Papers

A new set of jet-medium transport coefficients stemming from jet-medium exchanges involving Glauber quarks encoded in $\hat{\mathcal{F}}_i$ [Phys. Rev. C 111, 054913 (2025), Phys. Rev. C 113, 055207] are obtained by computing the tree-level leading-order $2\to2$ scattering rates and their moments using the approach developed in Refs. [arxiv 2608.17160, arxiv 2608.17161]. Sizeable deviations away from the leading logarithmic dependence of jet-medium transport coefficients and scattering rate are observed here, as previously mentioned in Refs.[arxiv 2608.17160, arxiv 2608.17161]. A closed-form expression for $\hat{\mathcal{F}}_i$ accurate to $\lesssim 10$% or better is obtained, enabling our approach to be used within Monte Carlo simulations of jet-medium interactions. Monte Carlo simulations incorporating $\hat{\mathcal{F}}_i$ are sensitive to flavor hydrodynamization dynamics as the medium created in nucleus-nucleus collisions transitions from the early-time Glasma dynamics to the quark-gluon plasma fluid.

SQL has been augmented with AI operators, enabling modern data analytics platforms to derive insights from both structured and unstructured data. We observe that while current Text-to-SQL systems can successfully generate these AI-augmented queries, reliably evaluating their correctness remains a critical open challenge. Current metrics, which rely on exact query results and deterministic execution, systematically fail against the flexible, non-deterministic outputs of AI operators. In this paper, we formalize these unique evaluation failure modes and introduce a Multilayered Evaluation Framework that decouples deterministic database logic from flexible AI semantics. We test our approach across both industry (BigQuery) and academic (ThalamusDB) systems. We demonstrate that traditional Execution Accuracy severely penalizes valid queries, achieving as low as a 25% detection rate for correct translations. Furthermore, even a state-of-the-art LLM-based autorater falsely rejects 32% of accurate queries due to the complexity of judging both relational and AI components simultaneously. By validating the standard relational logic and the AI operations separately, our framework achieves state-of-the-art overall accuracy across both platforms (up to 97.2%), proposing a reliable standard for benchmarking AI-powered SQL generators.

Generative AI changes the conditions under which graduate mathematics is learned, assessed, written, and defended. The central claim of this paper is that mathematics graduate programs should respond to the moment by clarifying what graduate mathematics education is trying to teach and assess. In most ways, the goals of mathematics education have not changed. Rather, with changing tools it has become more essential than ever to make clear the goals of mathematical training. We use the term mathematical judgment to refer to the capacity to evaluate mathematics (e.g., claims, definitions, examples, proofs, analogies, computations, uses of tools, research directions) as mathematically sound, useful, well-posed, and appropriately justified. The recommendation is to center training on mathematical judgment, and we examine possible policies for graduate programs in Mathematics to this end.

A general framework for inferring the computational role of spike-timing-dependent plasticity (STDP) does not currently exist. Here, we develop a game-theoretic description for a canonical, convergent neural circuit with general STDP and postsynaptic-potential (PSP) kernels. Parity-matched STDP-PSP interactions induce a potential game among presynaptic neurons, whereas parity-mismatched interactions induce a zero-sum game. These components, respectively, implement contrastive PCA and flow selection; their weighted interplay determines learning dynamics and computation for arbitrary STDP rules.

Multi-party human-robot collaboration poses a dual challenge: robot decisions should remain interpretable and auditable, while executed actions must satisfy safety constraints during physical interaction. Combining explainable decision-tree policies with control-barrier-function (CBF) filtering provides a promising architecture but creates two learning mismatches in multi-agent reinforcement learning. Safety projection changes the action applied to the environment, while the coupled proposal graph can misalign independently optimized actor updates with a team-level update. We present safety-aligned gradient enforcement (SAGE) to address both mismatches. Its shield-annealed internalization layer (SAIL) uses a differentiable finite-penalty proposal map while retaining the exact CBF quadratic program for execution, preserving constraint-normal sensitivity to internalize repeatedly active safety constraints. Team-averaged Lyapunov policy optimization (TALO) constructs a team-aware update reference and applies a Lyapunov half-space correction to regulate independent actor updates. Physical experiments with two humanoid robots and a human partner demonstrate deployment feasibility. Across nine simulation scenarios, SAGE achieves a 71.0% success rate with 0.5 collision steps per thousand environment steps. Ablations show that direct CBF filtering reduces collision frequency by 98.5% but decreases success from 67.3% to 59.3%. SAIL reduces proposal violation by 48.8% and proposal-execution correction by 85.2%, while TALO reduces the update-consistency gap by 50.8%.

A large catalog of asymptotically free and anomaly-free SU(N) chiral gauge theories that admit a large-N limit was constructed by Eichten, Kang, and Koh. Here we compute the global structure of their symmetry groups and tabulate the 't Hooft anomalies of the faithful symmetries, including those that are visible only in backgrounds carrying fractional flux. Most theories in the catalog contain fundamentally charged matter and support no genuine one-form electric center symmetry. However, fractional fluxes still arise from the faithful quotient and contribute to the anomaly in the same way as a two-form background for a one-form symmetry. The anomalies are computed using two methods, first with the descent procedure, and second by placing the theory on a four-torus, which supports twisted fluxes for background gauge fields. We discuss a selection of candidate IR behaviors and present a general discussion of how matching the anomalies associated with fractional flux can constrain infrared descriptions beyond the imposition of ordinary zero-form anomaly matching at the level of triangle diagrams. Based on large-N reasoning, we formulate a new proposal for the IR behavior of a pair of models, consistent with the ordinary zero-form 't Hooft anomaly matching conditions. In this case the matching condition involving fractional flux is satisfied automatically as a consequence of ordinary matching.

We investigate the nonequilibrium evolution of density correlations in a thermal one-dimensional Bose gas following a trap quench. We focus on the Tonks–Girardeau limit and develop an analytical description based on a finite-temperature extension of Quantum Generalized Hydrodynamics. We benchmark our predictions against exact numerics for a quench from a double well to a harmonic trap and for a trap release, and assess their temperature range of validity. Our results capture the dynamical enhancement of correlations during the expansion stages, namely the crossover from thermally suppressed to power-law bulk correlations. For the trap release, the conformal distance associated with fixed bulk points decreases as $1/t$, so the correlations progressively probe a universal regime~$\sim1/x^2$ that is independent of temperature. We thus show that, in the hydrodynamic long-time regime, the bulk density correlations approach their zero-temperature equilibrium form independently of the initial temperature.

The 30 and 40 GHz instruments of the QUIJOTE radio astronomy experiment use very sensitive receivers which need to be characterized both to ensure their proper functioning and for their calibration. Given the age of the PXI-1031DC device used up until now for this purpose, the design of a new device, the Calibrator TFGI QUIJOTE, has been proposed. This new device has been designed with the idea of being simple and modular, easy to use and upgrade. It is built on a Raspberry Pi 5 system, using an MCC 118 as analog reading module. This system, widely known among both makers and professionals, not only meets the requirements of modularity and ease of use, but also ensures that the device is much more economically competitive than the PXI. A first version of the device has been built and tested, obtaining good results in accuracy and ease of use.

We propose a decoupled, layerwise method for structurally sparsifying the fully connected layers of pretrained neural networks. Rather than penalizing all layers jointly, our approach extracts shallow two-layer subnetworks, normalizes the inner weights, and applies a structured group penalty to the outer weight matrix of each block, processing layers sequentially to prune neurons and reduce the width of each layer. We prove that the constrained decoupled objective is equivalent at optimality to a specific joint penalty on the inner and outer weights, for any positively homogeneous activation, and thus admits a clean projected and proximal formulation. Our central finding is that this decoupled reformulation is more robust than coupled methods. In numerical experiments it provides a wider usable range of the regularization strength and a lower rate of catastrophic over-pruning than the tested joint baseline while maintaining comparable accuracy. We establish these properties in controlled classification and sparse-recovery studies, and examine their scope in a high-dimensional PINN stress test and in the feed-forward layers of OPT-1.3B.

The discovery of the third known giant planet, $\beta$ Pic d, offers a new opportunity to investigate how planet-disk interactions build the minor-body architecture of the young Beta Pictoris system. We study the formation and redistribution of scattered-disk objects, falling evaporating bodies, a proto-Oort-cloud population, transitional interstellar objects, and interstellar objects. Using $N$-body simulations, we evolve a primordial planetesimal disk extending from 0.5 to 1000 au for 25 Myr under the perturbations of $\beta$ Pic b, c, and d, together with reconstructed stellar encounters from Gaia DR3. The resulting evolution is strongly radius-dependent. The planets efficiently restructure the inner disk, while the outermost primordial population remains largely unchanged beyond $325$ au. $\beta$ Pic c remains the main driver of the star-grazing population, showing that the falling-evaporating-body pathway persists after the inclusion of $\beta$ Pic d. At larger radii, $\beta$ Pic b and d dominate planetary scattering and produce a surviving disk edge near 50 au, consistent with the observed debris-disk truncation. Over 25 Myr, $8.74\%$ of the initial disk enters the scattered-disk population, $2.04\%$ reaches the proto-Oort-cloud region, and $6.44\%$ becomes unbound. Planetary stirring also generates a transient one-armed spiral among particles originating at $\sim60$-150 au, overlapping part of the observed "Cat's Tail" and the radial location of the prominent CO clump near 85 au. The reconstructed stellar encounters included here produce only weak perturbations, although the encounter catalog is complete only over the first ~2.5 Myr of the present epoch. The early evolution of the Beta Pictoris minor-body system is therefore dominated by its planetary architecture, offering a view of distinct minor-body populations while they are still being formed and redistributed.

We study the first-passage time statistics of an active Ornstein–Uhlenbeck particle on a finite one-dimensional interval with an absorbing boundary at one end and a reflecting boundary at the other, subject to stochastic resetting. In the weak-activity regime, we develop a perturbative solution of the Fokker–Planck equation and use a renewal framework to obtain analytical expressions for the first-passage-time distribution and its low-order moments. In the absence of resetting, activity can either increase or decrease the mean first-passage time, depending on the relation between the persistence time and the diffusive time to absorption. When resetting is introduced, we determine both the onset of beneficial resetting and the finite range of resetting rates for which resetting lowers the mean first-passage time of the active system. We further show that the resetting transition depends on the velocity resetting protocol. Finally, by comparing the active particle with resetting to the passive Brownian particle without resetting, we construct a phase diagram that identifies the regions of parameter space in which activity and resetting together reduce the mean first-passage time.

The dominant method of processing sonar data is using image-based representations, requiring the preprocessing of image data on autonomous systems. We propose an alternative data processing method for remote sensing applications via the use of data in Comma-Seperated Value format. Experimentation on our alternative approach shows a reduction of processing time by 91.18%, an improvement in accurate object detection by Machine Learning, and an increase in SNR (Signal-to-noise ratio), PSNR (Peak signal-to-noise ratio), and other evaluation metrics.

Manipulating previously unseen objects remains challenging, as their dynamics depend on latent physical properties, such as friction and mass distribution, that cannot be inferred from perception alone. Prior experience across objects can provide an initial estimate of unseen object dynamics, but this estimate remains uncertain and can degrade further during sim-to-real transfer. Adapting the dynamics through interaction can progressively refine the estimation, however, updating the model may invalidate the planned trajectory. Successful and efficient manipulation therefore requires both rapid dynamics adaptation and a planning strategy that can incorporate this evolution. In this work, we introduce MetaPusher, a meta-learning and adaptive planning framework for nonprehensile manipulation of unseen objects without prior object-specific interactions. A meta-learned dynamics model rapidly adapts from interactions during task execution, while an adaptive kinodynamic planner updates long-horizon plans by reusing and refining its existing search tree. This coupling enables manipulation and adaptation without a separate data collection phase. We evaluate MetaPusher on unseen objects in simulation and in sim-to-real scenarios, comparing against fine-tuning and active learning methods, MPPI-based control, and a reinforcement learning policy. It achieves lower prediction error and improves task success rate by up to 20%.

Chui's conjecture asks whether the average electrostatic field generated by $N$ unit point charges on the unit circle is minimized when the charges are equally spaced. Abakumov, Borichev, and Fedorovskiy proved an analogue of this conjecture in weighted Bergman spaces, showing that the corresponding norm is minimized uniquely by the $N$th roots of unity. In this paper, we give two alternative approaches to this weighted minimization problem. The first is a direct calculus-based proof. We reduce the norm to a pairwise interaction energy, identify the equally spaced configuration as a critical point, and analyze the resulting gradient system and Hessian. The second approach connects the problem with discrete energy minimization on spheres by Cohn and Kumar. By rewriting the Bergman interaction as a completely monotonic function of squared Euclidean distance, we recover the minimizing property of the regular $N$-gon. This second argument also extends the minimization result beyond the concave-weight setting of Abakumov, Borichev, and Fedorovskiy, including the power weights $g_\alpha(t)=t^\alpha$ for every $\alpha>0$.

Van der Waals (vdW) heterostructures provide a versatile platform for exploring emergent quantum phenomena, yet their fabrication is often limited by interfacial contamination and incompatibility with surface-sensitive characterization. Here, we demonstrate muscovite mica as a polymer-free platform for assembling and patterning vdW heterostructure devices. Mica-mediated transfer enables clean exposed surfaces suitable for atomic-resolution scanning tunneling microscopy (STM) following mild thermal annealing, without aggressive post-processing. Mica also serves as a mechanically robust support for sequential pickup assembly, eliminating intermediate release steps. Furthermore, exfoliated mica flakes function as removable shadow masks for contact deposition, enabling straightforward patterning without conventional lithography. We quantify the temperature dependence of mica-mediated graphene pickup and demonstrate gate-dependent transport in a three-terminal graphene/hBN device fabricated using the approach. By integrating transfer, sequential assembly, and contact patterning within a single materials platform, mica provides a simple and accessible route to unencapsulated vdW heterostructures for surface-sensitive spectroscopy and quantum-device applications.

A novel reinterpretation on the effective one-body (EOB) mapping of compact-body binary (CBB) coalescence was previously considered. Instead of the conventional “marble in a funnel" description of EOB, CBB coalescence is modeled as the concurrent rotation and contraction of a hollow shell with a constant reduced mass measure. Merger is achieved, therefore, when the shell radius reaches the total mass radius. While past works analyzed the energetics sector of the model and the radiated gravitational waves (GWs), this work considers the geometric waveform modeling aspect. More specifically, we mature the mass-shell waveform analysis initially discussed with toy-model conditions by imposing necessary non-conservative mechanics, leading to the analytical expression of usable mass-shell modeled GW-forms. This is to set the foundation for eventual computational endeavors with the analytical waveforms presented here, enabling a comparison against current and state-of-the-art models.

Relativistic magnetic reconnection, a prime mechanism for particle acceleration, has been extensively studied in electron–ion and pair plasmas, but how the dissipated energy is shared among species in electron–positron–ion plasmas, and in particular how ions are energized within a pair-dominated layer, remains poorly understood. We develop an analytic theory that couples single-particle orbit dynamics to the collective plasma response, and we validate it with particle-in-cell simulations spanning pair-dominated to electron–ion compositions. Particles are directly energized in electric-dominated regions until the reconnected magnetic field deflects them into the outflow; the characteristic escape length is in turn set by the relativistic skin depth of the reconnection-heated mixture, including the ion contribution. Writing $\sigma_e^*$ for the magnetic energy available per particle in units of $m_ec^2$ and $\mu=m_i/m_e$, the mean lepton energy scales as $\sigma_e^*$, whereas the mean ion energy scales as $\mu^{1/3}(\sigma_e^*)^{2/3}$ when ions escape non-relativistically from the electric-dominated regions and approaches the lepton energy when their escape is ultrarelativistic. The predicted energy fractions agree with simulations across composition, mass ratio, and magnetization. The ion share of the dissipated energy can fall far below the commonly assumed ion–lepton equipartition, reducing the energy available for hadronic and neutrino emission in reconnection-powered astrophysical sources.

AI productivity is often measured by task completion time, economic value, or improvements in outcome quality. However, these measures usually treat collaboration as a black box where they capture what output was produced, but not the interaction cost required to produce it. Motivated by economics literature, we introduce a productivity-oriented framework for evaluating human-AI collaboration as outcome quality relative to interaction cost. Across two datasets spanning four tasks, we show that: (1) sessions with identical quality ratings can differ by up to 70 times in interaction cost; (2) quality-cost relationships vary by task, with some tasks rewarding extended interaction and others favoring fast convergence; (3) subjective user ratings are not reliable substitutes for productivity; and (4) productive sessions are characterized by agents probing earlier and users spending less effort repairing the interaction. By distinguishing productive success from costly success, our framework makes interactional cost visible and shows how dialogue analysis can inform the evaluation and design of AI systems.

Let $\mathcal{A}_0$ be the set of finitely generated free groups on some countable alphabet and and define $\mathcal{A}_r$ inductively as the union of $\mathcal{A}_{r-1}$ with the set of amalgamated free products $G_1\ast_{\ZZ}G_2$, where $G_1,G_2\in \mathcal{A}_{r-1}$. We show, for all $r$, that every element of $\mathcal{A}_r$ has linear conjugator length, quantifying a result of Larsen and extending a result of Kharlampovich and Myasnikov. This is achieved by a linear conjugator length bound on a family of iterated HNN extensions of free groups. Also, in an appendix, we adapt a proof of A. Wei$\ss$ to show that all Generalized Baumslag-Solitar Groups have linear conjugator length. As a consequence, we obtain that one-relator groups with non-trivial center have linear conjugator length, giving evidence for a conjecture of Bridson, Riley, and Sale.

Molecular profiling from routine histopathology could expand access to precision oncology when sequencing is unavailable, tissue is limited, or training cohorts are small. We developed a hybrid quantum-classical strategy that replaces softmax attention in a transformer for histopathology-based gene expression prediction with a quantum-derived doubly stochastic matrix (QDSM). Across 29 cancer cohorts from The Cancer Genome Atlas and an independent pancreatic cancer cohort from the Clinical Proteomic Tumor Analysis Consortium, QDSM attention produced selective gains, with the largest relative improvements in smaller, data-limited cohorts, including adrenocortical carcinoma and uveal melanoma. Rather than improving transcriptome-wide performance uniformly, QDSM redistributed predictive accuracy across genes and pathways, improving biologically relevant targets in some tumor contexts while worsening others. In adrenocortical carcinoma, preferentially improved genes were enriched for adverse overall-survival associations, linking enhanced molecular inference to prognostically relevant biology. In pancreatic cancer transfer experiments, QDSM improved selected metabolic and lineage-associated genes but did not consistently improve performance under cross-cohort shift. Leave-one-cancer-out mixed-effects analysis showed that baseline molecular features predicted part of the gene-level benefit, while residuals identified cancer-specific programs that improved more or less than expected. Separate experiments on IBM quantum processors recovered the doubly stochastic matrix primitive underlying the attention mechanism. These findings position QDSM attention as a context- and target-dependent strategy for image-based molecular profiling and molecular triage when direct testing is unavailable, incomplete, or impractical.

Robots carrying out tasks in dark environments need to localize from a single RGB camera, in light so low that the per-pixel signal approaches the sensor's own noise, on a power-constrained onboard computer, in real time. Each of these constraints has matured pipelines, but the intersection does not. Offline low-light reconstruction now recovers structure below -4 dB but is far too slow to run in real time, while the real-time monocular systems a robot can actually carry (DROID-SLAM, DPV-SLAM, etc.) degrade or fail when SNR gets low. We measured how they fail: across the nine lowest darkness levels of our scenes, DROID-SLAM returns a full-length trajectory carrying no information about the camera's motion on all nine, VGGT-SLAM and CUT3R on eight, pi^3 on seven, and DPV-SLAM on four. We present SYS, a monocular pipeline built on a low-light feed-forward pointmap front end with an explicit match gate, which returns three tracked trajectories and no uninformative ones, at the lowest error of any method where it tracks (24-47% of the no-information ceiling against 56-73% for the strongest baseline), and at the narrowest coverage. On a real robot video take in which 86.5% of delivered frames are entirely black, every configuration of ours stops after the lit beginning, while DROID-SLAM and DPV-SLAM each emit a pose for all 1178 frames. Our method contribution is an embedded execution path for the Jetson AGX Orin: running the map, keyframes and backend at 384 pixels with tracking at 256, together with two fixes to the per-frame pose solve, is a replicated Pareto improvement, 1.28x throughput at 0.964x error on one scene and 1.42x at 0.68x on a second, with 47% less peak GPU memory and 29% less energy per pose. We evaluate on a calibrated, bit-exact regenerable noise ladder, on relabelled real-world dark exposures, and on a new dark-room video ladder recorded from a Boston Dynamics Spot robot.

Fine-tuning has emerged as a widely adopted approach for adapting LLMs to a variety of downstream tasks. However, how it reshapes their internal mechanisms remains poorly understood. To address this, we investigate how fine-tuning alters internal representations in LLMs, including attention patterns and layer-wise activations, and examine whether these changes are linked to task-relevant components identified by EAP (e.g., attention heads and logit-level activations) that drive task performance. We find that EAP-identified components are concentrated within specific layers, indicating a degree of functional localisation in how models internalise task-specific behavior. Notably, the distribution of these components across layers is largely uncorrelated with the layers undergoing the most substantial representational changes during fine-tuning. Furthermore, we observe that overlap in EAP-identified components across tasks does not translate into cross-task performance transfer if the tasks are different in nature (e.g. classification vs. generative tasks). More specifically, fine-tuning on one task can lead to a degradation of performance on another when the two tasks exhibit a high degree of overlap in their EAP-identified components.

Training a visuomotor policy calls for abundant demonstrations that closely match the target environment, yet collecting them anew remains expensive. Existing demonstration synthesis methods reduce this cost but remain constrained by high manual effort, limited visual fidelity, or heavy reliance on physics simulators. This paper introduces GaussianFactory, a high-fidelity data engine that mass-produces demonstrations with a single video scan as its only human input and no physics engine in the generation loop. Specifically, GaussianFactory reconstructs the scene as an editable 3D Gaussian Splatting (3DGS) replica and samples from the object-combination tasks the scene affords. For each task, it plans grasps and trajectories purely kinematically on the geometric reconstruction, rendering photorealistic demonstrations that visually match the target environment. Physical dynamics enter the pipeline only where contact force interactions dictate the outcome—during grasp formation, via a learned contact model pretrained once on an interaction dataset. To evaluate the downstream utility of the synthesized demonstrations, we implement an end-to-end scan-to-deployment workflow in two setups: a simulated scene that stands in for the real world to enable reproducibility, and a real-world workspace with a physical UR10e robot. In each setup, a standard diffusion policy trained solely on the synthesized demonstrations achieves 95.1% and 84.2% success rates, respectively.

For a suitable class of structures $\mathcal{K}$, let ${\rm Sh}(\mathcal{K}_\lambda^{\rm op},J_{\rm at})$ denote the atomic sheaf topos over the opposite of the category $\mathcal{K}_\lambda$ of structures in $\mathcal{K}$ of cardinality at most $\lambda$, with embeddings as morphisms. In this article, we analyze the relationship between model-theoretic properties of $\mathcal{K}$ and logical properties of ${\rm Sh}(\mathcal{K}_\lambda^{\rm op},J_{\rm at})$. For an infinite cardinal $\lambda$, we show that $\mathcal{K}$ is Galois $\lambda$-stable if and only if the atomic sheaf topos ${\rm Sh}(\mathcal{K}_\lambda^{\rm op},J_{\rm at})$ satisfies the axiom of choice for $\lambda^+$-indexed families. We further show that every structure in $\mathcal{K}_\lambda$ extends to a maximal structure if and only if ${\rm Sh}(\mathcal{K}_\lambda^{\rm op},J_{\rm at})$ satisfies the internal axiom of choice. If $\mathcal{K}_{\rm fin}$ is a class of finite structures, the cardinalities of the structures in each connected component of $\mathcal{K}_{\rm fin}$ have a finite upper bound if and only if ${\rm Sh}(\mathcal{K}^{\rm op}_{\rm fin},J_{\rm at})$ satisfies the internal axiom of (countable) choice. Furthermore, via the topological Galois representation, we establish a correspondence between the criteria for the axiom of choice for atomic sheaf toposes and those for permutation models in set theory.

Scientific sensor data is foundational across disciplines including spacecraft engineering, clinical medicine, and atmospheric science. In each context, pipelines are constructed to ingest raw archives, assess data quality, perform feature engineering and semantic annotation, and record provenance. However, these pipelines are often implemented as monolithic, domain-specific scripts with implicit assumptions and limited reusability across fields. This work introduces SensorWF, a FAIR-annotated workflow framework for generalizable scientific time-series analysis. The framework features a five-module reusable core (M1-M5) with typed input/output contracts. A domain adapter pattern isolates all domain-specific logic within M1, enabling modules M2-M5 to operate identically across disciplines. Domain assumptions are encoded in a machine-readable component registry, enabling reuse across sensor domains without modifying the analytical core. SensorWF also generates runtime PROV-O/ProvONE provenance traces with SHA-256 checksums for all file-path entities and emits SSN/SOSA-aligned OWL ontologies as primary outputs. To assess generalizability, SensorWF is instantiated in three distinct scientific domains: spacecraft telemetry, ambulatory ECG, and atmospheric climate, with synthetic fault injection and multi-detector anomaly detection demonstrated as use-case extensions. Results show that a single codebase, parameterized solely through M1 adapters of approximately 170-500 lines each, supports analytical pipelines across domains with varying sampling rates, channel counts, and fault taxonomies. All code, the component registry, ontology artifacts, and datasets are made available as an open scientific object. Our codebase is publicly available at https://purl.archive.org/sensor-wf.

Recent advances in large language models have improved their effectiveness as back-end components for voice assistants, particularly in intent understanding and context-aware input classification. However, online-hosted models introduce network dependency and variable inference latency, limiting their suitability for time-critical autonomous driving applications. In this work, we address these issues by developing Jarvis, an offline voice assistant for high-level behavioral commands of autonomous vehicles. Its architecture integrates speech recognition and synthesis with natural language command classification into a lightweight, local framework. Jarvis core component is a text-to-command classifier, built using a domain-specific fine-tuning of the Mistral 7B model, demonstrating low-latency inference. Our experimental evaluation demonstrates that our solution outperforms larger online-hosted models, achieving 97.63 % intent recognition accuracy with an average processing latency of 1.39 s, making it well-suited for operations requiring quick response times. To support further research and fine-tuning, we provide an open-source implementation.

Deep reinforcement learning (DRL) has achieved strong performance across a wide range of continuous-control problems. These continuous-control policies, however, are often defined as direct mappings from an observed state to an action or action distribution, requiring a single feed-forward network to construct an optimal control decision in one pass. While effective, this formulation leaves little opportunity for the policy to reconsider or progressively improve an action once an initial prediction has been formed. In this work, we explore an alternative approach: rather than learning only to directly predict an action, can a policy learn to iteratively improve one, and can this iterative process provide advantages during policy learning? We introduce Iterative Action Refinement (IAR), an iterative action-construction method that constructs control actions through a sequence of learned residual corrections. Starting from an initial proposal, a shared refinement network repeatedly conditions on the observed state and the current action proposal, allowing each refinement step to revise the action constructed by preceding steps. The final refined proposal is then used to determine the action executed by the agent. We integrate this iterative action-construction mechanism with Proximal Policy Optimization (PPO), yielding REFINEPPO. We evaluate REFINEPPO across 14 benchmark control tasks, complemented by controlled ablations of refinement depth and update schedules and analyses aimed at understanding why iterative refinement is effective. Across these environments, REFINEPPO matches or exceeds the performance of standard PPO while demonstrating faster convergence on several tasks.

While learning from human motions has enabled highly dynamic humanoid skills such as dancing and martial arts in obstacle-free space, traversal through densely cluttered environments remains underexplored. These spaces are three-dimensional and geometrically constrained, requiring scene-aware locomotion that tightly couples whole-body motion with scene geometry for obstacle avoidance. To address these challenges, we present Moving Through Clutter (MTC), a learning-from-demonstration framework for scene-aware humanoid locomotion. To bypass costly physical scene construction, MTC uses procedurally generated Virtual Reality environments for immersive data collection. To transform these human motions into training-ready humanoid motions, we propose a scene-aware motion retargeting algorithm that converts human demonstrations into humanoid trajectories while strictly enforcing robot-scene clearance to guarantee collision-free traversal. These reference trajectories are then used to train a scene-aware locomotion policy that deploys on a Unitree G1 humanoid. Evaluated on our proposed MTC-Challenge for multi-obstacle traversal, the policy demonstrates a 70.2% collision-free rate across diverse scenarios, successfully traversing complex environments through diverse whole-body skills, including crawling through low-clearance passages and squeezing through narrow gaps.

Firing neurons generate action potentials that propagate along axons to transmit signals supporting cognitive functions. These electrical currents generate magnetic fields, yet direct detection of these neuronal magnetic fields by MRI remains elusive. This Mini Review investigates why this goal has been proven difficult to achieve and whether an emerging approach, quantum sensing MRI, can overcome the challenge.

The Auger Radio Infill SKALA Extension (ARISE) was installed at the Pierre Auger Observatory in 2025 to detect the radio emission in the band of $50-350\,$MHz of near-vertical cosmic-ray air showers of energies around and above $100\,$PeV. The array comprises 18 SKALA~v2 antennas deployed around one of the surface-detector stations of the $433\,$m array (SD-433), which provides the trigger for ARISE. We present first measurements with ARISE and show the detection of radio pulses associated with extensive air showers. Radio signals are found in coincidence with SD-433 events, and a comparison of the arrival directions reconstructed from the radio and surface detector signals confirms the measurement of air showers with ARISE. These observations provide a first assessment of the detector setup and demonstrate the potential of ARISE for radio detection of near-vertical air showers.