High Energy Astrophysical Phenomena
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 14 September
A crossover transition from hadronic to quark matter during the merger of neutron stars can lead to interesting observational consequences in the emergent gravitational radiation. In particular, the increased pressure in the crossover density region ($2-5$ times the nuclear saturation density) can lead to an extended duration of high-frequency ($\sim 2-3.5$ kHz) gravitational wave emission during the post-merger epoch. We study the universality relations for a variety of equations of state with and without a crossover transition to a quark mater based upon the QHC19 and QHC21 EoSs. We find that the long-duration post-merger gravitational-wave emission is more pronounced in the QHC21 EoS. We then consider universality relations among various observable quantities that have been identified in simulations with hadronic equations of state. We show that there are deviations from the universality relations for some quantities that may ultimately provide an observational signal that such a transition has occurred.
Functional Analysis
Hadamard Rigidity and Sharp Stability in the Completely Bounded Bohnenblust–Hille Inequality
The completely bounded Bohnenblust–Hille inequality of Arunachalam, Dutt, Escudero Gutiérrez and Palazuelos controls coefficient summability with optimal constant one, independently of the ambient dimension. We classify all nonzero equality cases over both the real and complex fields: they are precisely scalar multiples of Hadamard chains supported on equal-sided Cartesian boxes. Thus equality determines both the support and the factorization of the coefficient tensor. For each fixed degree $d\ge2$, we also establish sharp structural stability. A form with completely bounded norm one and deficit $\varepsilon$ lies within $C_d\sqrt\varepsilon$, in the critical coefficient $\ell_{2d/(d+1)}$ norm, of both a flat unimodular path and a normalized unitary chain on the same box. The corresponding unimodular and scaled unitary edges satisfy a matching normalized Frobenius estimate. At zero deficit both approximants coincide with the original Hadamard chain. All constants are independent of the ambient dimension, and the exponent $1/2$ is optimal for both approximations over either field.
Read the full abstract
Plasma Physics
Characteristic Mapping Method for Vlasov-Poisson with BGK-collisions
This work presents the first steps for simulating kinetic plasmas with collisions using the characteristic mapping method (CMM). The CMM is a semi-Lagrangian method that uses the group structure of volume-preserving diffeomorphisms to efficiently represent the flow-maps of transport problems. Individual submaps can be composed to follow the flow backward in time to its initial foot point, numerically establishing the back-to-label symmetry. The novelty of the presented work is the handling of the source term by storing the Duhamel integrals of each time sub-interval, allowing efficient temporal-integration. Furthermore, we use the Lagrangian structure to avoid implicit time-stepping in the hydrodynamic regime, which is known to be stiff. We benchmark our method on the Boltzmann-BGK and Vlasov-Poisson-BGK equations and consider different test cases, the Sod shock tube problem and nonlinear Landau damping for different Knudsen numbers. We show third-order spatial and temporal convergence and illustrate the fine-scale zoom property of CMM for the bump-on-tail instability.
Read the full abstract
Analysis of PDEs
Effective Dissipation and Asymptotic Stability in a Finite-Level Quantum System with Wave-Induced Memory
We study quantum systems modeled by Liouville-Von Neumann equations, with a finite number of energy levels. The system is coupled to a wave equation, intended to reproduce environmental effects on the quantum system. The whole dynamic is energy conservative, but damping effects are expected to be embodied into the energy exchanges with the environment. We analyse how these effects can effectively damp quantum coherences. We exhibit key mechanisms and structural properties for such a damping to arise, with a neat distinction between pure and mixed cases
Read the full abstract
Analysis of PDEs
Endpoint Differentiability Moduli for Fully Nonlinear Elliptic Equations
A classical consequence of Caffarelli's fully nonlinear regularity theory [L. A. Caffarelli, Ann. of Math. (2) 130 (1989), no. 1, 189-213] is that viscosity solutions of uniformly elliptic equations $F(D^2u)=f$, with $f\in L^p$, $p > n$, are locally $C^{1,\gamma}$ for every $\gamma < \min\{\alpha_H,\sigma_p\}$. Here $\alpha_H=\alpha_H(n,\lambda,\Lambda)\in(0,1)$ denotes the universal Hölder exponent for gradient regularity of $F(D^2h)=0$, while $\sigma_p=1-n/p$ is the scaling exponent of the source term. In the source-limited regime $\sigma_p < \alpha_H$, the singularity of $f$ is the decisive obstruction and the endpoint $\gamma=\sigma_p$ is attainable. In the homogeneous-limited regime $\alpha_H\le\sigma_p$, however, classical theory only yields $\gamma < \alpha_H$, leaving the limiting differentiability estimate unquantified. This is the endpoint gap addressed here. When $\alpha_H < \sigma_p$, we prove that solutions admit pointwise Taylor expansions satisfying $|u(x)-u(x_0)-Du(x_0)\cdot(x-x_0)|\lesssim |x-x_0|^{1+\alpha_H}\left(1+\log\frac{1}{|x-x_0|}\right)^m$. Thus the homogeneous differentiability scale is reached up to an explicit logarithmic defect. At the critical threshold $\alpha_H=\sigma_p$, finite logarithmic powers no longer close the iteration; nevertheless, a slower selection of scales yields $|u(x)-u(x_0)-Du(x_0)\cdot(x-x_0)|=O\left(|x-x_0|^{1+\alpha_H}\exp\left(A\sqrt{1+\log\frac{1}{|x-x_0|}}\right)\right)$. Both estimates improve the full family of classical sub-endpoint $C^{1,\gamma}$ bounds, $\gamma < \alpha_H$, by quantifying differentiability at the limiting homogeneous exponent. The proof introduces a new scale-selection mechanism for endpoint Campanato-type recurrences, suggesting a flexible tool whenever the limiting smoothness is dictated by the homogeneous theory itself. We also discuss the role and possible optimality of the resulting logarithmic defects.
Read the full abstract
General Relativity and Quantum Cosmology
General first-order constitutive equations for a relativistic dissipative multi-component fluid in the presence of a weak background electromagnetic field
In this work we establish general constitutive equations for a relativistic plasma composed of an arbitrary number of classical, charged, and chemically non-reacting species in the presence of a background electromagnetic field. The intensity of the electromagnetic force acting on either species is assumed to be comparable with the gradients of the state variables and is thus considered as a first-order driving force for dissipation within the gradient expansion. Assuming that the species share the same temperature and velocity flow when in equilibrium, we determine the entropy production and establish constitutive equations in terms of first-order frame invariant quantities. Conditions on the corresponding transport coefficients are given in order to comply with Onsager's reciprocal relations and the second law of thermodynamics. In particular, we discuss cross effects for the binary mixture which may be relevant for astrophysical and cosmological applications. Finally, we derive a complete set of evolution equations governing the dynamics of a binary mixture propagating on an arbitrary spacetime background.
Read the full abstract
Solar and Stellar Astrophysics
The Post-Outburst Spectrum of L1415-IRS
We present near-infrared spectroscopic observations of L1415-IRS, a Class I type young stellar object that underwent a factor of $>$30 photometric outburst in the early 2000s. Ours is the first reported spectroscopy of this deeply embedded source. Both components of L1415-IRS present continuum plus $H_2$ emission. The southern component has additional weak HI Br$\gamma$ and $CO$ emission. Based upon our spectrum and evidence of a declining lightcurve, we hypothesize that either: the outburst was not large enough and therefore hot enough to produce the FU Ori type absorption spectrum commonly interpreted as an accretion disk atmosphere, or the source has already recovered from FU Ori-type equatorial accretion to re-establish T Tauri-type magnetospheric accretion.
Read the full abstract
Cryptography and Security
Understanding the Usability of Cryptographic Verification Tools
Cryptographic protocol verification tools are widely used to analyze the security of complex protocols, yet how users interact with these tools remains comparatively understudied. We present an exploratory human-centered study of experienced users of Tamarin, ProVerif, and related protocol verifiers. Our survey included researchers, graduate students, and practitioners with hands-on experience using Tamarin, ProVerif, or related tools. The findings reveal usability barriers across the verification workflow, including difficulties debugging non-termination and performance issues, and the lack of systematic methods for validating formal models against real protocols. When proofs fail without concrete attacks, users commonly simplify models, add helper lemmas, and revisit modeling abstractions. Participants also called for actionable diagnostics, clearer explanations of results, visualization, and automation for recurring proof tasks. Our findings suggest that persistent usability challenges arise from the gap between protocol-level reasoning and the verifier's formal model, proof procedures, and diagnostic output. We derive concrete design priorities for improving the accessibility, interpretability, and usability of cryptographic protocol verification tools.
Read the full abstract
Artificial Intelligence
Cross-Anatomy Transfer Versus Sparse Interpolation in Digital-Twin-Oriented Aortic Fluid-Structure Interaction Surrogates
Surrogate credibility for fluid-structure interac- tion (FSI) requires distinguishing transfer across independent anatomies from interpolation within an already sampled surface. Four de-identified human aortic models from the Vascular Model Repository were reconstructed into separate lumen and nominal 1.5-mm wall domains and analyzed under matched first-cycle two-way FSI. A geometry-only LightGBM prior, selected by leave-one-anatomy-out development on three anatomies, was zero-shot evaluated on a fourth, then probed with a post-zero- shot sparse field-completion case study over six targets. Zero-shot transfer was poor across all targets. At a five-percent anchor level (203 anchors, 3,852 evaluation nodes), prior-plus-adaptation reached an oscillatory shear index (OSI) R2 of 0.603. However, same-anchor controls tuned only on the three development anatomies were stronger for several outcomes: inverse-distance weighting reached R2 = 0.829 (OSI), 0.617 (peak von Mises stress), 0.676 (mean stress); radial basis function interpolation reached 0.917, 0.714, 0.778. Sparse within-anatomy labels thus support field completion, but this four-anatomy cohort gives no evidence the cross-anatomy prior adds value beyond direct interpolation. We frame this as a first computational stage toward a measurement-linked digital twin: the surrogate/update layer is evaluated here, while larger cohorts, converged FSI, measurable patient-side inputs, and physics-informed learning remain future work, not a claim of a complete clinical twin. Our code, data and computation files are available at https://github. com/ali-nourbakhsh2005/Aortic-FSI-Sparse-Field-Completion
Read the full abstract
Software Engineering
FairLint-DL: An IDE-Native Tool for Fairness Debugging of Deep Learning Software
Existing fairness analysis tools predominantly operate as post-training evaluation frameworks, requiring practitioners to complete the full model development lifecycle before assessing bias. We present FairLint-DL, a Visual Studio Code extension that implements a shift-left approach to fairness testing by enabling pre-training, IDE-native bias detection directly on tabular datasets. FairLint-DL trains a configurable deep neural network as a proxy model and applies information-theoretic Quantitative Individual Discrimination (QID) metrics. Grounded in Shannon and min-entropy, QID quantifies the causal influence of protected attributes on predictions. The system implements a two-phase gradient-guided search algorithm for discovering discriminatory instances, a causal debugging pipeline that localizes bias to specific network layers and neurons via sensitivity analysis, and dual explainability engines using SHAP and LIME for feature-level attribution. Evaluation on three tabular benchmarks (Adult Census Income, German Credit, and Bank Marketing) reveals fairness concerns that vary widely across datasets: on Adult, 96.0% of analyzed instances exhibit QID above the 0.1-bit significance threshold, with a mean QID of 0.619 bits and a disparate impact ratio of 0.581, violating the four-fifths legal rule. FairLint-DL produces these results within 12 seconds on cached models, demonstrating the feasibility of integrating fairness analysis into the developer workflow without significant overhead.
Read the full abstract
Mesoscale and Nanoscale Physics
Probing proximity-induced superconductivity in bilayer graphene using gate-defined quantum dots
Van der Waals heterostructures offer a direct way of combining two-dimensional (2D) materials with different electronic properties, such as 2D semiconductors, metals, and superconductors, in a single device. Bilayer graphene (BLG) is particularly attractive in this context, as its electrically tunable band gap enables local control of tunnel barriers and quantum dots. Here, we realize an all-2D hybrid platform based on BLG proximitized by superconducting NbSe$_2$. Using local electrostatic gates, we define tunnel barriers and quantum dots at different distances from the lateral superconductor-semiconductor interface. The quantum dots serve as local spectroscopic probes of the proximitized BLG channel segment, forming tunable superconductor-quantum dot-normal conductor junction devices. Coulomb blockade and finite-bias spectroscopy reveal a proximity-induced superconducting gap of up to $80\,\mathrm{\mu eV}$ and allow to track its evolution with increasing distance from the NbSe$_2$ contact. We find that the local density of states remains suppressed over distances exceeding 1 $\mu$m, consistent with superconducting proximity through a highly ballistic BLG channel. Our results show that BLG-superconductor hybrids offer a controllable platform where quantum dots and quantum point contacts can be well combined with superconductivity.
Read the full abstract
Robotics
ConGraspXL: Controllable Constraint-Conditioned Dexterous Grasping Motion Synthesis
Dexterous grasping is usually conducted for specific tasks, leading to heterogeneous constraints such as specific approach directions, desired contact regions, specified wrist trajectories, and functional hand poses. Our previous work, GraspXL, achieves scalable grasping motion synthesis for diverse objects and hand morphologies, while lacking controllability for synthesis under such various task-driven constraints. In this paper, we propose ConGraspXL, which extends GraspXL with controllable constraint-conditioned grasp motion synthesis that accommodates diverse task-driven constraints and their combinations. We introduce a hierarchical constraint formulation, enable flexible constraint composition with a masked residual interface, and improve control precision with dynamic hand centers and feed-forward wrist guidance. Without losing the strong generalization capabilities of GraspXL, ConGraspXL enables precise and flexible controllability for various individual constraints and their combinations, providing a plug-and-play low-level grasp controller for downstream applications such as whole-body grasp completion, functional grasping, and human-motion imitation.
Read the full abstract
Functional Analysis
Transfinite contractive propagation and the ball fixed point property in \texorpdfstring{$C(K)$}{C(K)}
We prove that, for a compact Hausdorff space $K$, the real Banach space $C(K)$ has the ball fixed point property if and only if $K$ is extremally disconnected. The new implication is obtained by constructing a fixed-point-free nonexpansive self-map of the whole closed unit ball whenever $K$ is a compact $F$-space which is not extremally disconnected. The construction uses a transfinite contractive propagation system with two increasing profiles. A successor–limit delay preserves their tail constraint and has no subsolution in the profile domain. Analysis records boundary deficits; positive operators on ordinal $C_0$-spaces and a common center determined by the two tails yield a nonexpansive synthesis satisfying an order domination inequality. The required topological families are obtained from a gap in a maximal Boolean chain in the zero-dimensional case, and from a transfinite extension of signs otherwise. The argument works in ZFC and resolves the question posed by Avilés, Japón, Lennard, Martínez-Cervantes, and Stawski.
Read the full abstract
Machine Learning
Generative models for simulation based filtering: Formulations and Empirical Comparisons
This letter presents a unified formulation and a controlled numerical comparison of generative-model approaches to the nonlinear filtering problem. Under this formulation the analysis step is realized by a transport of the forecast distribution to the posterior, the approaches differing only in how that transport is selected and learned. We derive three new filters, based on stochastic interpolants, their deterministic flow-matching limit, and Schrödinger bridges realized through forward–backward SDEs. We develop a two-stage tuning procedure that separates the training of the generative model from its online refinement. The resulting methods are compared against the optimal transport filter (OTF), the Knothe–Rosenblatt filter (KRF), the sequential importance resampling (SIR) particle filter and the ensemble Kalman filter (EnKF), in terms of accuracy, computational time, and sensitivity to ensemble size and state dimension. The results indicate that every generative filter resolves multimodal posteriors that the EnKF and SIR do not, that no single generative framework dominates, the preferred method being set by the available online budget and ensemble size, and that the filters differ in the regularity of the particle trajectories they produce.
Read the full abstract
Populations and Evolution
Navigating the Delicate Geometry of Beehive Mite Infestation with Optimal Control
The parasitic mite Varroa destructor poses a severe existential threat to global honey bee Apis mellifera populations. In this paper, we present a dynamical systems model of hive-mite interactions incorporating a eusocial Allee effect to evaluate the efficacy of chemical interventions. We partition treatments into “soft” miticides (targeting phoretic mites) and “harsh” miticides (penetrating the capped brood to target reproductive mites). Our bifurcation analysis reveals a fundamental trade-off: while harsh treatments effectively eradicate the protected mite reservoir by shifting the transcritical bifurcation boundary, they impose sublethal toxicity on the bees. We analytically show that exceeding a critical dosage threshold culminates in a catastrophic saddle-node bifurcation that guarantees colony collapse. To navigate this toxicity limit, we formulate an optimal control problem using Pontryagin's Maximum Principle. Numerical solutions reveal that a dynamic “shock and maintain” cocktail strategy optimally balances reservoir clearance with hive viability. Expanding the model to include treatment-resistant strains demonstrates that single-chemical reliance forces the hive onto a highly toxic, nearly unsustainable chemical treadmill. Finally, our mathematical framework yields a striking ecological prediction: because the eradication boundary scales intimately with the colony's carrying capacity, unchecked mite pressure will likely exert evolutionary forces that select for smaller, naturally resistant hives over the massive colonies favored by commercial agriculture.
Read the full abstract
Sound
SongCraft: Unified Song Generation and Editing with Reconstructive Learning
Song generation and editing have mostly been treated as separate tasks. Existing editing methods often require noise injection and regeneration or curated paired training data. We propose a unified approach for song generation and editing based on reconstructive pretraining, in which a model is trained to reconstruct audio from varying numbers of interpretable conditions. With conditions such as text and lyrics, the model learns to generate diverse songs. With dense conditions specifying fine-grained music attributes, the model learns to reconstruct the target and enables editing by modifying any single attribute while keeping others fixed. This leads to SongCraft, a latent flow matching based model trained for both generation and fine-grained editing. To improve song generation quality, we further introduce word-level phoneme alignment that improves pronunciation learning and accelerates convergence, beat conditioning that improves general musicality, and representation alignment on VAE latent space that produces semantically meaningful latents for improved generation quality. Experiments show that SongCraft achieves the lowest word error rate among evaluated song generation baselines while maintaining competitive audio quality. We further show that a single model can support editing of lyrics, vocal melody, beats, and singer identity, and we also study the trade-off between reconstruction quality and editability.
Read the full abstract
Machine Learning
Robust Fault Detection in Mechanical Multimodal Time Series via Self-Supervised Cross-Modal Reconstruction
Fault detection is essential in industrial systems, enabling early identification of abnormal behaviour and improving safety, reliability, and operational efficiency. Modern systems increasingly rely on heterogeneous sensing modalities that capture complementary aspects of the underlying physical process. However, existing data-driven anomaly detection methods often process each modality independently or use simple feature-level fusion, limiting their ability to exploit cross-modal relationships that characterize normal system behaviour. Their performance also commonly assumes similar training and deployment distributions, whereas real-world operation is affected by changing operating conditions, environmental influences, and system degradation that induce distribution shifts and reduce detection performance, especially in unseen regimes. In this work, we propose a multimodal anomaly detection framework based on cross-modal reconstruction of heterogeneous time-series sensor data. Rather than modeling each modality independently, the framework learns system dynamics by reconstructing each modality from the others, thereby exploiting complementary information across modalities. This integrates information across sensing channels without requiring explicit temporal alignment or identical sampling rates, while improving robustness to sensor noise, missing measurements, and modality-specific disturbances. To address distribution shifts during real-world deployment, anomalies are identified using cross-modal reconstruction error and an adaptive test-time thresholding mechanism that adjusts to changing operating conditions. Experiments on three industrial case studies show strong fault detection performance and substantially improved robustness under out-of-distribution conditions, with the largest gains observed in the most challenging operating regimes.
Read the full abstract
Distributed Computing
Cognitive Admission Control: Risk-Conditioned Assurance for Consequential Actions in Agentic Distributed Systems
In agentic distributed systems, an agent may be authorized to mutate external infrastructure while lacking evidence that the mutation is ready to execute. Cognitive Admission Control (CAC) makes this evidence requirement explicit. A policy maps a typed action and its modeled risk to assurance obligations specifying predicates, evidence classes, scope, freshness, and witness-set constraints. A deterministic evaluator distinguishes satisfied, violated, and unresolved obligations; unresolved conditions produce targeted evidence-acquisition requests. Successful admission produces a certificate binding the action, its witness manifest, and dispatch-time guards. We formalize the admission calculus and the assumptions connecting it to mediated execution. The guarantees are policy-relative: physical safety additionally requires sound evidence, an adequate environment model, and preservation of relevant conditions through the effect. A TypeScript prototype is evaluated in 2,730 controlled local trials with independent effect observation and matched fault schedules. Across 390 CAC trials, 120 effects complete without modeled harm and no harmful effects occur. A live-policy baseline achieves the same completion count but admits the constructed correlated-witness failure. Mechanism ablations isolate guard, evidence-class, structural-cut, and remediation behavior. A further 9,000 measurements exercise the complete local dispatch path with persistent replay protection. These results establish tested implementation behaviors and local costs, not production failure rates or comparisons of language-model capability.
Read the full abstract
Computation and Language
Efficient One-to-Many Translation with Joint Multi-Stream Diffusion
One-to-many machine translation (MT) is computationally expensive for autoregressive (AR) systems, which suffer from linear latency scaling with both sequence length and the number of target languages. We explore how diffusion can enable multilingual translation with a discrete diffusion framework that refines all target languages in parallel, achieving sublinear latency scaling with the number of targets, and supports deployment as a single unified model to replace multiple independent systems. Conditioned on a continuous semantic anchor rather than source tokens, our framework supports zero-shot transfer to unseen source languages without retraining, maintaining approximately $75\%$ of its supervised translation quality on zero-shot sources. We investigate the quality-latency frontier and find that with accelerated sampling, it achieves comparable supervised quality to AR baselines with a $2 \times$ speedup and $11.9\%$ better zero-shot BLEU. These results highlight the potential of joint multi-stream diffusion as a practical and flexible alternative for efficient one-to-many translation.
Read the full abstract
Mesoscale and Nanoscale Physics
An improved auxiliary boson method for quantum dots with finite Coulomb interaction U
In this work, we present an improved version of the formulation of the four auxiliary bosons originally proposed by Kotliar and Ruckenstein, adapted to describe a quantum dot in the Kondo regime with finite electron-electron interaction U. The proposed refinement provides a more accurate treatment of strong electronic correlations while preserving the analytical tractability of the original approach, thereby enabling a reliable description of transport properties in nanoscale systems. Benchmark calculations show good agreement with the results of the numerical renormalization group, supporting the validity of the improved formulation. Our approach therefore provides an efficient and robust framework for investigating transport through strongly correlated quantum dots.
Read the full abstract
Computer Vision
Racing in Volume with Flow Ensembles
Streaming 4D reconstruction has been demonstrated only indoors, on dense camera rigs surrounding subjects that move at human pace. Outdoor 4D reconstruction exists but relies either on cameras mounted on the moving vehicle itself, or on limited-coverage arrays observing quasi-static subjects offline. The case that actually matters for spectators is a fast-moving subject, watched from a sparse ring of allocentric cameras, streaming. No method targets this, and no benchmark exists to evaluate one. To this end, we introduce FastFlowGS, a streaming 4D Gaussian Splatting method for reconstructing fast-moving subjects from a small set of fixed external cameras, and Monaco4D, a photorealistic Unreal Engine 5 benchmark for high-speed outdoor reconstruction. FastFlowGS fuses sparse matches, semi-dense tracks, and dense optical flow by lifting each signal to 3D with geometric uncertainty and combining them through a Kalman-style temporal update. Monaco4D provides Formula 1 sequences under varied illumination from trackside, onboard, and drone viewpoints with dense ground truth. On CMU-Panoptic, FastFlowGS exceeds the strongest baseline by 12.6% VMAF at 35% greater efficiency. On Monaco4D, where existing streaming methods degrade severely, it improves dynamic-region PSNR by up to 18.6% with 28.3% lower per-frame optimization time. Dataset and additional details can be found at https://humansensinglab.github.io/monaco4d/.
Read the full abstract
Machine Learning
Agentic Search Spaces for Tabular Machine Learning
Despite the rapid progress of LLM-based agents for planning, code generation, and debugging, their practical value for tabular machine learning remains underexplored. In this paper, we investigate a concrete use case: whether state-of-the-art agentic AI systems can design extended HPO search spaces for established tabular models that outperform the standard search spaces provided by the model authors. Specifically, we represent each tabular model as a modular pipeline covering preprocessing, embeddings, architecture, training, and inference. We then task the agent to propose candidate code implementations for each module and use a classical HPO algorithm to jointly optimize over these candidates and the model's default hyperparameters. Compared with the base HPO spaces, the expanded search spaces improve the performance of nearly every model family across a suite of 45 datasets, with average relative gains of 0.6%, rising to 2.0% on small-to-medium regression datasets. Notably, these gains come at no extra tuning cost: the enlarged spaces outperform the base under the same tuning and ensembling budgets. The gains transfer to the recent TabArena benchmark, where the agentic spaces improve the official Elo scores of four of the five model families and the two strongest agentic ensembles surpass the best AutoGluon ensemble of conventional models. Overall, our study suggests that LLM agents can provide practical value for tabular ML by expanding the design space.
Read the full abstract
Quantum Physics
Geometric View of Iterative Fixed-Node Dynamics
The sign-structure of a quantum many-body ground state is a central quantity in fixed-node approaches to mitigating the Fermion sign problem. For continuum electronic-structure Hamiltonians, the correct sign-structure is sufficient to compute exact ground state properties; on the other hand, lattice fixed-node approaches retain an additional dependence on trial wave-function amplitudes which are improved upon non-optimally while preserving their sign structure. Here we consider an iterative map obtained by repeatedly replacing the fixed-node trial state with the ground state of its associated lattice fixed-node Hamiltonian (independent of whether and how a practical algorithm could implement this iteration). We show the fixed points of this map are eigenstates of the Hamiltonian and reduced Hamiltonian resulting in at most one fixed point in the interior of each sign chamber. We prove that the ground state fixed point is stable and that the boundary of the ground state sign chamber is repulsive. This shows the amplitude dependence beyond having the correct ground state sign-structure disappears under self-consistent iteration. In other sign chambers, energy descent can be obstructed by the imposed sign structure; this tension leads to the iteration driving select amplitudes to zero producing `support collapse' onto chamber boundaries and corresponding to reduced Hamiltonians. Excited-state fixed points are therefore provably unstable and flow toward lower-energy boundary fixed points. We show sign chamber boundaries have a directional stability; boundary points attractive in one direction are repulsive under a sign flip. By revealing the geometry and stability structure of iterative fixed-node dynamics, our results clarify the foundations of a central approach to the Fermion sign problem and motivate potential new algorithmic strategies.
Read the full abstract
Algebraic Topology
On smooth and bundle structures on topological manifolds homotopy equivalent to $S^{4k-1}$-bundles over $S^{4k}$
We first verify the expected calculations of the topological structure set, in the sense of surgery theory, for the total spaces of $S^{4k-1}$-bundles over $S^{4k}$ for $k\ge3$, in which manifolds homotopy equivalent to the total spaces are organized. Next, we investigate the question of which of the elements in these structure sets can be realized as such bundles. Moreover, we study the forgetful map from the smooth version to the topological version of the structure set; we determine its image. All elements in the image turn out to have the underlying manifold such a bundle.
Read the full abstract
Computer Vision
Sequence Recognition in Bharatnatyam dance
Bharatanatyam is the oldest Indian Classical Dance (ICD) which is learned and practiced across India and the world. Adavu is the core of this dance form. There exist 15 Adavus and 58 variations. Each Adavu variation comprises a well-defined set of motions and postures (called dance steps) that occur in a particular order. So, while learning Adavus, students not only learn the dance steps but also take care of its sequence of occurrences. This paper proposed a method to recognize these sequences. In this work, firstly, we recognize the involved Key Postures (KPs) and motions in the Adavu using Convolutional Neural Network (CNN) and Support Vector Machine (SVM), respectively. In this, CNN achieves 99% and SVM's recognition accuracy becomes 84%. Next, we compare these KP and motion sequences with the ground truth to find the best match using the Edit Distance algorithm with an accuracy of 98%. The paper contributes hugely to the state-of-the-art in the form of digital heritage, dance tutoring system, and many more. The paper addresses three novelties; (a) Recognizing the sequences based on the KPs and motions rather than only KPs as reported in the earlier works. (b) The performance of the proposed work is measured by analyzing the prediction time per sequence. We also compare our proposed approach with the previous works that deal with the same problem statement. (c) It tests the scalability of the proposed approach by including all the Adavu variations, unlike the earlier literature, which uses only one/two variations.
Read the full abstract
Artificial Intelligence
BLINDSPOT: A Benchmark for Safety and Refusal Calibration in Long-Horizon Tool-Using Agents
Large language model (LLM) agents increasingly operate over long-horizon interactions involving tool use, persistent state, evolving authorization, and external environment feedback. In such settings, safety failures may emerge only after multiple turns, yet existing evaluations often reduce agent behavior to task or attack success, obscuring whether an agent acts, refuses, or remains appropriately calibrated as the interaction evolves. We introduce Blindspot, a benchmark for trajectory-level safety calibration of long-horizon tool-using agents. Blindspot evaluates complete user-agent-environment trajectories through adaptive adversarial interaction, stateful tool execution, and execution-grounded adjudication. Its current instantiation contains 22 attack families and 35 scenarios across seven domains, yielding more than 2,500 long-horizon trajectories with an average interaction length of 14.7 turns. Each trajectory is assigned one of five outcomes: Safe Completion, Correct Refusal, Unsafe Completion, Over-Refusal, or Indeterminate. Unlike fixed attack datasets, Blindspot is an extensible live-simulation framework in which attacks, scenarios, tools, policies, domains, and agent configurations can be added without redesigning the evaluation pipeline. We evaluate 13 proprietary and open-weight LLMs using eight metrics covering unsafe completion, appropriate refusal, benign utility, over-refusal, repeated-run robustness, and post-refusal failure. Preliminary results reveal substantial differences in safety-utility calibration across models and show that failures can emerge only after several initially safe interaction steps. These findings motivate treating agent safety as a trajectory-level property rather than a single-turn or binary success criterion.
Read the full abstract
Information Retrieval
Evaluating Brand Retrieval and Ranking in Large Language Model Recommendations
Large language models (LLMs) are increasingly used for product recommendation, but evaluating their recommendations presents challenges that differ from conventional information retrieval and recommender systems. LLMs can generate recommendations without an explicit candidate set, and repeated responses to the same query can produce different brands and rankings. We introduce a framework for evaluating open-ended LLM brand recommendations that defines the competitive set independently of model outputs and estimates recommendation prevalence and prominence through repeated sampling. We operationalize these constructs using Brand Recommendation Probability (BRP@$k$) and Mean Reciprocal Rank (MRR@$k$), and apply the framework to six LLMs across five product categories. Category-only queries reveal substantial omission of established brands and limited evidence that recommendation prominence follows conventional brand popularity. Instead, prominence is associated with broader marketplace-visibility signals, particularly search interest and online brand conversation. Needs-based queries show that contextualizing users' goals and constraints changes which brands are retrieved, while diagnostic positioning probes demonstrate that brands omitted from ordinary recommendations can remain conditionally retrievable when distinctive cues are supplied. These findings highlight the need to evaluate LLM recommendation as a stochastic retrieval-and-ranking process rather than from individual generated lists. We provide open-source software and data to support reproducible evaluation of LLM-generated brand recommendations.
Read the full abstract
Data Structures and Algorithms
SETH-based Lower Bound for Dynamic Degeneracy
In this work, we consider the problem of maintaining an approximate value of degeneracy of a given dynamic $n$-vertex graph $G$ updated by edge insertions and deletions. From the work of Christiansen and Rotenberg [ICALP 2022], it follows that one can design a dynamic data structure for this problem with worst-case update time $\text{poly}(d_{\mathrm{max}}, \log n)$ that maintains an integer between $d$ and $2d+3$ where $d$ is the degeneracy of $G$, under the assumption that $d$ never exceeds $d_{\mathrm{max}}$. We complement their result by providing a conditional lower bound: we prove that, unless SETH fails, for any $\varepsilon, \delta > 0$, $k \in \mathbb{N}$, and function $f\colon \mathbb{N}\to \mathbb{N}$, there is no data structure which maintains a $(2-\varepsilon)$-approximation of the degeneracy of $G$ with initialization time $f(d_{\mathrm{max}})\cdot n^k$ and amortized update time $f(d_{\mathrm{max}})\cdot n^{1-\delta}$.
Read the full abstract
Software Engineering
Assurance Envelopes for Autonomous Coding Agents: Minimum-Cost Evidence for Software Change
When a coding agent returns to existing software, it inherits evidence from earlier engineering work: tests, type checks, proofs, static analyses, and traces. Reloading all of it is wasteful, but dropping a piece the change depends on can leave a required property unsupported. Given the properties a change must preserve, its obligations, we ask which least-cost subset of the available evidence re-establishes them, and we call such a subset a task-conditioned assurance envelope. Evidence and the rules that combine it form a typed inference graph; an obligation is met when forward chaining from the selected evidence reaches it, and we validate every selection by that closure rather than by trusting the optimizer. The software-derived graphs in our evaluation come from preserved outcomes of prior AI coding-agent runs; we freeze those artifacts and ask which accumulated evidence should be restored for a later task. Small graphs from Rust, IronBlocks, and Pong outcomes show that the minimum envelope depends on the task, that none may exist when current evidence cannot re-establish a required property, that some properties need several pieces of evidence together, and that expanding the requirements adds evidence rather than replacing it. A prespecified synthetic benchmark of 249 instances characterizes computation: a baseline that discards the 'several pieces together' structure necessarily fails to re-derive them; every completed exact cross-check agreed with the CP-SAT optimizer; and median solve time stayed below 20 ms at 500-evidence graphs, except that graphs with many alternative derivations per target timed out at far smaller sizes, so structure, not raw size, drives difficulty. The contribution is a bounded application of established optimization to selecting assurance context for a software change; discovering the obligations and downstream agent benefit remain open.
Read the full abstract
Artificial Intelligence
CLEAR: Cross-Source Evidence Adjudication for Large Language Models in Medicine
Medical knowledge evolves continuously, whereas the parametric knowledge encoded in large language models (LLMs) is fixed at training time. External retrieval, including retrieval-augmented generation (RAG), can provide access to newly available evidence, but retrieved information may be irrelevant, incomplete, or conflicting. As a result, external retrieval can in turn degrade the factual accuracy and evidence grounding of LLM outputs. To address this challenge, we propose CLEAR, an agentic framework for cross-source evidence adjudication in LLMs in medicine. CLEAR independently generates candidate answers from three complementary pathways—parametric knowledge, locally curated corpora, and dynamically retrieved evidence—reflecting three common sources of information available to LLMs. An aggregation verifier jointly evaluates the candidates, supporting evidence, provenance, and source-quality information to identify agreement and conflict across sources. An adjudication module then determines whether the current conclusion should be preserved or revised through complementary override-guard and challenge-audit mechanisms, while unresolved conflicts trigger targeted follow-up search and re-adjudication.