the challenges of general physical-field reconstruction, wave responses are complex-valued, frequency-sensitive, and highly oscillatory, while costly simulation and sensing often leave only extreme-sparse observations. Existing low-rank, operator, and diffusion approaches are largely designed for real-valued, smoother fields; dense pixel-space diffusion is particularly inefficient for oscillatory complex fields and difficult to scale to 3D. We propose HarmoCore, which places a generative prior in a compact, continuous, and structured wave-field latent. HarmoCore represents joint real–imaginary channels with Functional Tucker cores over shared continuous spatial bases, learns a frequency-conditioned core diffusion prior, and performs Diffusion Posterior Sampling directly in core space. At fixed sensor coordinates, the multilinear decoder induces an explicit likelihood guidance operator, avoiding dense pixel-space correction. Optional target-equation residual guidance further promotes physical consistency. Experiments on 2D Helmholtz, 2D synthetic wave fields, and 3D Helmholtz show substantial gains under 1%–2% sensing while remaining practical in three dimensions.
Read the full abstract
Formal Languages and Automata
Two-State Max-Plus Comparison Is Decidable
Daviaud, Guillon, and Merlet proved that comparison of max-plus automata is undecidable under a fixed state bound of 553 and explicitly left the range from 2 to 552 states open. We resolve the two-state endpoint. More strongly, given an arbitrary finite max-plus automaton $A$ and a max-plus automaton $B$ with at most two states, it is decidable whether $[\![A]\!](w)\leq [\![B]\!](w)$ for every word $w$. The structural reason is a one-dimensional projective normal form for two-state dynamics. Outside an effective bounded region, a transition has one of three tail behaviors: it propagates the unbounded projective gap with gap-independent height increment, forgets the gap with gap-independent height increment, or reads the gap magnitude into the height increment and then forgets it. In particular, any transition whose output depends on the unbounded gap necessarily destroys that gap. This yields an exact one-counter realization of $B$. Effective semilinearity of context-free Parikh images then reduces comparison to Presburger arithmetic. As a consequence, two-state max-plus comparison, equivalence, and positivity are decidable.
Read the full abstract
Robotics
ADAPT: Agile Diffusion Action Priors for Robust and Steerable Online Text-Driven Humanoid Control
We present ADAPT, an end-to-end framework for interactive, text-conditioned humanoid whole-body control. Unlike dominant text-to-motion pipelines that generate kinematic motions for a separate tracker, ADAPT solves language control with an end-to-end closed-loop control framework, where the robot must continuously respond to changing commands while maintaining balance, natural motion, and smooth transitions. ADAPT learns a diffusion-based action prior from text-labeled humanoid state-action trajectories, enabling diverse motion skills to be directly executed from language commands. To improve long-horizon robustness and smooth prompt switching, we train a lightweight residual reinforcement learning policy on top of the frozen diffusion controller. We further show that the same diffusion policy can be reused as a steerable text-conditioned motion prior for downstream task adaptation. Experiments demonstrate robust language-grounded skill execution, smooth interactive transitions, and style-preserving downstream control.
Read the full abstract
Cryptography and Security
Automating Static Code Analysis Through CI/CD Pipeline Integration
In the contemporary landscape of software devel-opment, securing sensitive data is paramount to safeguarding organizational reputation, preventing financial losses, and pro-tecting individuals from identity theft. This paper addresses the pervasive challenge of identifying and rectifying security vulnerabilities early in the development process, emphasizing the role of Static Application Security Testing (SAST) tools. While SAST tools play a crucial role in detecting vulnerabilities, widespread adoption has been hindered by usability issues, including high false positive rates and a lack of native pipeline support. This paper proposes a novel, generalized, and automated process for aggregating SAST tool outputs and integrating them into developers' familiar issue-tracking software. The process streamlines the identification and communication of security vulnerabilities during the development lifecycle, facilitating more efficient remediation efforts. We demonstrate the successful implementation of the proposed process with the SonarQube SAST tool in a GitLab-based development environment. Developers were positive about the structured implementation, real-time feedback, and proactive vulnerability management. However, despite some challenges such as a potential learning curve and tradeoffs between secure coding and workflow disruption, the overall positive impact on security awareness and responsiveness suggests that the proposed process holds promise in enhancing the security posture of software development practices
Read the full abstract
Statistics Theory
Robust dimension-free estimation of simple random tensors: optimal guarantees under heavy tails and adversarial contamination
We study robust estimation of simple random tensors of arbitrary order $q\in\mathbb{N}$ under finite-moment assumptions and adversarial contamination. We propose the first robust estimator achieving near-optimal dimension-free statistical rates in this setting. The estimator attains the near-optimal corruption rate whenever $p\ge2q$ moments are finite and continues to provide nontrivial guarantees throughout the weak-moment regime $q\le p\le2q$. Being based on directional trimmed means and minimax aggregation, our estimator is adaptive to $p$ and upper bounds on hypercontractive constants without resorting to interval-intersection procedures. Our analysis extends the trimmed-mean framework underlying recent advances in robust mean and covariance estimation to arbitrary tensor order. In particular, we establish concentration inequalities for higher-order counting and truncated empirical multi-vector product processes. We believe these inequalities could be of independent interest beyond the present application, including algorithmic robust estimation.
Read the full abstract
Chemical Physics
Diagnosing Latent Energy Decomposition in Machine-Learning Interatomic Potentials via Interacting Quantum Atoms
Machine-learning interatomic potentials (MLIPs) can reproduce potential energies and forces accurately, but their internal energy allocation is often difficult to interpret. E3D-IQA is introduced as a diagnostic framework connecting the latent edge-energy representation of an Allegro-type MLIP with Interacting Quantum Atoms (IQA) energy decomposition. The Allegro edge-energy path is retained as a latent pair contribution, while a node-energy path is trained against the IQA intra-atomic energy. IQA interatomic energies are not direct training targets; instead, the learned edge energies are evaluated after training against the IQA pair terms. Tests on H/C/N/O organic reaction structures show that intra-atomic supervision is essential: energy and force training alone does not recover an IQA-like one-body/two-body allocation. With intra-atomic supervision, node energies reproduce IQA intra-atomic terms, and latent edge energies show meaningful correspondence with IQA interatomic terms. Residual errors are concentrated in positive or weak pair interactions, exposing internal allocation failures that remain hidden in total-energy and force metrics. Adding structures labeled only with energies and forces improves transfer to larger molecules and reduces decomposition errors. E3D-IQA therefore provides a route for diagnosing and guiding latent scientific representations using partially labeled quantum-chemical datasets.
Read the full abstract
Differential Geometry
Intrinsic Finite-Step Characterizations of Discrete General Helices
We study polygonal general helices in Euclidean three-space using the turning and signed torsion angles of the discrete Frenet frame. For helices whose axis is not orthogonal to the edge tangents, we prove that the global constant-angle condition is equivalent to the existence of a conserved Frenet-frame vector. On the generic branch, elimination of the auxiliary coefficient yields an intrinsic finite-step compatibility relation involving three consecutive turning angles and two consecutive torsion angles. Complementary phase and linear-subspace formulations cover the antipodal-binormal case. These characterizations reconstruct the helical axis and the helix angle and yield a sharp bound for each turning angle. We also give a spherical formulation through the tangent indicatrix: its vertices lie on a plane section of the unit sphere, which is a small circle in the non-orthogonal case, whereas the orthogonal case is exactly planar. A nonconstant Frenet-data example illustrates the criterion. Finally, for uniform chordal sampling of a smooth curve, the discrete Lancret-type quotient and the reconstructed helical direction converge with second-order accuracy.
Read the full abstract
Algebraic Geometry
Deciding superellipticity and computing the Weierstrass normal form
Let \( \mathcal{S}_{g,n} \subset \mathcal{M}_g \) be the locus of curves of genus \( g \geq 2 \) admitting a model \( y^n = h(x) \) with \( h \) separable; such curves $C$ have a cyclic group \( C_n \leq \operatorname{Aut}(C) \) of order \( n \) with \( C/C_n \cong \mathbb{P}^1 \). % We give an algorithm which, given an absolutely irreducible plane model \( F(x,y) = 0 \) of a curve \( C \) over a field \( k_0 \) of characteristic zero, decides for which \( n \) the curve lies in \( \mathcal{S}_{g,n} \) and returns a model \( y^n = h(x) \) together with the birational transformation to it.
Read the full abstract
Quantum Physics
Parametric amplification of continuous-variable entangled state for loss-tolerant quantum distributed sensing
Quantum metrology exploits quantum states to achieve an estimation sensitivity beyond classical limits. In the continuous-variable (CV) regime, the squeezed state has been used to implement deterministic quantum sensing, but the quantum metrology sensitivity of this state is significantly affected by losses or detection inefficiencies, which restrict its applications. In this work, quantum distributed sensing is proposed using optical parametric amplified multimode entanglement generated from squeezed states. It is found that the sensitivity is robust to loss or detection inefficiency when large-gain optical parametric amplification (OPA) is introduced, where a two-mode Einstein–Podolsky–Rosen-entangled state and a four-mode cluster state are exploited for analysis. The quantum sensitivity is greatly improved compared to that without OPA in almost all loss scenarios. The quantum Fisher matrix is calculated for both states to obtain the optimal bound in comparison with our scheme, and it is found that even with a small or moderate OPA gain, quantum sensing can be improved compared with the traditional scheme. The states are also compared with corresponding single-mode squeezed states, finding the parameter ranges where entangled states perform better. This study provides a method for realizing large-scale quantum metrology in real-world applications despite losses or detection inefficiencies.
Read the full abstract
Statistical Methodology
Deterministic Leave-One-Cluster-Out Cross-Validation for Multilevel Bayesian Structural Equation Models
We introduce a closed-form, refit-free procedure for leave-one-cluster-out (LOCO) cross-validation in multilevel Gaussian Bayesian structural equation models (SEMs), together with predictive scoring of every nested submodel. Conditional independence of clusters given the parameters expresses the cluster-deleted posterior as a functional of the full posterior. The LOCO predictive density is then a harmonic mean of the cluster likelihood, computable from a single fit in INLAvaan, the integrated nested Laplace approximation package for Bayesian SEM. We evaluate the harmonic-mean expectation in closed form through a fully exponential Laplace approximation of the reciprocal cluster likelihood under a Gaussian posterior; candidate structural restrictions follow by Gaussian conditioning of the same Laplace summary. The resulting Taylor elpd (expected log predictive density) scores are fully deterministic, requiring neither refitting nor Monte Carlo sampling, so the variance pathology of the naive harmonic-mean estimator does not arise. A direct-sum decomposition of the compound-symmetric cluster covariance makes the cost independent of cluster size, orders of magnitude below brute-force refitting. We validate against brute-force refits and Markov chain Monte Carlo in simulation, and illustrate the procedure on school-safety climate in PISA 2022 and on 16 candidate structures linking personality and well-being in the MIDUS sibling sample.
Read the full abstract
Robotics
Behavior–Realization Separation for Constrained Physical Human–Robot Interaction
Physical human–robot interaction software often couples desired-behavior specification with constrained realization; we treat these as separate layers. A behavior layer supplies a desired contact-port acceleration $a_k^{\mathrm{id}}=f_\theta(e_k,\dot e_k,F_{h,k})$. A realization layer converts it into constrained robot commands and reports total desired-versus-realized acceleration error instead of hiding it in saturation. A same-objective unconstrained counterfactual separates regularization from constraint intervention, while plant data expose model error. This paper implements a receding-horizon quadratic program realizing memoryless affine behaviors. Changing the behavior modifies objective coefficients through $(C_\theta,G_\theta)$ while the robot-command variable and feasible set remain unchanged. A planar study instantiates impedance and admittance; the same running layer accepts an impedance–admittance–impedance reassignment without reconstruction, under its existing rate limit. On a torque-controlled 7-DOF Franka FR3 in MuJoCo, the runtime freezes task-space dynamics per solve and enforces torque feasibility across its horizon. Under a sustained 20~N push, it holds a slack-relaxed workspace boundary to within approximately 0.1–0.2~mm, versus 4.4~cm (impedance) and 4.7~cm (admittance) overshoot from instantaneous clipping. A derated actuator budget then activates the torque constraint: horizon-wide enforcement keeps its frozen-model plan feasible to $2.1\times10^{-4}$~N$\cdot$m, whereas a first-step-only ablation plans up to 11.329~N$\cdot$m beyond budget; on the executed nonlinear plant, where both share the same local-model error, the gap is smaller but still favors horizon-wide enforcement (0.161 vs.\ 0.380~N$\cdot$m). These results are a focused proof of behavior–realization separation.
Read the full abstract
Differential Geometry
Rigidity of Euclidean Minimal Hypersurfaces under Nonuniform Diagonal Dilations
Let $n\ge3$ and $D_t=\operatorname{diag}(t^{g_1},\ldots,t^{g_n})$ be a positive diagonal dilation family. We study connected embedded Euclidean hypersurfaces whose diagonal images are minimal. The level-set minimality operator splits into coefficients indexed by the pair sums $g_i+g_j$. Under pair-sum nonresonance, minimality at only $\binom n2$ distinct dilation parameters forces all pair coefficients to vanish. A dimension-reduction argument then shows, without any hypothesis on the coordinate components of the normal, that the second fundamental form vanishes identically. This yields an affine characterization. Repeated-weight helicoidal examples in every dimension and a resonant quadratic cone show that curvature cancellation can survive in genuinely nonuniform families. An application gives a finite-output-level rigidity criterion and an explicit representation for weighted-homogeneous production functions with minimal isoquants.
Read the full abstract
Information Retrieval
From Saliency to Discriminability: Rank-Preserving Visual Token Pruning for VLM Rerankers
Large vision-language models used as listwise rerankers must jointly process visual tokens from tens of candidates per query, making token pruning essential for practical deployment. Existing pruning methods retain tokens by attention saliency, yet we show that saliency is systematically misaligned with ranking contribution: visually prominent tokens often capture order-neutral patterns shared across candidates. This mismatch is layer-dependent: saliency becomes informative only where attention is concentrated, and normalized attention entropy diagnoses the reliability shift (Pearson r=0.87). We propose RaDiCal (Rank-Discriminative Calibration), a training-free framework that uses normalized attention entropy to decide when saliency can be trusted, fusing it with an attention-free rank-discriminative prior and selecting pruning layers from the same trust landscape. Across three retrieval benchmarks and multiple VLM architectures, RaDiCal matches Dense MRR@10 on Flickr30K and surpasses it on MSCOCO at a 20% token budget, ranks first among all pruning methods on FashionIQ, and holds within 1.2 pp on Flickr30K and MSCOCO at 10% retention. It cuts FLOPs by 39–45% and delivers 1.28–1.45$\times$ measured speedups across two VLM architectures without dataset-specific retuning.
Read the full abstract
Computer Vision
DGNet: Dual-knowledge Guided Network for Infrared Small Target Detection
InfRared Small Target Detection (IRSTD) is a prominent and challenging task in computer vision. In recent years, text-guided methods have significantly improved detection performance. However, they still suffer from two key limitations. First, a single text description simultaneously modeling both background and target leads to semantic entanglement, which contradicts the objective of background suppression and target enhancement. Second, reliance on image-specific textual prompts (requiring additional external models such as CLIP during inference) results in deployment constraints. To address these issues, we propose a novel Dual-knowledge Guided Network (DGNet) based on multiple generalizable texts. Specifically, we design a Prior-knowledge Wavelet Modulation (PWM) module, which leverages dual textual priors that separately characterize large-scale backgrounds and sparse targets to effectively disentangle and modulate entangled semantics in the frequency domain. Furthermore, we introduce a Consensus-knowledge Directional Alignment (CDA) loss, which models the initial state and the ideal target across samples as `complex background' and `bright target', respectively, thereby constructing a clear and unified directional optimization trajectory for the model. Extensive experiments on three public datasets demonstrate the superior performance of DGNet and the effectiveness of each component. The source code is available at https://github.com/iLearn-Lab/MM26-DGNet.
Read the full abstract
Artificial Intelligence
Triple-Bottom-Line Sustainability of Language Models for Edge AI: A Comparison Between SLMs and Quantized LLMs
Edge-AI model selection is commonly driven by one isolated metric - accuracy, latency, memory, energy, or safety, even though a deployable language model must balance all five. Our work focuses on answering the question whether na- tively trained small language models (SLMs) or large language models (LLMs) compressed through post-training quantization offer the more sustainable edge- deployment trade-off. We introduce a reproducible Holistic Sustainability Score (HSS) organized around the triple bottom line: an economic pillar for capability and systems efficiency, an environmental pillar for operational GPU energy and a social pillar for harmful-prompt robustness. Five BF16 SLMs and five LLMs under different quantization approaches - BF16, INT8, NF4 4-bit, GPTQ 4-bit, and GGUF Q4 produce 30 measured configurations. Capability is assessed on five zero-shot benchmarks; efficiency uses latency, throughput, peak VRAM and energy; and safety is approximated by attack success rate on five harmful prompts. Qwen3-30B-A3B/GGUF Q4 ranks first in the combined pool (93.38), followed by Mistral-Small-24B/GGUF Q4 (92.40), while Phi-4-mini/BF16 is the highest- ranked SLM in that pool (89.49). Thus, the hypothesis that native SLMs must be the most sustainable edge choice is not supported universally; optimized quantized LLMs can win overall, while SLMs remain competitive through lower resource demand. Quantization is a systems-level choice rather than a monotonic precision- efficiency trade-off and HSS remains relative to its comparison pool and proxy definitions.
Read the full abstract
General Relativity and Quantum Cosmology
Perturbations and stability of black holes with static scalar hair in general GLPV theories
We derive the background equations and complete odd- and even-parity quadratic actions for static, spherically symmetric black holes with radial scalar hair in general quartic-quintic Gleyzes-Langlois-Piazza-Vernizzi (GLPV) theories, including Horndeski. On regular, nondegenerate branches, the formulation applies across Killing horizons and yields local no-ghost conditions and radial and angular eikonal characteristics. Tensor modes in both parity sectors share the same squared radial speed. When the radial coordinate is timelike, generic antisymmetric mixing induces nonstandard large-multipole scaling for both even-parity angular branches. Requiring both branches to have finite, nonzero eikonal phase speeds selects the Horndeski-related compatibility condition between the quartic and quintic beyond-Horndeski functions. This condition aligns the covariant degeneracy directions and permits a local disformal map to Horndeski when regular and invertible. For a finite, nonzero scalar kinetic term at the horizon, every nontrivial branch of the exact quadratic GLPV black hole is locally unstable or has a degenerate tensor cone near a simple outer horizon. Apart from identified exceptions, this obstruction extends to general shift- and reflection-symmetric quadratic GLPV theories and the regular Horndeski-related branch of the quartic-quintic class. Finally, for scalar-Gauss-Bonnet black holes with a vanishing horizon scalar kinetic term, we construct analytic power-law quartic beyond-Horndeski deformations whose associated quintic function is fixed by the same condition. At sufficiently small coupling, all local no-ghost and high-frequency gradient-stability conditions hold throughout the exterior. For a linear Gauss-Bonnet coupling, we estimate an interior scale below which the background and stability expansions lose perturbative control; this does not imply a physical instability.
Read the full abstract
Computer Vision
Separating perception from reasoning in vision-language models: a model-free render ceiling for crystal structures
Vision-language models (VLMs) often predict plausible structural labels from images without recovering the three-dimensional geometry behind them. End-to-end scores cannot tell whether an error comes from ambiguous views, lost visual information, or flawed reasoning. Multi-view 3D reconstruction faces its own trade-off: rigid symmetry priors erase genuine distortions, while unconstrained triangulation absorbs measurement outliers. We separate visual perception from structural reasoning through an inspectable 3D geometric intermediate. Its foundation is a model-free render ceiling: an exact reference for what calibrated, species-labeled projections determine, obtained by ruling out projective phantoms. Across a 1,950-structure census, every tested combination of four or five calibrated views uniquely determined the 3D sites from complete, species-labeled centroids. Against this reference, replacing images with exact text coordinates lowers the error rate of all 17 evaluated VLMs on crystals. The three frontier models gain 10.0-19.5 percentage points yet still miss 8.1-21.0% of cases. To recover geometry from imperfect observations, our symmetry-aware split reconstruction jointly estimates a symmetric base, coherent 3D physical displacements, and view-specific 2D corruption, adaptively selecting between atomic and collective distortion modes. Relative to Huber triangulation, it reduces coordinate error by 36.7% under Gaussian noise and 41.2% under corruption. Adaptive mode selection further cuts collective error up to 30.4% relative to atomic split across two external cohorts without sacrificing localized fidelity. Given reconstructed rather than Huber coordinates, downstream models improve. Geometric recovery thus becomes an explicit, testable foundation for structural understanding.
Read the full abstract
Artificial Intelligence
Drift-Aware LLM Routing with Sparse Contexts and Shared Budgets
A multi-model language service must route each request while preserving workload-level budgets for compute, latency, memory, or monetary cost. Two features make this problem materially harder than static model selection. Prompt representations are high dimensional, so only a small subset of embedding directions may predict the incremental value of a model, and both the request mix and the model frontier drift after launches, fine-tunes, quantization changes, and system updates. We formulate nonstationary sparse contextual routing with multiple knapsack constraints and an optional shadow-audit stream that evaluates a small fraction of prompts on several models. We propose Drift-Aware Sparse Routing (DRS). The policy estimates reward and resource use from a rolling audit window, routes using pessimistic reward and optimistic cost estimates, updates resource shadow prices online, and applies a hard meter before commitment. The analysis separates control from statistics. On any event with uniform prediction radii $\{\beta_t\}$, regret against a paced dynamic fluid benchmark is bounded by the sum of the radii, a capacity-buffer term, and an $O(\sqrt{T})$ pacing term. Under a sparse linear model and bounded drift $V_T$, rolling estimation gives \[ \widetilde O\left( T\sqrt{\frac{s}{\rho W}}+WV_T+\sqrt{T} \right), \] where $s$ is sparsity, $\rho$ is the audit rate, and $W$ is the window length. Optimizing $W$ yields the usual stationary $O(\sqrt{sT/\rho})$ rate when $V_T=0$ and a $O(T^{2/3}(s/\rho)^{1/3}V_T^{1/3})$ adaptation term under drift.
Read the full abstract
Computer Vision
Do Satellites See Commuters? A Critical Benchmark of Vision Foundation Models
Satellite foundation models offer a globally available alternative to census data for commuting origin-destination (OD) generation, yet no study has systematically compared encoder paradigms within a single downstream pipeline. We ablate four satellite vision encoders: language-supervised (RemoteCLIP), self-supervised (DINOv3), and geographically grounded (SatCLIP, AlphaEarth) within an identical WeDAN graph diffusion framework across 1,925 US counties, 325 UK districts, and 14 global cities under five random seeds. Three main findings emerge. First, language-supervised features achieve the strongest in-distribution performance (RemoteCLIP CPC 0.602), while geographically grounded encoders transfer more reliably zero-shot: AlphaEarth improves CPC by 33% over RemoteCLIP on UK districts. Second, pretraining corpus scale alone is insufficient: DINOv3, trained on a substantially larger satellite corpus, underperforms RemoteCLIP by 0.091 CPC in-distribution and collapses to CPC 0.022 globally. Third, no encoder transfers usefully to global cities (best CPC 0.122 for RemoteCLIP, 0.022 for DINOv3), confirming cross-continental OD generation remains an open problem. We additionally clarify the semantics of the census noise parameter $\eta$, whose ordering reverses under cross-continental evaluation, a distinction critical to correctly interpreting prior results. Our code is available at: https://github.com/cruiseresearchgroup/satellites-see-commuters.
Read the full abstract
High Energy Physics (Phenomenology)
Mixing-angle sensitivity of inclusive $\Xi_{bc} $ and $\Omega_{bc}$ lifetimes
We study the inclusive lifetimes of $\Xi_{bc}^{+}$, $\Xi_{bc}^{0}$, and $\Omega_{bc}^{0}$ with emphasis on the mixing between the spin-zero and spin-one $bc$-diquark configurations. The matrix elements are evaluated in both the homogeneous bag model (HBM) and the nonrelativistic quark model (NRQM). At the benchmark mixing angles $-23.0^\circ$ for $\Xi_{bc}$ and $-23.5^\circ$ for $\Omega_{bc}$, the corresponding lifetime triplets are $(\tau_{\Xi_{bc}^{+}},\tau_{\Xi_{bc}^{0}},\tau_{\Omega_{bc}^{0}})=(0.421^{+0.098}_{-0.093}{}^{+0.007}_{-0.014},0.075^{+0.035}_{-0.017}{}^{+0.004}_{-0.001},0.223^{+0.074}_{-0.046}{}^{+0.018}_{-0.012})\,\mathrm{ps}$ in HBM and $(\tau_{\Xi_{bc}^{+}},\tau_{\Xi_{bc}^{0}},\tau_{\Omega_{bc}^{0}})=(0.367^{+0.076}_{-0.070}{}^{+0.001}_{-0.005},0.100^{+0.032}_{-0.014}{}^{+0.006}_{-0.001},0.229^{+0.066}_{-0.042}{}^{+0.017}_{-0.011})\,\mathrm{ps}$ in NRQM. The first errors are the fixed-angle uncertainties; the second are the envelopes from varying each mixing angle by $\pm10^\circ$ with all other inputs fixed. Relative to the unmixed spin-one reference points, the $\Xi_{bc}^{0}$ and $\Omega_{bc}$ lifetimes decrease by approximately $17\%$-$19\%$. Neither the minimum nor the maximum of the central lifetime curves occurs at a pure $S_{bc}=0$ or $S_{bc}=1$ configuration; the extrema instead lie at mixed angles with $|\phi_{{\cal B}}|\simeq27^\circ$-$63^\circ$. The two pure-spin cases alone are therefore insufficient to characterize the angular dependence.
Read the full abstract
Robotics
Fleets Need a Context Plane: Rethinking Cooperative Perception for Autonomous Drones
Cooperative perception allows a drone fleet to combine observations from multiple viewpoints. However, existing systems typically fix their feature-sharing policies at design time or adapt to only one context signal. This is a poor fit for aerial fleets, whose missions, bandwidth, formation geometry, and scene coverage can change during flight. We quantify the cost of context-blind sharing on UAV3D by controlling feature exchange at evaluation time using a released DiscoNet checkpoint, without retraining. Mission-aware sharing matches full-sharing accuracy while using only 5-10% of the bytes. The best tested peer selection policy changes with the byte budget, and choosing the wrong policy loses up to 7.7 AP. Moreover, under a constrained budget, two policies with the same full-scene accuracy differ by 5.9 AP within the mission region, showing that multiple context axes must be considered jointly. We therefore propose the context plane, a bounded, structured interface for runtime context. Each drone publishes a descriptor of at most 1 KB at 10 Hz, and lightweight, replaceable policies use the fleet context to decide what each drone computes, shares, and fuses. Existing sharing schemes become fixed policies within this interface. In our ROS 2 prototype on a Jetson AGX Orin, the context plane uses approximately 0.01% of the data-plane bandwidth, and each policy decision takes 0.10 ms. These results show that an explicit context interface can support low-overhead runtime adaptation without modifying or retraining the perception model.
Read the full abstract
Computer Vision
Teaching Vision-Language Models to Use the Scale They Are Given: Label-Free Equivariance Training for Metric Physical Reasoning
Metric questions about video, such as the speed of a moving object, require a vision-language model to convert visual measurements into physical units using a real-world reference supplied in the prompt. We find that current models use this reference only partially. When every world-space quantity in the prompt is multiplied by a common factor, the prompt still describes the same video and the correct answer changes by exactly that factor, but the predictions of eight models change by less, and their accuracy stays concentrated near the scale that the depicted objects usually have. Asked the same physics in a scale-free form, the two models we test recover the closed-form scaling laws on most items, which indicates that the deficit lies in metric grounding and not in knowledge of the physical mechanism. Because the scaling relation is exact, it can serve as supervision without metric annotations. Equivariance Self-Distillation (EquiSD) projects a model's own prediction onto the functions that satisfy this relation and fine-tunes the model on the resulting targets, with one query per training question and no ground truth. Trained on synthetic video only, EquiSD brings a 3B model close to the exact relation on held-out simulated videos, also at scales not seen in training, and improves its accuracy across scales. Without adaptation, it also improves accuracy across scales on the QuantiPhy benchmark, where its gain reaches 93% of that obtained by supervision with exact simulator answers.
Read the full abstract
High Energy Astrophysical Phenomena
Search for Neutrinos from Tidal Disruption Events with IceCube
Tidal disruption events (TDEs) are theorized to produce high-energy neutrinos through photohadronic interactions between accelerated protons and multi-wavelength photons in the accretion disk and outflows. Detecting these neutrinos would provide insight into the dynamics of TDEs. Taking advantage of the recent increase in observed TDEs from wide field-of-view telescopes, we conduct a dedicated search for neutrinos coincident in optical/UV and X-ray wavelengths. We searched for neutrino emission from 89 TDEs selected based on X-ray and optical/UV observations using time-dependent likelihood analysis methods in two parts. First, we searched for emission from individual sources, where we fit the time window of expected neutrino emission. Second, we performed a study of jetted and non-jetted TDE subpopulations using a stacking search with a fixed one year time window. No significant neutrino excess was observed in either search. We set upper limits to the contribution of jetted and non-jetted TDEs detected in optical/UV and X-ray wavelengths to the diffuse astrophysical neutrino flux assuming TDEs are standard candles.
Read the full abstract
Computer Vision
Physically Plausible Video Generation via Visual-Semantic Chain-of-Events Conditioning
Physically Plausible Video Generation (PPVG) seeks to synthesize videos consistent with physical principles, yet remains challenging due to underspecified natural language conditioning. Advanced chain-of-thought (CoT) frameworks augment prompts with physical knowledge. However, such prompts describe physical phenomena holistically, overlooking intermediate states and transition dynamics. In this paper, we reformulate PPVG as event-centric generation by representing physical evolution as a chain of causally connected and physically constrained events. Our framework comprises three key modules: (1) Physics-driven Event Chain Reasoning. This module decomposes physical phenomena into causally connected events represented by evolving scene graphs. Formula-derived physical quantities are bound to relevant objects and interactions, characterizing the direction and magnitude of each event transition. (2) Transition-aware Routed Keyframe Conditioning. This module routes each event to a specialized keyframe synthesis operator for appearance variation or object transformation. Consecutive keyframes are injected as residual guidance during denoising, enabling smooth visual transitions between event-boundary states. (3) Physics-injected Contrastive Semantic Guidance. This module constructs physics-informed positive and counterfactual negative prompts for classifier-free guidance, steering generation toward plausible dynamics and away from physics-violating counterparts. Experiments on PhyGenBench, VideoPhy, PhyWorldBench, and Physics-IQ demonstrate that our framework generates videos with superior physical plausibility across diverse domains.
Read the full abstract
Dynamical Systems
Empirical variational principles for preimage entropies
Preimage entropy measures the complexity generated by the inverse-image structure of a non-invertible dynamical system. For a continuous map $f:X\to X$ on a compact metric space, Hurley's pointwise topological preimage entropies $h_m(f)$ and $h_p(f)$ are natural invariants measuring the complexity of preimage sets. The question of whether they admit unconditional variational principles in terms of suitable measure-theoretic counterparts remains open. In this paper we resolve it by using empirical metric preimage entropies $h^*_{m,\mu}(f)$ and $h^*_{p,\mu}(f)$, defined by restricting preimage fibers to orbit segments whose empirical measures are close to a prescribed invariant measure $\mu$. We prove the variational principles $$ h_m(f)=\sup_{\mu\in\mathcal M_f(X)}h^*_{m,\mu}(f), \qquad h_p(f)=\sup_{\mu\in\mathcal M_f(X)}h^*_{p,\mu}(f) $$ for every continuous map on a compact metric space. We also show that, in general, the set of all invariant measures in these formulas cannot be replaced by the set of ergodic invariant measures. We then compare the empirical entropies with the pointwise metric preimage entropy $h_{m,\mu}(f)$. For every ergodic invariant measure $\mu$, we prove $h^*_{p,\mu}(f)\ge h_{m,\mu}(f)$. Moreover, if $f$ has uniform separation of preimages, then for every ergodic invariant measure $\mu$, $$ h^*_{m,\mu}(f)=h^*_{p,\mu}(f)=h_{m,\mu}(f). $$ Examples show that $h_{m,\mu}(f)$ is not comparable with the empirical quantities in general. We further introduce a resolving-partition property, weaker than uniform separation of preimages, under which the variational principle for $h_m(f)$ and $h_{m,\mu}(f)$ holds. Finally, we establish corresponding variational principles for preimage pressure and give an example showing that uniform separation of preimages does not imply forward expansiveness.
Read the full abstract
Artificial Intelligence
SciTrue: Reliable Scientific Claim Validation with Frontier and Open Language Models at the NTCIR SciClaimEval Task
We describe the SciTrue team's participation in both subtasks of the NTCIR-19 SciClaimEval task~\cite{sciclaimeval}, which asks systems to verify scientific claims against the tables and figures of a paper. Rather than tuning a single model, we benchmark eleven frontier and open multimodal models under one honest, per-sample protocol and combine them with light, transparent post-processing. On the official, blind test leaderboard (Section~\ref{sec:results}), SciTrue placed first by a clear margin in three of the four evidence-category/subtask combinations, and tied for first on the primary metric in the fourth. Three findings explain the result. First, strong instruction-tuned models are already competitive: Claude Opus~4.8 and Gemma-4-31B each exceed the strongest public baseline (o4-mini), and GPT-5.5 and Claude Fable~5 lead both subtasks (97.7 on Subtask~2). Second, the task's pairing structure is the largest lever: a leak-free pair prior that recovers the Supported/Refuted pairing from the claim text alone (a visible field) and assigns Supported to the higher-confidence evidence raises Subtask-1 pair-accuracy from 72.2 to 93.5, far more than any model swap or ensemble weighting. Third, a case-by-case audit finds that most residual errors are visually-undetectable label-mapping swaps or dataset label noise, so measured accuracy understates the true ability and the fixable-by-modeling headroom is small. Controlled fine-tuning, distillation, and agentic consistency-checking support the same conclusions, and we document throughout a measurement leak—label information reaching a system through the packaging of the data rather than its content—in which the released file ordering encodes the label, including one instance that briefly misled our own pipeline.
Read the full abstract
Machine Learning
EEG-AS: Instance-Level Foundation Model Selection for EEG Foundation Models via Behavior Reconstruction
Electroencephalography (EEG) is a non-invasive technique for measuring neural activity and has been widely used in neuroscience applications. Recent advances in EEG foundation models have enabled strong performance across diverse neural decoding tasks. However, no single foundation model consistently performs best across datasets or individual EEG instances, while instance-level model selection remains largely unexplored. To address this limitation, we formulate EEG foundation model selection as an instance-level Algorithm Selection (AS) problem. We propose EEG-AS, an instance-level algorithm selection framework that characterizes each EEG instance using inference-available latent EEG embeddings, handcrafted neurophysiological features, and an anchor foundation model. During training, EEG-AS learns to reconstruct unavailable foundation-model behaviors from privileged prediction tokens conditioned on an anchor foundation model, while during inference it estimates these behaviors without executing the entire model portfolio, enabling efficient selection from seven EEG foundation models. Experiments on seven public EEG benchmarks demonstrate that EEG-AS substantially narrows the gap between the Single Best Solver (SBS) and the oracle upper bound for each instance. These results highlight the effectiveness of instance-level AS for adaptive deployment of EEG foundation models.
Read the full abstract
Artificial Intelligence
Self-Reports Are Not Verification: Environment-Grounded Auditing of LLM Operators in Evolutionary Search
Language model agents increasingly propose actions, observe external feedback, and explain their own behavior. Their confidence and rationales are convenient monitoring signals, but convenience is not verification. We introduce an environment-grounded audit in which every intermediate proposal receives an exact outcome. A language model operates an evolutionary Contexto search whose feedback function assigns every valid guess an exact rank without human annotation. Across 200 runs spanning five configurations and three model families, four reporting configurations produce 12,249 self-reports. We test three assumptions: stated confidence is calibrated, inherited rationales affect later proposals, and fitness-based selection improves report quality. All three fail. Operators overstate top-100 success by factors of 4.8 to 9.3, while calibration and discrimination dissociate across model families. Controlled interventions on 754 inherited rationales bound any measured benefit of the genuine rationale to roughly 250 ranks. Neither fitness-based nor random selection produces a detectable selection differential or parent-to-offspring transmission in report accuracy, despite sharply different search behavior. Agent self-reports should therefore be treated as claims to verify against the environment, not as evidence of their own reliability.