Cosmology and Nongalactic Astrophysics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
An ultra-slow-roll phase during inflation could potentially produce the right abundance of primordial black holes (PBHs) for them to form all of dark matter (DM). We consider such a scenario, carefully treating the transitions from slow to ultra slow roll inflation and back, using a parametrisation of the inflaton potential that can describe inflation from the time when CMB scales crossed out of the horizon until its end. A $\delta N$ analysis shows that this model can possess $O(1)$ non-Gaussianity. When computing the primordial black hole abundance, we keep the full non-linear relation between curvature perturbations and the density contrast and consider the non-Gaussianity to all orders. We find that $O(1)$ non-Gaussianity is sufficient to enhance the PBH abundance by orders of magnitude, leading to reduced gravitational wave (GW) signals for fixed PBH abundance. The signal to noise ratio is computed for future gravitational wave observatories. We find that, in spite of the reduced strength, the GW signal is well above the sensitivity of LISA if PBHs form a significant component of DM.
Materials Science
How to enhance the applicability of Murnaghan equation of state
The paper presents the brief overview of known isothermal equations of state for solids and of theoretical background to construct them, focusing on widely used Murnaghan equation. Since the latter has a limited range of application, failing to reproduce reliably both near-equilibrium and high-pressure regions simultaneously, the paper proposes simple, but rather efficient way to improve it. The suggested modification favors its applicability at high pressures, allowing to expand considerably the region of reliable reproducing of experimental or computational data. The proposed improved equation has been tested in fitting of U(V) data, obtained within ab initio computations, as well as of experimental V-P data for a representative selection of solids. In comparison with ordinary Murnaghan equation, it has found significant advantages, extending the range, where the computational data may be reproduced trustworthy (from the equilibrium state to the pressures, appreciably higher, the ordinary equation could). As for the fitting of experimental data, the proposed improved Murnaghan equation reproduces them, in the presented examples, at least not worse than other well-known equations (Vinet, Birch-Murnaghan, Mao) do. Some aspects of fitting process, as well as some alternative approaches to obtain the proposed equation of state in various forms are also discussed.
Read the full abstract
Audio and Speech Processing
Open-Set Vessel Re-Identification from Underwater Ship-Radiated Noise with a Raw-Waveform Selective-Kernel Acoustic Neural Network (SKANN) and a Cross-Passage Evaluation Protocol
Underwater acoustic target recognition has converged on closed-set classification by vessel type, a task that does not answer whether a monitoring system has heard this hull before. We formalise open-set, cross-passage vessel re-identification on public hydrophone data and specify a protocol that removes the two easiest routes to a high score: hull-disjoint splits keyed to MMSI/IMO, galleries and queries from disjoint passages of each hull, source-pure galleries, and an audio-adjudicated transit-deduplication gate. We describe SKANN, a raw-waveform encoder whose front end is a four-scale bank of learned filters fused by selective-kernel attention, trained with an angular-margin objective and an augmentation regime that perturbs recording chain, ambient noise and multipath while preserving the narrowband lines that carry identity. On a 40-hull IARA gallery (96 queries, 98 passage candidates), cross-passage rank-1 is 0.25 for the embedding and 0.26 for an automated narrowband-tonal comparator; the two are statistically indistinguishable at the top of the ranking, the embedding orders the rest of the list more reliably (AUC 0.82 vs 0.76), and their score fusion reaches rank-1 0.35 – the only contrast that attains nominal significance, presented as evidence of partial complementarity, not as a recommendation. Transit deduplication alone removes a 16-21 point apparent rank-1 advantage, larger than any between-method difference. Two further findings delimit what public data can support: ShipsEar cannot separate hull identity from recording channel under an identity protocol, and cross-network fine-tuning helps vessels seen during fine-tuning but is a null result on unseen ones. The results support analyst triage over a ranked shortlist, not identification. Checkpoint, validation embeddings, transit map and per-query outputs are released under CC-BY-4.0 (doi:10.5281/zenodo.22160138).
Read the full abstract
Robotics
OpenWAM: An Open, Modular Exploration Towards Systematic World-Action Model Pretraining
World-Action Models inherit world knowledge from video-generative priors, and channel it into executable control signals through embodied experience. Existing systems, however, are monolithic: the generative backbone, visual representation, architecture, information flow, inference procedure, and training data are tightly coupled, obscuring which design choices matter and why. We introduce OpenWAM, an open research stack that turns world-action pretraining into a controlled experimental program. OpenWAM-Infra factorizes the WAM design space into composable modules with unified training, inference, deployment, and evaluation. On this substrate, OpenWAM-Study examines three questions through controlled experiments: what to inherit, how world and action learning interact, and how their synergy scales; and distills three principles: upstream knowledge transfers through a sufficiently capable generative backbone and a compact, information-rich latent space; world-action synergy requires dedicated action capacity, explicit world-to-action information flow, and synchronized joint denoising; and embodied pretraining principally improves out-of-domain generalization, with one-stage co-training over egocentric and robot data integrating world coverage and action grounding. Composing these principles, we build OpenWAM-{\alpha}, an open WAM pretrained on roughly 6,400 hours of egocentric human and robot data and evaluated across simulation and real-world benchmarks. Across the eight simulation benchmarks and the real-robot experiments, which together span embodiments from single-arm and bimanual manipulation to dexterous hands, OpenWAM-{\alpha} delivers consistently excellent performance, sustaining its top-tier standing from simulation to the physical world. We release the full stack, including infrastructure, evaluation protocols, pretrained models, and data recipes, to facilitate future research.
Read the full abstract
Quantum Physics
Riemannian Optimization for Multi-Player Quantum Games on Product Unitary Manifolds
Quantum game theory is an extension of classical game theory that uses quantum principles in game theory. The Eisert-Wilkens-Lewenstein (EWL) quantum game is an early example of the two-player classical Prisoner's Dilemma transformed into a quantum Prisoner's Dilemma. In the EWL game, the players choose pure quantum strategies represented by unitary matrices. This extension can resolve the classical dilemma by enabling cooperative equilibrium with higher payoff. In this paper, we first discuss the Extended EWL (EEWL) for multiplayer quantum games with mixed strategies. In EEWL, each player controls a set of unitary operators as quantum actions and uses a classical mixed strategy over these actions. The payoffs are defined as expectation values of Hermitian reward operators acting on a shared quantum state, which is generated and measured according to the EEWL protocol. We then propose the Unitary Strategy Matrix Exponential Algorithm (USMEA), a geometry-aware sequential algorithm for the EEWL mixed-strategy setting, in which each player jointly learns a trainable set of local unitary actions and the associated classical mixing probabilities. Thereby it acts as a learning-and-control layer for multi-agent quantum decision systems. We analyze the convergence properties of USMEA under standard smoothness and step-size conditions and validate the theory with numerical experiments. These results show how classical optimization methods can be systematically integrated into the design and analysis of engineered quantum strategic interactions.
Read the full abstract
Statistical Mechanics
Fragility from Enthalpy Fluctuations and Fictive-Temperature Dynamics
Fragility and fictive temperature are two central measures of glass formation, but their connection is usually expressed indirectly through structural relaxation times or cooling-rate dependence. Here we formulate this connection in terms of configurational enthalpy fluctuations. The time evolution of the fictive temperature during cooling or aging is controlled by the enthalpy–enthalpy correlation function that also determines the frequency-dependent configurational heat capacity. Fragility enters through the strong temperature dependence of the structural relaxation time and therefore controls how rapidly the corresponding enthalpy-relaxation spectrum shifts to long times near the glass transition. This leads to a direct time-domain description of the onset and growth of the fictive-temperature lag, including the Debye limit, and connects calorimetric relaxation, fragility, and nonequilibrium structural arrest within a common framework.
Read the full abstract
Information Retrieval
Uncertainty Quantification for LLM Agents: A Taxonomy, an Evaluation Protocol, and an Empirical Study
Large language models (LLMs) are no longer deployed only for single-turn conversation but increasingly act as agents that plan, call tools, retrieve evidence, maintain memory, and interact over long horizons, often together with other agents through multi-turn conversations. Therefore, knowing when to trust the agentic system is a prerequisite for safe deployment. However, existing work on quantifying uncertainty for LLMs was built almost entirely for single-turn question answering. This paper argues that errors and uncertainty arise from multi-turn conversations, environments, and tools rather than from a single-turn question answering setting. It comes late, however, and is compounded in a single score that is too coarse to represent the unreliability. We organize the literature with a three-axis taxonomy, (1) what the uncertainty is, (2) how it is estimated, and (3) where uncertainty arises during an agent pipeline. We investigate step-level and trajectory-level calibration and show with a simple counterexample that the first does not imply the second. Experiments on real agent traces across four models and up to a 50-step budget show that the proposed metric and reporting protocol (Trajectory-Checkpoint Expected Calibration Error, TC-ECE) can be computed and that step errors are coupled along a trajectory. We find that confidence estimates from the agent's own responses do not consistently outperform a simple baseline. The experiments also show that averaging all trajectories together can hide overconfidence at later stages, which becomes visible when results are analyzed across different horizons. In simpler terms, this paper identifies where the uncertainty comes from in the agentic system pipeline, how to teach agents to know when they are wrong, and why one confidence number is not enough.
Read the full abstract
Computer Vision
Social Intuition vs. Machine Reasoning: Anticipating Human-Robot Interaction from multiple modalities
Anticipating whether a person will interact from one's own perspective is a highly intuitive task for humans, that relies on a combination of cues. We investigate how humans perform at predicting a person's intention to interact from a service robot's point of view, using pose-only or full video input, then benchmark different lightweight pose-based models and state-of-the-art vision-language models. We conducted our benchmark on the HUI360 dataset on a fixed pilot subset of 100 test tracks (25 positive, 75 negative). We found that with pose-only input, human annotators outperform lightweight trained pose models but not by large margins (+0.08 in F1-Score). But when given full egocentric video with a target bounding box, human annotators perform substantially better and largely outperform the Vision-Language Models (+0.2 in F1-Score). We also compared VLMs of different size and under different input conditions, and found that the best results do not correlate with model size. Our result confirms that predicting interactions is a challenging task for social robots and that reasoning-capable models are necessary but their actual reasoning capabilities alone do not suffice to match the social intuition of humans.
Read the full abstract
Optics
Mechanical quality factor estimation from a nonlinear optical cavity response
Intrinsic damping rate and quality factor ($Q$) are key parameters in optomechanical systems. However, in high-Q devices, their measurement is often challenging because probe-induced dynamical backaction (DBA) and the need to resolve and accurately track the mechanical resonance limit the performance of state-of-the-art techniques. Here, we introduce an all-optical ringdown method in the unresolved-sideband regime that exploits the nonlinear, time-averaged optical response induced by strong mechanical motion. We derive an analytical model of the time-averaged optical cavity response probed by a weak frequency-scanning laser. During the evolution of the mechanical oscillator, the excited motion modulates the Lorentzian lineshape into a double-horned profile whose peak separation enables an accurate estimation of the modulation amplitude. Tracking this response during the mechanical ringdown allows the intrinsic damping rate to be determined without resolving the mechanical oscillation. We apply the method to a density-modulated phononic crystal membrane placed inside a high-finesse cavity and obtain mechanical quality factors consistent with independent measurements on the same platform. Our method requires a minimal optical setup, while offering an in situ diagnostic and reduced impact from probe-induced DBA and robustness against measurement noise. Finally, the model can describe other resonant systems dispersively coupled to coherent frequency modulations, providing new insights and measurement strategies that expand beyond the optomechanical platform.
Read the full abstract
Statistical Methodology
Estimation and Inference for Latent Markov Models by Fourier Recursions
This paper proposes a new theoretically exact Fourier recursion framework for a broad class of latent Markov models (LMMs), encompassing models widely used across a broad range of fields in economics. It can be viewed as a counterpart of the celebrated Kalman filter for non-Gaussian and nonlinear LMMs. Closed-form recursive updates of Fourier coefficients jointly deliver filtering, likelihood evaluation, and simultaneously accumulate the score and Hessian online. We introduce a unified truncated implementation that ensures uniform error control and numerical stability, preventing approximation errors from accumulating through the recursion. We establish asymptotic properties of the feasible maximum likelihood estimator for LMMs, provide a recursion-based consistent estimator for the Fisher information and discuss the possible model misspecification. These results provide the first general asymptotic theory for feasible approximate maximum likelihood estimation in LMMs. Simulations demonstrate its accuracy and stability, while an application to U.S. bankruptcy data recovers a persistent latent bankruptcy-pressure process.
Read the full abstract
Quantum Physics
Unitary equivalence of Schrödinger and Heisenberg pictures survives in nonlinear quantum mechanics
The equivalence of Schrödinger and Heisenberg pictures is often thought to break down in nonlinear quantum mechanics due to the absence of a state-independent unitary propagator. We show that for Hamiltonians depending on the state through instantaneous expectation values, the equivalence is preserved. While the unitary intertwiner becomes initial-state dependent, it remains well-defined and ensures identical physical predictions in both pictures. We illustrate this with the analytically solvable examples of a nonlinear spin precession and a mean-field harmonic force, as well as in local field theory with mean-field coupling.
Read the full abstract
Solar and Stellar Astrophysics
An Eccentric Massive Protobinary Assembled via a Core-merger Parabolic Encounter
Most massive stars form in binary systems, which profoundly influence their subsequent evolution. However, how such systems form remains poorly understood, with several competing scenarios proposed, including disk fragmentation, core fragmentation and capture. Determining the orbital architectures of massive binaries, particularly during their earliest embedded phases, is therefore crucial for distinguishing among these formation pathways, but direct measurements of their three-dimensional motions have remained exceptionally challenging. Here we present high-resolution, multi-epoch sub-millimeter-to-centimeter ALMA and JVLA observations of the massive protobinary IRAS 07299$-$1651, complemented by JWST and VLT infrared imaging. We detect orbital proper motion of the binary components, enabling a full three-dimensional orbital reconstruction. Combining orbital fitting, multi-wavelength continuum modelling, hydrogen recombination line kinematics and jet observations, we find that the preferred orbital solutions are highly eccentric and close to parabolic, while both circumstellar disks are strongly misaligned with the orbital plane. These properties are naturally explained by a “core-merger” scenario in which the two protostars originated independently from initially unbound cores that recently underwent a near-parabolic encounter, producing an eccentric binary with a current separation of about 200 au. These findings suggest that the core-merger process may represent an important pathway for forming eccentric massive binaries.
Read the full abstract
Software Engineering
A Surrogate-based Approach for Fast Multi-objective Architectural Refactoring Optimization
Software model optimization is a process that generates architecture alternatives aimed at improving quantifiable non-functional properties of software systems, such as performance and reliability. Multi-objective evolutionary algorithms are commonly used to explore the search space and help designers identify trade-offs among competing non-functional properties (e.g., through a Pareto front). However, such algorithms face efficiency challenges in complex software models and large design spaces, since evaluating the fitness (i.e., the quality) of each architecture requires analysis tools that become computationally expensive when repeatedly invoked during the search process. In this paper, we explore the construction of surrogate models based on regression techniques to approximate the outputs of these analysis tools at significantly lower computational cost, while maintaining reasonable output accuracy. Our experimental results suggest that surrogate models provide savings of up to $30\%$ in computational time and maintain the Pareto front quality provided by evolutionary algorithms. Also, we observed some differences in the architectural models produced by our approach. Overall, surrogate models constitute a promising approach for scaling multi-objective architecture optimization to larger spaces and complex architectural models.
Read the full abstract
Numerical Analysis
Frequency Domain Biot–Allard Equations for Isotropic and Anisotropic Poroelastic Media: Two-field formulations and iterative splitting
We present a frequency-domain formulation of Biot's dynamic poroelastic equations with frequency-dependent dissipation (Biot–Allard) for anisotropic, heterogeneous media with memory effects. Two equivalent two-field representations–a displacement-pressure and a velocity-pressure-rate formulation–enable stabilized iterative splitting. While coupling operators generally lack an adjoint or skew-adjoint relationship at finite frequencies, the velocity–pressure-rate representation restores a skew-adjoint structure in the quasi-static limit. We prove continuity of the coupling operators and coercivity of the diagonal blocks, essential for convergence of the L-stabilized splitting scheme. The frequency-domain setting eliminates convolutional memory terms, incorporates attenuation and dispersion via complex-valued parameters, and reduces the time-dependent problem to a family of elliptic boundary-value problems suited for parallel computation and multi-frequency inversion. A conforming Galerkin finite element discretization preserves block structure, and numerical experiments confirm robustness and capture frequency-dependent attenuation. To illustrate discretization independence, we include a large-scale wave simulation using a pseudo-spectral method. This work provides a rigorous and efficient framework for modeling wave phenomena in complex porous media.
Read the full abstract
Mathematical Physics
Towards Heterogeneous Exascale CFD with a Single Fortran Code Base: OpenMP Target Acceleration of the High-Order Unstructured Solver UCNS3D
Heterogeneous exascale systems are reshaping computational fluid dynamics, yet rewriting mature high-order solvers for accelerators can fragment code bases and weaken reproducibility. We present a single-source OpenMP target modernisation of UCNS3D, a high-order unstructured finite-volume solver for compressible flows. The common CPU/GPU implementation preserves its numerical formulation, Fortran code base, and workflows. Rather than porting isolated kernels, we enable the complete explicit time-advancement path: high-order reconstruction, gradient evaluation, inviscid and viscous fluxes, boundary treatment, halo exchange, and solution update. The implementation uses persistent target data regions, flat run-time arrays, compile-time sizing of temporary storage, explicit local algebra, target-callable routines, and GPU-aware MPI with device-resident buffers. These choices address the irregular stencils, complex data structures, and substantial temporary storage of high-order unstructured CFD. Verification using the compressible Taylor-Green vortex shows CPU and GPU dissipation histories agreeing to machine precision on LUMI and with published reference data. A single-node run on the ARCHER2 GPU platform provides an independent portability check. End-to-end performance is evaluated on fully populated nodes using Taylor-Green vortex, LM1021 sonic-boom, and NASA high-lift CRM benchmarks. Relative to the previous production implementation, the refactored CPU path is 1.27-1.67 times faster, while GPU offload delivers same-node speed-ups of 2.71-4.05 and 84-101 percent strong-scaling efficiency. The results show that standards-based OpenMP provides portable, production-scale acceleration without sacrificing numerical fidelity, CPU performance, or software sustainability.
Read the full abstract
Networking
Det-5G: Closing the Determinism Gap in 5G-Advanced for Industrial Closed-Loop Control
The ultra-reliable low-latency communication (uRLLC) capability of 5G has created significant opportunities for industrial wireless connectivity, yet widespread use of cellular networks for closed-loop control remains challenging. Closed-loop control requires more than low packet latency and high reliability: cyclic command/feedback exchanges must complete within predictable time bounds despite changing channel conditions, recovery transmissions, mobility, and multi-device contention. This paper introduces Deterministic-5G (Det-5G), a unified radio resource allocation framework for industrial closed-loop control. Det-5G treats the complete bidirectional control cycle as the scheduling object and combines coordinated downlink/uplink allocation, adaptive bundled transmissions, group-oriented downlink communication, and optimized multi-user uplink scheduling over 5G air-interface. Its performance is evaluated through a combination of closed-form analysis and Monte Carlo scheduling experiments, with comparisons against conventional dynamic grant-based scheduling, semi-persistent scheduling/configured grant operation, and fixed proactive repetition. The evaluation shows that Det-5G improves predictability of cycle completion, maintains the target reliability under changing link conditions, and scales more effectively to multi-device control than conventional reactive scheduling, while adapting radio resource use instead of continuously provisioning for the worst case as in fixed repetition. These characteristics make cycle-oriented scheduling a pragmatic solution for reducing the determinism gap that limits the use of 5G for closed-loop control in different verticals, especially as it evolves through 5G-Advanced toward 6G.
Read the full abstract
Instrumentation and Detectors
Review of Hybrid and Opaque Scintillator Technologies Advancing Neutrino and Dark Matter Detectors
Organic liquid scintillators are used since several decades, especially in neutrino physics. They excel at the detection of low-energy particles where energy and timing information is required. Organic liquid scintillators have advantages such as high light yield, radio purity, cost effectiveness, and more. However, they also exhibit disadvantages, most prominently a lack of vertex resolution and particle identification. Various novel ideas have emerged over the course of the last decade to improve the performance of organic liquid scintillators. Among them are most prominently hybrid and opaque scintillators. In this proceedings, these novel scintillators are reviewed and their current applications in the research fields of neutrino and dark matter physics are illustrated.
Read the full abstract
Quantum Physics
Wavelength-Uniform Quantum Algorithms for Mixed-State Quantum Dynamics
One of the main challenges in numerical simulation of quantum dynamics is the prohibitive cost in the semi-classical regime, in which the de Broglie wave length is small compared with the characteristic length scale and the solution is highly oscillatory. For the von-Neumann equation for mixed-state quantum dynamics, this difficulty is overcome by using the Weyl variables, under which the solution is not oscillatory. Furthermore, we use the integral representation of the potential difference, which robustly captures the classical limit as the semi-classical parameter approaches zero. By using exact Hermite moments and quantum singular value transformation to treat the polynomial and sparse coordinate matrices of the dense projected smooth non-polynomial potentials respectively, we obtain a quantum algorithm efficient for all ranges of wave lengths, with complexity polynomial in the spatial dimension and discretization and query bounds containing no negative powers of the small wavelength. Thus it can capture the correct physical observables even if the spatial grid does not resolve the frequency, hence defying the Nyquist-Shannon sampling theorem.
Read the full abstract
Dynamical Systems
Generic reconstruction of rational maps from multipliers of periods one and two
For every integer $d\geq2$ and every field of characteristic different from $2$, we prove that on the moduli space $\mathcal{M}_d$ of degree-$d$ rational maps, the multiplier spectrum morphism formed from the periodic points of periods one and two is birational to the closure of its image. Consequently, over every algebraically closed field of characteristic different from $2$, it is generically injective, which proves a recent conjecture of Ji and Xie in characteristic zero. The proof uses a fixed-index normal form, a non-archimedean degeneration of two-cycles, and birational reconstruction of an affine fixed-point configuration from pair invariants.
Read the full abstract
Combinatorics
Shellability of semidistributive lattices
We prove that join-semidistributive shellable lattices are join-extremal. In particular, every semidistributive shellable lattice is extremal. Together with known results, this gives the equivalence between extremality, left modularity, EL-shellability and shellability for semidistributive lattices. This answers open questions of Mühle and Segovia.
Read the full abstract
Machine Learning (Statistics)
Distributed Lag Neural Additive Models
We introduce Distributed Lag Neural Additive Models (DLNAMs), neural-additive analogues of Distributed Lag Non-linear Models (DLNMs) for learning nonlinear effects distributed over lags. DLNAMs replace a prespecified spline cross-basis with neural components that learn exposure–lag response surfaces, avoiding choices of basis family, dimension, and knot placement while preserving additive interpretability and familiar distributed-lag summaries. Exp-centered input layers, smooth activations, and learned subnetwork mixtures produce smooth, locally adaptive representations; pointwise uncertainty combines a conditional last-layer Laplace approximation with between-member ensemble variation. In simulations, DLNAMs generally outperformed DLNM comparators, including penalized and treed variants, in recovering known response functions, with lower bias, stronger boundary recovery, and better-calibrated cumulative intervals; gains were largest for more demanding functions. The architecture performed consistently across sample sizes, outcome families, lag horizons, and jointly fitted multi-exposure settings, retaining recovery performance as exposures were added; fit-specific changes were largely confined to optimization, and applications recovered established empirical patterns.
Read the full abstract
Quantum Physics
Multi-copy and Catalytic Superactivation of Genuine Multipartite Nonlocality
Genuine multipartite Bell nonlocality (GMNL) aims at capturing correlations between many distant observers that are globally Bell nonlocal. A commonly used definition for GMNL is based on biseparable models constructed from non-signalling correlations. Here we uncover an effect of superactivation of GMNL in this framework. Specifically, by locally wiring together two copies of a biseparable correlation, i.e. not GMNL, one can obtain another correlation that is GMNL. We show that this is possible for any number of parties, and for quantum-realizable correlations. Finally, we show that superactivation of GMNL is also possible at the single-copy level, via a catalytic protocol involving only local wirings. These results question the operational meaning of this definition of GMNL.
Read the full abstract
Computation and Language
Beyond Single-Negative Preference: Multi-Negative DPO for LLM-Centric Historical Entity Linking
Large language models (LLMs) have recently shown promise for historical entity linking, but preference optimization for this task is often formulated with only one negative candidate per training instance. This discards information from the remaining candidates retrieved for the same mention. We introduce multi-negative direct preference optimisation (MDPO), a reference-based pairwise objective that compares the correct entity with all valid rejected candidates associated with each mention. MDPO preserves the Bradley-Terry formulation of DPO while exploiting the complete candidate set through masked, length-normalised sequence scores. We evaluate MDPO on hipe-2020 and newseye, covering French, German, English, Swedish, and Finnish historical newspaper text. Experiments show that MDPO improves over supervised fine-tuning and single-negative DPO, with particularly strong gains for NIL mentions, semantic ambiguity, OCR noise, and historically difficult names. Further analyses disentangle candidate-generation and selection errors, showing that candidate retrieval remains a key bottleneck for end-to-end entity linking. These results demonstrate that incorporating all within-instance negative candidates is a simple and effective improvement for LLM-based historical entity linking.
Read the full abstract
Instrumentation and Detectors
Ton-scale Xenon Gas TPC for $0\nu\beta\beta$ Search at Atmospheric Pressure
We explore aspects of an unorthodox ton-scale xenon gas time projection chamber detector operated at normal temperature and pressure (NTP), designed for $0\nu\beta\beta$ discovery at $\sim10^{27}$ year sensitivity. At fixed active mass, a greater transparency to $\gamma$-ray backgrounds and better track clarity exists at NTP relative to higher density. Operation at NTP also alleviates difficulties with pressure containment and significantly reduces HV requirements. With metre-scale electron tracks at a few MeV, event topology can be used to efficiently reject the copious $\gamma$-ray backgrounds by factors ranging from 550–1200, largely compensating for the greater surface area of the detector. Event topology is captured from anode plane signals combined with sensing the secondary ion component arriving at the cathode plane. A molecular gas admixture limits diffusion of the secondary electron tracks to less than one cm rms and provides stable proportional avalanche gain for measurement of energy. An energy resolution $\delta$E/E $\leq$ 1% FWHM seems possible. A 100 kg scale demonstrator for this detector technology could lead directly to a ton-scale search.
Read the full abstract
Rings and Algebras
The Classification of the Rota-Baxter Operators on Four-dimensional Grassmann Algebra
The purpose of this paper is to classify all the Rota-Baxter operators explicitly on Grassmann algebra in dimension four over the reals.
Quantum Physics
Cat-Code-Protected Controlled Quantum Communication via Non-Local CNOT Gates over Star Quantum Networks
Controlled quantum communication enables secure state transfer between a sender and receiver with the assistance of one or more controllers. However, practical implementation over optical fibre networks is severely hindered by amplitude damping, which reduces fidelity exponentially with distance. We address this challenge by combining two powerful techniques: cat-state encoding for error correction and optimal non-local CNOT gates for distributed gate implementation. The protocol eliminates the need for the physical position qubit itself to travel through the optical fibre, reducing damping events. We show, through density-matrix simulations, that our cat-code-protected protocol with a non-local CNOT operation achieves higher fidelity at 50 km, significantly outperforming the standard protocol. We analyse the protocol's security against beam-splitter attacks and show that CHSH tests provide security against beam-splitter attacks on the distributed entanglement resource despite the cat code's error correction. Our results establish that cat-code-protected controlled quantum communication is feasible with current technology and structurally extensible to multiple controllers, providing a theoretical framework for studying error-corrected controlled quantum communication over long-distance star quantum networks.
Read the full abstract
Software Engineering
A Text Mining and Classification Approach for Analyzing Architecture Decision Records
Architectural decision records (ADRs) have become a popular lightweight mechanism for documenting architectural knowledge in software projects. However, there is limited empirical evidence on the kinds of architectural concerns captured in ADRs and how well their contents align with established architectural knowledge concepts and documentation practices. In this paper, we propose an automated text-mining and classification approach for analyzing ADRs at scale. We apply this approach to a dataset of ADRs extracted from ~550 open-source repositories, combining topic modeling, LLM-based classification, and template compliance checks. Our analysis examines decision taxonomies and quality attributes, and the degree to which ADRs adhere to the MADR template. Our findings show that ADRs frequently capture existence, technology, and process-related decisions, while alternatives, decisions drivers, and some quality concerns remain under-documented. We also observe recurring mismatches between ADR contents and template sections. These insights into current documentation practices provide architects with valuable information to reflect on how ADRs are and should be used to effectively deal with architectural knowledge. Furthermore, our automated approach is adaptable to other architectural tasks.
Read the full abstract
High Energy Physics (Phenomenology)
Vector-meson properties in a data-driven AdS/QCD model
We investigate the properties of $\rho$, $\psi$, and $\Upsilon$ mesons within a holographic AdS/QCD framework, capturing both their vacuum phenomenology and in-medium dynamics. By reconstructing flavor-specific gauge kinetic functions within a model originally developed to study thermodynamic functions, we compute the vector-meson mass spectrum, the decay constants and the hadronic vacuum polarization contributions to the anomalous magnetic moment of the muon at zero temperature. Furthermore, we evaluate the thermal spectral functions to analyze mass shifts, resonance broadening, and the melting of vector bound states in a hot and dense medium. Finally, using the Kubo formula, we extract the electrical conductivity of the plasma as a function of temperature and chemical potential.
Read the full abstract
Computer Vision
TRAIL: Trajectory-Aware Visual Place Recognition against Unordered Databases
Modern Visual Place Recognition (VPR) methods excel on standard benchmarks yet remain brittle in feature-poor environments. By treating each query image in isolation, they discard the sequential context in any real trajectory. We formalize a task that exploits this context: given a query sequence, localize the final image against an unordered reference database – which, unlike sequence-to-sequence methods, requires no sequential structure in the database. We propose TRAIL (TRajectory-Aware Image Localization), a principled framework based on Conditional Random Fields (CRF) that combines learned functions for visual similarity and for camera-motion consistency, refining a distribution over candidate references as each query arrives. A lightweight post-processing layer atop any pre-trained VPR backbone, TRAIL improves a state-of-the-art baseline by up to 8.3 percentage points on our primary benchmark, transfers to unseen datasets without retraining, and delivers its largest gains where visual cues are scarce.
Read the full abstract
Chemical Physics
Low-rank approximation of Moment Tensor Potential enables reducing training set size without loss of accuracy
In this study, we implement a low-rank approximation of Moment Tensor Potential (MTP) based on the tensor train (TT) decomposition. The implemented tensor-factorized MTP (TFMTP) and the original MTP model are actively trained via a MaxVol-based algorithm during molecular dynamics simulations of a four-component molten salt mixture, LiF-NaF-KF (FLiNaK), and geometry optimizations of a five-component equiatomic Mo-Nb-Ta-W-V random alloy. We demonstrate that under a 1.5-fold compression, TFMTP requires two times fewer configurations for fitting than the original MTP model, while maintaining an indistinguishable level of accuracy. These actively trained MTP and TFMTP models are further used to evaluate the density and viscosity of FLiNaK at temperatures ranging from 600 to 1200 K, as well as the elastic constants and bulk modulus of the Mo-Nb-Ta-W-V alloy at zero temperature. For both atomic systems, the differences in physical properties predicted by MTP and TFMTP are negligible.