Peer Revenue Returns subject to peer review

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Wednesday 2 September

Cosmology and Nongalactic Astrophysics

Universal CMB Phase Coherence from Single-Field Inflation

The phase coherence of primordial perturbations shapes the acoustic peaks of the CMB and is regarded as a key signature of inflation. Inflationary scenarios at different energy scales, thus with different durations, are commonly expected to leave different levels of coherence. We show this expectation fails: because the curvature perturbation is conserved on super-Hubble scales, the coherence at recombination inherits a value fixed by the observed primordial spectrum alone. This holds irrespective of the inflationary energy scale and the subsequent expansion history.
Read the full abstract

All authors Siméon Vareilles, Thomas Colas, Julien Grain, Federico Piazza, Vincent Vennin

High Energy Astrophysical Phenomena

Unlocking the QPE Mystery: Star-Disk Collisions in Realistic AGN Disks

Quasi-periodic eruptions (QPEs) are luminous, recurring soft X-ray outbursts observed in the nuclei of low-mass galaxies. They display two remarkable trends: outburst durations are $\sim$10-20% of the recurrence timescale, and longer bursts are more luminous. A promising theory that naturally explains the quasi-periodicity invokes collisions between a star on an extreme mass-ratio inspiral (EMRI) orbit and the accretion disk around the supermassive black hole. However, it remains unclear how this model reproduces the observed trends. We therefore carry out two-dimensional, multi-frequency radiation hydrodynamic (RH) simulations of star–disk collisions. Crucially, we adopt a more realistic circumnuclear disk structure from previous Radiation MHD simulations of sub-Eddington accretion disks. We find that the thick, puffed-up disk atmosphere, extending to $z/r\sim1$, causes different portions of the bow shock to break out at different times, producing prolonged thermal emission as the shock emerges through the breakout surface at $z/r\sim0.7$. The X-ray flare duration is set by the shock propagation time through the optically thick disk–a $\sim$10% of the orbital timescale, reproducing the observed duty cycle. A more oblique star-disk interaction yields a longer, more luminous flare. The realistic AGN disk models also exhibit a surface density $\Sigma\propto r^2$, giving a collisional energy $E\propto P^{2/3}$ that may explain the luminosity–period trend, especially for the weaker QPEs. Overall, we suggest that a more realistic circumnuclear disk structure can explain several observed QPE trends-and QPEs may, in turn, constrain the disk structure.
Read the full abstract

All authors Zhaohuan Zhu, Xiaoshan Huang, Yan-Fei Jiang, Shunquan Huang

Astrophysics of Galaxies

Bridging integrated-light and resolved-star studies of Local Group dwarf spheroidals. A novel use of deep imaging based on amateur telescopes

Low-mass galaxies beyond the Local Group can only be studied through their integrated light, yet it remains unclear whether structural parameters inferred from integrated light are directly comparable to those obtained from resolved-star studies. Using new deep imaging obtained with small-aperture amateur telescopes, we analyse the classical dwarf spheroidals Leo II, Sculptor, and Fornax, whose stellar distributions have been extensively characterised through resolved-star studies. We find that the radial light distributions and effective radii inferred from integrated light are in excellent agreement with those derived from resolved stars, showing that resolved-star and integrated-light analyses can be placed within a common observational framework applicable across a broad range of distances. The two approaches are highly complementary: resolved-star analyses trace galaxies to very large radii but are limited by stellar crowding in their central regions, whereas integrated-light observations are insensitive to crowding and allow the inner stellar distribution to be traced to substantially smaller radii. The integrated-light analysis extends the structural analysis to radii up to an order of magnitude smaller than previous star-counting studies while significantly reducing the scatter in the measured profiles.
Read the full abstract

All authors Sergio Guerra Arencibia, Ignacio Trujillo, Mireia Montes, Ignacio Ruiz Cejudo, Carlos Marrero-de la Rosa, Vicente Fontana, Aleix Roig

General Relativity and Quantum Cosmology

Excitation factors and exceptional points of Kerr-de Sitter quasinormal modes

We compute the excitation factors of Kerr-de Sitter black hole quasinormal modes using two different techniques: a Heun function representation and an expansion of the solution in terms of hypergeometric functions. Exceptional points at which the complex frequencies of different overtones coincide are a generic feature of the Kerr-de Sitter quasinormal mode spectrum. Near exceptional points, the modes exhibit a "hysteresis phenomenon" that was previously found for Kerr black holes perturbed by massive fields and the absolute value of the excitation factors of the individual modes is significantly enhanced, while they acquire nearly opposite phases. We extrapolate the excitation factors to small values of the cosmological constant and find good agreement with previous calculations in the Kerr limit.
Read the full abstract

All authors Dario Rossi, Naritaka Oshita, Leonardo Gualtieri, Emanuele Berti

High Energy Physics (Theory)

Gravitational Backreaction in de Sitter: A Canonical ADM Approach

We develop a canonical ADM framework for gravitational backreaction during inflation. Our approach extends the formalism used to compute cosmological correlators by applying it to the evolution of the quantum background, which we identify with the metric one-point function. The aim is to assess the quantum stability of inflationary geometries under gravitational backreaction. Within this framework, the backreaction is determined by the expectation values of the Hamiltonian constraint and the equations of motion for the spatial metric. Working in the exact de Sitter limit, we apply the framework to a massless minimally coupled spectator scalar field and the physical graviton polarizations. At one loop, their backreaction renormalizes the relation between the cosmological constant and the Hubble parameter without inducing any secular departure from de Sitter evolution, in agreement with previous results. We perform the calculation using a hard momentum cutoff and show that a fixed physical cutoff allows the homogeneous background equations to be renormalized with time- and background-independent coefficients, in contrast to a fixed comoving cutoff. Finally, we demonstrate that the explicit representation of the backreaction depends on the gauge condition imposed on the metric fluctuations, as well as on the background-fluctuation split. Nevertheless, these different representations are related by a gauge transformation and are therefore physically equivalent.
Read the full abstract

General Relativity and Quantum Cosmology

Tidally-enhanced resonances in extreme-mass-ratio inspirals: A tertiary path to chaos

Extreme-mass-ratio inspirals (EMRIs) provide a unique laboratory for probing strong-field gravity and complex relativistic dynamics. We study a tidally deformed EMRI composed of a stellar-mass secondary orbiting a supermassive (non-)rotating black hole embedded in an external, adiabatically varying tidal environment. These systems provide a restricted, yet astrophysically motivated, realization of the relativistic three-body problem, expected to appear in active galactic nuclei, with a clear hierarchy of masses and radiation-reaction timescales. The external tidal deformation breaks the axisymmetry of the Kerr spacetime, rendering the geodesic dynamics non-integrable and giving rise to chaotic motion. The resulting signature of non-integrability is characterized through Poincaré maps and rotation curves constructed from the ratios of the fundamental frequencies of bound radial, polar, and azimuthal motion. We identify two prominent plateaus whose widths increase with the tidal field amplitude, signaling a transition from weak to strong chaos. We then demonstrate the sensitivity of the chaotic dynamics to the orientation of the orbit relative to the tidal field. We further analyze the proper-time evolution of the action-angle variables, showing that the angle combinations associated with the dominant commensurabilities are phase locked, thereby allowing the associated tidal contributions to induce secular changes in the constants of motion, whereas off-plateau angle combinations circulate. These results clarify the dynamical significance of the prominent plateaus and provide a novel phase-space characterization of tidal resonances in EMRIs. Finally, we discuss the potential role of radiation-reaction effects in driving EMRIs through tidal island crossings and the implications for gravitational-wave inference with future detectors.
Read the full abstract

All authors Kyriakos Destounis, Takuya Katagiri, Sajal Mukherjee, Kostas D. Kokkotas

Earth and Planetary Astrophysics

Radial velocity detection of the TRAPPIST-1 planetary system with SPIRou and NIRPS

The TRAPPIST-1 system is well-known for its seven transiting Earth-sized exoplanets. It has been extensively studied and characterized, notably with transit timing variations (TTVs) to precisely measure the mass of the planets. Using near-infrared spectroscopic observations obtained as part of the SPIRou Legacy Survey and the NIRPS Guaranteed Time Observation programs, we aimed to verify those values through radial velocity (RV) measurements of the system. Our RV analysis reveals that the current data do not have the precision required to individually detect the TRAPPIST-1 planets. However, we confidently detect ($\Delta\ln\mathcal{Z}=7.53$, 1860:1 odds) the combined RV signature of the planets by informing their relative masses on the TTV analysis, with TRAPPIST-1 b as a proxy of the whole system. For the first time, the RV signal of the TRAPPIST-1 system is recovered: we find a RV semi-amplitude of $K_b=3.65^{+0.78}_{-0.83}$ m s$^{-1}$ corresponding to a planetary mass of $M_{p,\,b}=1.31\pm0.29$ M$_\oplus$, demonstrating that the RV measurements are consistent with the TTV model ($M_{p,\,b;\,\text{TTV}}=1.374\pm0.069$ M$_\oplus$). Additionally, the NIRPS RVs constrain the presence of giant planets beyond the snow line, excluding Saturn-mass planets out to 2.7-yr orbits and Neptune-mass objects out to 20 d. Through RV, we determined the stellar activity period to be of $3.22^{+0.22}_{-0.20}$ d. Its agreement with photometric measurements (K2 and TESS) confirms stellar rotation as the origin of the $\sim3.3$-d periodicity observed for TRAPPIST-1. We further investigated stellar activity with SPIRou polarimetric measurements, placing an upper limit on the longitudinal field ($|B_l|<40$ G, $3\sigma$). This limit is compatible with a weak multipolar large-scale magnetic geometry, as observed in some of the later-type rapidly rotating M dwarfs.
Read the full abstract

All authors Alexandrine L'Heureux, René Doyon, Charles Cadieux, Pascal Petit, Olivia Lim, Étienne Artigau, Neil J. Cook, Eric Agol, Xavier Bonfils, Julien Morin, Stefano Bellotti, Claire Moutou, Leslie Moranta, Jean-François Donati, Luc Arnold, Xavier Delfosse, Guillaume Hébrard, Romain Allart, François Bouchy, Lucile Mignon, Izan de Castro Leão, Dany Mounzer, Khaled Al Moulla, Frédérique Baron, Björn Benneke, Mathis Bouffard, Casey Brinkman, Bruno L. Canto Martins, Andres Carmona, Ryan Cloutier, Marion Cointepas, Nicolas B. Cowan, Eduardo Cristo, Roseane de Lima Gomes, Jose Renan De Medeiros, Xavier Dumusque, Dasaev O. Fontinele, Thierry Forveille, Yolanda G. C. Frensch, Jonathan Gagné, Jonay I. González Hernández, Nicole Gromek, Melissa J. Hobson, Vigneshwaran Krishnamurthy, Pierrot Lamontagne, Lison Malo, Eder Martioli, Yuri S. Messias, Louise D. Nielsen, Ares Osborn, Léna Parc, Caroline Piaulet-Ghorayeb, Nuno C. Santos, Bennett Neil Skinner, Avidaan Srivastava, Atanas K. Stefanov, Alejandro Suárez Mascareño, Gregg Wade, Joost P. Wardenier, Drew Weisserman

Computation and Language

Unifying Conformal Language Tasks with In-Context Ensembles

Many NLP tasks, such as summarization and extractive question answering, reduce to retrieving relevant content from documents under two constraints: coverage, retaining enough pertinent information to achieve some goal, and conciseness, removing as much irrelevant information as possible. Conformal prediction methods have been used to guarantee coverage, and must be optimized for conciseness through design of a score function. State-of-the-art scoring functions use hand-engineered LLM prompts asking the model to rate the importance of content, but manual prompt engineering is labor-intensive and task-specific. We introduce the Conformal Relevance framework which uses in-context learning example curation and ensembling to create a score function which maintains coverage while improving conciseness with minimal manual input. We demonstrate this framework's application on seven NLP tasks, and also theoretically study the impact of diversity for ensembled conformal scores, giving a complementarity condition that characterizes when ensembling improves worst-case sentence scores, and a saturation bound on ensemble improvement.
Read the full abstract

All authors Xiao Shi Huang, Chen-Yuan Lin, Bruce Kuwahara, Kin Kwan Leung, Jesse C. Cresswell

Solar and Stellar Astrophysics

Benchmarking Machine Learning Emulators of Stellar Evolution for Precision Asteroseismology

Fast and accurate stellar evolution emulators—surrogate models that approximate expensive simulation outputs with machine learning (ML)—are powerful tools for modern stellar characterization, hierarchical inference, and population synthesis. We analyze the grid density required for reliable emulation by training ML algorithms on main-sequence models with masses M=[0.7,1.2] solar masses. This range is challenging to emulate due to rapidly varying evolutionary behavior caused by the radiative-to-convective core transition, as well as the requirement to match the part-per-thousand seismic precision that has been delivered for such stars from the NASA Kepler mission. Generating grids from analytical models, as well as MESA, YREC, MIST, and ASTEC, we compare linear interpolation, k-nearest neighbors, random forests, and neural networks (NNs) in interpolating the stellar observables: T_eff, L, Delta nu, and nu_max. While NNs outperform other methods, sparse grids induce localized failures in the core-transition region, resulting in unstable derivatives, ensemble disagreement, and fragmented posterior distributions during inference. Performance gains from denser grids are non-uniform, suggesting that adaptive grid generation should be favored over uniform refinement. Finally, we show that NN ensembles allow for localized uncertainty propagation, more accurately reflecting emulator reliability across parameter space than global uncertainty estimates. As we consider only the two-dimensional case of varying only stellar mass and age along the main sequence, these results represent a lower bound on the challenge in emulating stellar evolution simulations for precision asteroseismology.
Read the full abstract

All authors Naomi Gluck, Earl P. Bellinger, Yan Liang, Ebraheem Farag, Nicholas Saunders, Selim Kalici, Christopher J. Lindsay, Sarbani Basu

Machine Learning

Causal Foundation Models

Causal inference is the practice of estimating the effect of a treatment or intervention from data. It traditionally requires a bespoke pipeline for every new problem: first proposing a causal mechanism, selecting a compatible estimator, and finally training it. Meanwhile, across diverse settings and modalities, much of machine learning has shifted to the paradigm of foundation models: networks pretrained once at scale and applied to new tasks without fine-tuning. Causal foundation models (CFMs) bring this paradigm to causal inference. CFMs are pretrained neural networks that estimate causal quantities, such as the average treatment effect, on entirely new datasets using in-context learning without requiring model updates. This work provides a practical introduction to this emerging area. We summarize the necessary background in causal inference and machine learning before discussing CFMs. Throughout, we include example code and Jupyter notebooks.
Read the full abstract

High Energy Physics (Phenomenology)

A Task Force on Strong Coupling Determinations from Event Shapes

The strong coupling constant $\alpha_s$ is a fundamental parameter of the Standard Model. Its precise determination is essential for accurately predicting, studying, and understanding processes at the Large Hadron Collider and future experiments such as the Future Circular Collider. Event shape and correlator observables measured at electron-positron colliders provide one of the cleanest environments for extracting $\alpha_s$, thanks to their sensitivity to $\alpha_s$ and the availability of high-precision data from the Large Electron-Positron Collider. More broadly, such observables provide an ideal setting to develop and test our understanding of the perturbative and non-perturbative elements of Quantum Chromodynamics, which will underpin the field's precision and discovery frontiers for decades to come. Despite these advances, significant discrepancies persist between different determinations of $\alpha_s$ from event shapes, both in the extracted central values and estimated uncertainties. This document motivates the establishment of a dedicated Task Force to coordinate a community-wide effort addressing these open questions. We report on the first two-day meeting held at CERN in November 2025, summarizing the scientific discussion and documenting the experimental analyses identified as priorities during the meeting, as well as the concrete list of tasks to be carried out by the theory community in preparation for future meetings.
Read the full abstract

All authors Anthony Badea, Luca Buonocore, Gherardo Vita, Simone Alioli, Zeno Capatti, Yi Chen, David d'Enterria, Thomas Gehrmann, Andre H. Hoang, Alexander Huss, Sebastian Jaskiewicz, Alexander Karlberg, Yen-Jie Lee, Michelangelo L. Mangano, Vicent Mateu, Jiahao Miao, Pier Francesco Monni, Paolo Nason, Luca Rottoli, Iain W. Stewart, Paolo Torrielli, Giulia Zanderighi

High Energy Physics (Theory)

Taxonomy of Potentials in Asymptotic Limits

In infinite-distance limits of scalar field moduli spaces in quantum gravity, particle masses, brane tensions, and scalar field potentials scale exponentially with geodesic distance. Previous work has shown that the exponential decay rates of particle masses and brane tensions admit a discrete classification. In this work, we extend this classification to the case of scalar field potentials. We find that leading contributions to the potential typically scale with tensions of codimension-1 branes as $V \sim T_{d-1}^2$ or codimension-0 branes as $V \sim T_d$, and these relations hold formally even when the associated branes are absent from the spectrum. As a result, the taxonomy rules for potentials are intimately connected to the brane taxonomy rules. More generally, potential contributions can be labeled by a collection of integers, which correspond physically to their transformation properties under Weyl rescalings and their order in the string loop expansion. This implies that the vector $\vec v = - \vec \nabla \log V$ is lattice-valued, and indeed it lies in precisely the same lattice as the analogous vectors $\vec \alpha = - \vec \nabla \log T_d$ for codimension-0 branes of tension $T_d$. We verify our taxonomy rules in various examples in string theory. We show that these rules imply that sums of positive potential terms satisfy the Strong Asymptotic de Sitter Conjecture, which requires $|\vec \nabla \log V| \geq 2/\sqrt{d-2}$ in asymptotic regimes of scalar field moduli space, and they imply that higher-derivative gravitational corrections are suppressed by powers of the species scale.
Read the full abstract

High Energy Physics (Phenomenology)

Yukawa coupling deviations from four-quark interactions

New physics generating four-quark interactions can be difficult to detect at the LHC, especially if it does not couple to first generation quarks. We investigate the possible role of such interactions in generating measurable deviations in Higgs-quark couplings. We find important constraints from flavour physics and top-pair production, but show that Higgs-charm and Higgs-bottom couplings remain the most sensitive LHC observables to some four-quark operators involving scalar currents. At FCC-ee, more operators and flavour structures can be tested. We further investigate simple single-particle models that can generate these interactions, including a second Higgs doublet, and scalar and vector diquarks. We find that in some cases, including new scalars most strongly coupled to the third generation of quarks, deviations in Yukawa couplings could offer the first signs of new physics generating four-quark interactions.
Read the full abstract

Quantum Physics

Computing with qLDPC Codes by Climbing the Chain Map Hierarchy

We develop a framework for logical computation with qLDPC codes that places logical Pauli, Clifford, and non-Clifford operations on the same footing. This brings the simple homological description of Pauli logicals to the patchwork landscape of logical Clifford and non-Clifford operations, providing a tool for the discovery of new logical gates. In particular, we define the chain map hierarchy: a family of chain complexes whose homology classes encode logical unitary and code surgery operations at any level of the Clifford hierarchy, precisely analogous to the chain complex description of Pauli logicals. Consequently, intuition for Pauli logicals can be leveraged to discover new logical operations on qLDPC codes. For instance, the familiar ability to deform Pauli logicals with stabilizers—i.e. boundaries of the chain complex—becomes a way to search for constant-depth unitary implementations of (non-)Clifford logical gates. Using this strategy, we discover constant-depth unitary implementations of the full logical Clifford group on any number of blocks of the 2D toric code, including within a single block, and addressable logical CCZ gates on any triple of logical qubits on any number of blocks of the 3D toric code. Beyond manifold codes, we find addressable non-Clifford gates on codes with many encoded qubits. The chain map hierarchy naturally encompasses and extends other constructions of logical gadgets, for instance providing a universal parameterization of cup products. As such, our work provides a unified, useful, and intuitive language for computing with qLDPC codes.
Read the full abstract

Machine Learning

Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillation

On-policy distillation (OPD) accelerates post-training by providing dense token-level supervision from a frozen teacher on the student's own rollouts. Vanilla OPD applies this supervision uniformly across prompts, without checking whether the teacher is reliable for each prompt. Because reverse KL is mode-seeking, a confidently wrong teacher can induce a strong yet misleading update. Distributional proxies, such as entropy or teacher-student likelihood agreement, measure uncertainty or agreement but do not directly verify outcome correctness. We introduce Teacher-Gated On-Policy Distillation (TGOPD), built on the principle that teacher reliability should be verified at the prompt level before dense supervision is admitted. TGOPD estimates reliability from a small set of verifier-scored teacher probes and routes each prompt exclusively to dense OPD when the reliability check passes or to verifier-grounded GRPO otherwise. Across 4B and 35B students in mathematics, code, and instruction following, TGOPD outperforms Vanilla OPD in all six single-domain settings and achieves higher seven-benchmark averages at both scales under multi-domain training. By using otherwise-idle teacher capacity for reliability estimation, TGOPD also reduces teacher-side compute waste in asynchronous OPD, increasing teacher-node GPU utilization from 9.8% to 78.9% in the measured 4B single-domain run.
Read the full abstract

All authors Zhiwei Zhang, Zechen Sun, Fei Zhao, Kang Peng, Bin Liang, Huayu Deng, Yao Hu, Kam-Fai Wong, Mu Chuan

Logic

Finite-index extensions of essentially free countable Borel equivalence relations

We show that finite-index extensions of essentially free countable Borel equivalence relations are essentially free, answering a question of Kechris.

Machine Learning

Evaluating Graph Neural Networks for Change-Criticality Classification in Maritime Navigation Charts

Graph neural networks (GNNs) are a class of neural networks suitable for learning on graph-structured data. Their application to spatial data is a natural extension, however its relatively unclear which message-passing operations, architectural configurations, and graph representation is best suited for classifying changes to objects in electronic navigational charts (ENCs)–geospatial vector datasets used for marine navigation. Maintaining these datasets is a challenge, and categorizing changes to objects in the ENC based on their significance to navigational safety is of particular importance. Here, we propose to represent these vector navigation datasets as a graph structure where the spatial objects serve as nodes and their spatial and semantic relationships form edges. We encode both the old ENC dataset and new ENC dataset into a pair of graphs and frame the task as a graph-pair classification problem. Building on this representation, we investigate the use of GNN architectures to classify whether the encoded graphs constitutes a critical or non-critical risk to navigational safety. We train and evaluate several GNN architectures and model configurations on ENC changes reviewed by maritime experts. Our results demonstrate that graph-based representations improve the classification of ENC updates, providing a scalable approach for automating or improving ENC maintenance workflows.
Read the full abstract

Superconductivity

Direct Validation of Superconductivity through Contact-Free Detection of Persistent Supercurrents Using Room-Temperature Quantum Magnetometry

The accelerated emergence of new materials is driving the search for high temperature superconductors, but rapid experimental validation remains a critical bottleneck, particularly for microscopic samples under high pressure. Here, we demonstrate one-step direct superconductivity validation through room-temperature, contact-free detection of remnant supercurrents. The technique utilizes a cryogen-free optically pumped atomic magnetometer to detect the temperature-dependent magnetic field produced by supercurrents of a superconductor. The superconducting transition is directly identified by the abrupt disappearance of magnetic field from the remnant supercurrent above the transition temperature and the reversal of the supercurrent direction upon reversing the applied magnetic field. Validated on YBCO microcrystals and REBCO tape, this technique detects pico-Tesla magnetic fields from supercurrents induced by the ambient Earth's magnetic field in a millimeter-sized REBCO square disk, as well as from sub-100 micrometer YBCO microcrystals compatible with high-pressure diamond anvil cells. The use of a ferrite flux guide enables sensitive detection from centimeter-scale distances. Requiring no electrical contacts, magnetic coils, or integrated magnetic sensors, this non-invasive, room-temperature platform offers a scalable approach for high-throughput screening and validation of superconductivity.
Read the full abstract

All authors Xinyi Zeng, Chengzhen Qin, Bowen Fan, Hammad Ahmed, Hui Fang, Stuart Long, Xiaonan Shan, Jiefu Chen, Shoujun Xu, Liangzi Deng, Ching W. Chu, Jiming Bao

High Energy Physics (Phenomenology)

Sub-TeV Singlino Dark Matter in light from Sagittarius A$^\ast$ and LUX-ZEPLIN Nuclear-Recoil Event

We investigate the impact of a dark matter density spike surrounding the Milky Way's supermassive black hole (SMBH) on the detectability of Singlino-dominated neutralino dark matter within the Next-to-Minimal Supersymmetric Standard Model (NMSSM). Similar density enhancements, or mini-spikes, around stellar-mass black holes (sBHs), have also been considered. Such a dark matter (DM) candidate typically produces weak indirect detection signals in conventional dark matter halos. Additionally, a Singlino-like lightest supersymmetric particle (LSP) is very difficult to probe at the LHC or through direct DM searches. Furthermore, a Singlino-like LSP with an almost degenerate Higgsino NLSP can produce an observable inelastic DD signal and is compatible with the anomalous high-recoil event recently reported by the LZ collaboration. Keeping these in consideration, we examine the prospects for detecting these sub-TeV dark matter scenarios through gamma-ray observations of the regions surrounding the SMBH Sgr A$^\ast$ and the sBH in the low-mass X-ray binary XTE J1118+480.
Read the full abstract

All authors Utpal Chattopadhyay, Debottam Das, Rahul Puri, Joydeep Roy

Machine Learning

No-Regret Bayesian Optimization with Finite-Library Input-Warped Kernels

Gaussian-process Bayesian optimization (GP-BO) excels at black-box optimization of costly functions, e.g., hyperparameter optimization (HPO) and multi-agent system (MAS) design. Convergence-rate guarantees exist for select methods, notably GP upper confidence bound (GP-UCB), but require a fixed kernel. Critically, the kernel encodes how input proximity affects objective value similarity. When raw coordinates poorly match this geometry - as with log-scaled hyperparameters or localized peaks - input warping can greatly improve sample efficiency, yet known GP-UCB proofs require a fixed kernel. We propose Finite-Library Input-Warped Bayesian Optimization (FLIWBO), which selects warps from a finite library of smooth input maps by any history-dependent rule. It adapts the input geometry to accelerate learning while retaining high-probability convergence guarantees under mild hypotheses, with an explicit $\sqrt(N_\varepsilon)$ library-size cost. Controlled diagnostics show that finite-library warping repairs planted geometry mismatches and identify FLIWBO failure cases. Across four repeated benchmarks - warped synthetic objectives, a confidence-fence trap, and Fashion-MNIST HPO - FLIWBO-UCB beats raw-coordinate GP-UCB under misspecified geometry, escapes traps that defeat even oracle-warp expected improvement, and recovers much of the gain from manual log scaling, while leading the tested methods that admit a matching regret guarantee. A 20-dimensional MAS design study further shows feasibility under costly noisy evaluations. Code for experiments is available: https://github.com/edvin-ketabati/bogp-paper-experiments.
Read the full abstract

General Finance

Tempting the Agent: The Economics of Reputation without Persistent Identity in AI Agent Markets

Reputation is a fundamental mechanism through which markets sustain trust when service quality cannot be perfectly assessed ex ante, constituting a form of intertemporal economic capital by attracting future demand. Its effectiveness as a disciplinary mechanism depends not only on past interactions but also on the persistence of the identity to which reputation is attached. When identities can be abandoned and recreated cheaply, reputational capital may itself become an object of opportunistic exploitation. This paper develops a dynamic economic framework to study when reputation is sufficient to discipline autonomous agents. We model reputation as capital attracting future economic activity. At each point, an agent chooses between operating honestly, investing in quality to preserve future gains, or executing a one-shot deviation to extract its reputation's value and restart from a penalized identity. Our analysis relates the temptation to opportunistic behavior to identity-reset costs, reputation persistence, demand sensitivity, and enforcement design, deriving comparative statics on optimal quality provision. Autonomous AI-agent operating on the blockchain are a relevant application: infrastructures such as ERC-8004, ERC-8183, and x402 combine reputation, identity, and payments in permissionless markets. Nonetheless, our framework applies to any environment where reputation generates future business and identities are replaceable.
Read the full abstract

Machine Learning

TRACE: A spatiotemporal contact memory graph network simulator for granular dynamics

Learned graph simulators provide an efficient alternative to high-fidelity solvers for granular dynamics. However, granular motion depends strongly on inter-granular contact history, which is difficult to preserve when particle contacts form, break, and rearrange. Existing simulators mainly store temporal information in node features or node-level memory. Here we introduce TRACE, a graph-network simulator that stores interaction history directly on contact edges. Each edge maintains a persistent memory updated by attention-based message passing and a gated recurrent unit, while an edge-identity dictionary preserves this memory as the contact graph changes. A physics-structured decoder predicts inter-granular normal and tangential contact forces, enforces the Coulomb friction limit, and applies equal-and-opposite internal forces. The model is trained with single-step pretraining followed by autoregressive rollout fine-tuning. We evaluate TRACE on 2D and 3D granular column-collapse benchmarks. In both cases, TRACE produces stable, physically consistent long-horizon rollouts, closely reproducing the final deposit geometry and the kinetic energy released during collapse. Compared with graph network simulator (GNS) and node-memory graph neural simulator (NMGNS), TRACE reduces long-rollout position error by 31-62% and final-deposit error by 58-89% across the two benchmarks, while using fewer parameters and maintaining near-zero particle interpenetration. TRACE also achieves 12.2$\times$ and 8.9$\times$ speedups over the material point method (MPM) reference solver in 2D and 3D, respectively. Our code is available at https://github.com/Data-Driven-Computational-Geotechnics/TRACE.
Read the full abstract

All authors Changjian Zhou, Negin Yousefpour, Jie Qi, Junfeng Fang, Guillermo A. Narsilio, Hans Petter Jostad

Social and Information Networks

Toward Collective-Centric Evaluation of Preference Inference for Participatory Democracy

To scale up collective decision-making, participatory democracy platforms such as Polis and Remesh enable online deliberation among thousands of participants. However, at this scale, participants cannot review every opinion submitted by others, producing highly sparse voting data that misrepresent patterns of consensus, conflict, and minority support. Platforms therefore increasingly rely on Preference Inference (PI) models to predict missing votes. Yet this automation is not neutral: inferred preferences can artificially amplify, suppress, or reorder existing patterns of support, ultimately reshaping how the outcomes of a deliberation are interpreted. More generally, we lack a systematic understanding of how existing PI methods affect the collective preference landscape. To address this gap, we benchmark several existing PI approaches in this context. Moving beyond conventional user-centric evaluations centered on the accuracy of individual predictions, we introduce a collective-centric evaluation framework that measures whether inferred votes preserve salient properties of the broader preference landscape. We further contribute the largest multilingual dataset of its kind: four consultations spanning over 90k participants, 1M votes, and 22 languages. Our experiments show that models with comparable predictive accuracy can differ substantially in the degree to which they preserve the collective structure. These results demonstrate that accuracy alone is insufficient for evaluating PI in democratic settings. By contributing a novel comprehensive and collective-centric evaluation benchmark for the task of PI, this work aims to support the development of AI systems that scale deliberation without compromising the integrity of its democratic outcomes.
Read the full abstract

All authors Pierre-Antoine Lequeu, Salim Hafid, Paul Lerner, Nazanin Shafiabadi, Laurène Cave, David Mas, Jean-Philippe Cointet, Benjamin Piwowarski, François Yvon

High Energy Physics (Experiment)

Measurement of Differential Cross Sections and Integrated Luminosity using Proton-Proton Elastic Scattering with HADES at T = 4.53 GeV and T = 1.60 GeV

This work presents an investigation of elastic proton-proton scattering based on data collected with the recently upgraded HADES detector using beams of protons with kinetic energies of $T = 4.53$ GeV and $T = 1.60$ GeV impinging on a liquid hydrogen target. This is the first measurement emerging from the upgraded detector setup, including a Forward Detector with straw-tube tracking planes designed for the PANDA experiment. The well-constrained two-body kinematic relations of elastic scattering have been exploited to study the detector alignment, the beam alignment, and to provide the final measurement of the beam energy. In addition, the recorded luminosity of the experiment has been determined by normalizing the HADES data in a range of the squared four-momentum transfer $t$ that overlaps with measurements by previous experiments worldwide. Furthermore, the differential cross sections have been determined over a large range of $t$ from which the slope parameter $B$ is extracted, yielding $B = 8.98 \pm 0.3$ (GeV/$c$)$^{-2}$ and $7.3 \pm 0.1$ (GeV/$c$)$^{-2}$ at $T = 4.53$ GeV and $1.60$ GeV, respectively. This contribution to the world database on elastic proton-proton scattering provides input to theoretical modeling and serves as a reference for luminosity determinations by future experiments in the same energy range.
Read the full abstract

All authors HADES Collaboration: R. Abou Yassine, J. Adamczewski-Musch, G. A. Appagere, M. Becker, A. Blanco, C. Blume, M. Bohman, Y. Bondar, L. Chlad, I. Ciepał, S. Deb, M. D. Doncel Monasterio, M. Duerr, W. Esmail, L. Fabbietti, M. Firlej, T. Fiutowski, A. M. Foda, J. Förtsch, P. Fonte, J. Friese, I. Fröhlich, T. Galatyuk, R. Gernhäuser, M. Grunwald, D. Grzonka, M. Gumberidze, S. Harabasz, T. Heinz, C. Höhne, F. Hojeij, R. Holzmann, M. Idzik, B. Kämpfer, K.-H. Kampert, B. Kardan, V. Kedych, S. Kim, V. Kladov, A. Kodym, M. Kohls, J. Kolas, G. Korcyl, G. Kornakov, W. Krueger, A. Kugler, R. Lalik, S. Lebedev, T. Leontiou, S. Linev, F. Linz, L. Lopes, M. Lorenz, A. Malige, P. Marciniewski, J. Markert, T. Matulewicz, J.G. Messchendorp, V. Metag, J. Michel, A. Molenda, J. Moron, C. Müntz, A. Mukherjee, M. Nabroth, A. Opíchal, J. Orliński, J.-H. Otto, M. Papenbrock, Y. Parpottas, M. Parschau, S. Pattnaik, C. Pauly, D. Pawlowska-Szymanska, V. Pechenov, O. Pechenova, G. Perez Andrade, J. Phan, K. Piasecki, J. Pietraszko, T. Povar, M. Predota, K. Prościński, A. Prozorov, W. Przygoda, B. Ramstein, N. Rathod, J. T. Rieger, J. Ritman, A. Rost, A. Rustamov, S. Sahu, P. Salabura, J. Saraiva, K. Scharmann, N. Schild, K. Schönning, E. Schwab, F. Seck, I. Selyuzhenkov, J. Smyrski, M. Sobiella, S. Spies, A. Sreejith, A. Strach, H. Ströbele, J. Stroth, P. Subramani, K. Sumara, O. Svoboda, K. Swientek, J. Taylor, P. E. Tegner, P. Tlusty, M. Traxler, S. Treliński, I. C. Udrea, F. Ulrich-Pur, V. Wagner, A.A. Weber, C. Wendisch, D. Wielanek, P. Wintz, A. Władyszewska, H.P. Zbroszczyk, M. Zieliński, P. Zumbruch

Machine Learning

Mesh-Native Physics-Informed Graph Surrogates for TCAD-in-the-Loop Design Space Exploration

High-fidelity TCAD simulation of drift-diffusion transport remains the workhorse of emerging FinFET device design, but it is computationally expensive, especially for 3D structures where runtime escalates steeply with mesh complexity. This sharply limits multi-objective design space exploration. Existing machine-learning surrogates map a fixed set of design parameters to a few scalar device metrics, discarding the underlying physics and losing transferability across device geometries and families. A physics-informed graph attention network (GAT) surrogate is proposed. It operates directly on the tetrahedral TCAD mesh and predicts, at every mesh node, the electrostatic potential together with the electron and hole quasi-Fermi levels, the fundamental unknowns of the drift-diffusion system. Training combines a data loss with finite-volume current-continuity residuals, embedding carrier-transport physics into the objective. Operating on the mesh as a graph, the surrogate inherits size generalization: a model trained on few-fin meshes applies unchanged to substantially larger arrays, bounded at inference only by GPU memory. Per-node uncertainty from a deep ensemble drives an active-learning loop that screens large candidate pools in seconds and forwards only the most informative designs for full simulation. Benchmarked against Sentaurus Device on multi-fin tri-gate FinFETs, the surrogate reproduces the three drift-diffusion fields with sub-volt per-field RMSE and reaches a per-design throughput orders of magnitude higher than the full simulator. The advantage grows with device size: on large multi-fin arrays that are prohibitively slow to simulate directly, inference still completes in under a second per device, enabling Pareto-front exploration across device scales infeasible for direct TCAD sweeps.
Read the full abstract

Machine Learning

Tail-Likelihood Reinforcement Learning

Reinforcement learning typically optimizes average reward. For generative policies, the average can hide an important distinction: two policies can achieve the same mean reward while having very different chances of producing a rare but high-reward rollout. This matters as sampling increases during training and inference, since its benefit depends on retaining probability mass on high-reward outcomes. We propose to optimize this coverage directly. Rather than considering only expected reward, we consider all of its upper tails: for each reward threshold, how likely is the policy to exceed it? This turns a continuous reward into a family of binary success events. We introduce Tail-Likelihood Reinforcement Learning (TailRL), which maximizes the log-probability of exceeding a randomly chosen reward threshold. Its gradient gives more weight to rare, high-reward rollouts and can be interpreted as a mixture of Best-of-k gradients. TailRL requires only a simple modification to the advantage function, making it compatible with existing reinforcement learning pipelines. Across object localization, maze navigation, GUI grounding, and code optimization, TailRL leverages rare high-reward training samples to avoid suboptimal solutions and yields models that benefit more from additional samples at inference time.
Read the full abstract

All authors Shrinivas Ramasubramanian, Daman Arora, Fahim Tajwar, Guanning Zeng, Qingyang Wu, Zhongzhu Zhou, Chenfeng Xu, Haiwen Feng, Yuda Song, Aarti Singh, Ruslan Salakhutdinov, J. Andrew Bagnell, Jeff Schneider, Andrea Zanette

Machine Learning

Modern Transformers Are Implicit Hybrids: From Functional Differentiation to Principled Hybrid Architecture Design

Hybrid architectures combining Full Attention (FA) and Linear Attention (LA) are increasingly prominent, yet their allocation remains heuristic. We seek an evidence-grounded basis in head-level functional organization learned by RoPE-based Transformers. Behavioral probes do not yield a complete taxonomy, so we propose two intervention metrics: RoPE Frequency Importance Score (RFIS), measuring how each frequency affects a head's attention distribution, and RoPE Positional Dependence (RPD), isolating dependence on rotary positional modulation. On Qwen3-series models and Llama3.1, RFIS suggests and RPD verifies a complete taxonomy of retrieval and positional heads separated by a salient mid-low-frequency band. Controlled Transformers show that this boundary follows the training-length positional scale; we term it the Global Positional Band (GPBand). The analysis suggests a potential cause of zero-shot length-extrapolation failure and yields two principles: positional modeling should operate only locally, with global access through position-independent retrieval; and both functions should be assigned at head granularity with layer-specific allocation. We instantiate them in Head-wise Hybrid Architecture (HwH), using NoPE FA for global retrieval and LA for local positional modeling. With an FA-to-LA ratio below 1:3, HwH retains strong language modeling and commonsense reasoning while improving retrieval and substantially strengthening zero-shot long-context extrapolation over Transformer, LA, and a layer-wise hybrid baseline. Ablations validate both principles and component roles, highlighting principled hybrid architecture design as a promising route toward future foundation models.
Read the full abstract

Quantitative Methods

Sparse concept attribution for histomorphological hypothesis generation from whole-slide classifiers

Histology images contain rich morphological information and can provide insights into pathological processes. However, deriving hypotheses relating morphological phenotypes to clinical attributes is bottlenecked by a manual image interpretation step. Here, we demonstrate that this process can be automated through interpretable deep learning. We present SCOPE, a method to interpret slide-level classifiers by combining pathology-specific vision–language models with sparse concept attribution onto a generalist histomorphological concept bank. To measure whether such explanations recover known morphology, we introduce MorphoRecoveryBench, a benchmark of seven tasks with pathologist-curated reference descriptions. On this benchmark, dense concept attribution is indistinguishable from a random baseline, whereas sparse attribution recovers substantial known morphology; decomposing the pooled slide embedding reaches similar explanation correctness at a fraction of the computational cost. Post-hoc interpretation of whole-slide classifiers can thus generate morphological hypotheses at scale, for expert validation.
Read the full abstract

All authors Tristan Lazard, Kenza Bouzid, Julius Hense, Shruthi Bannur, Daniel Coelho de Castro, Daniel Shao, Rajesh Jena, Drew Williamson, Stephanie Hyland

Machine Learning

From Euclidean to Graph-Structured Data: A Survey of Collaborative Learning

The conventional approach to machine learning, that is, collecting data, training models, and performing inference in a single location, faces fundamental limitations, including scalability and privacy, that restrict its applicability. To address these challenges, recent research has explored collaborative learning approaches, including federated learning and decentralized learning, where individual agents perform training and inference locally, with limited collaboration. Most collaborative learning research focuses on Euclidean data with regular, grid-like structure (e.g., images, text). However, these approaches fail to capture the relational patterns in many real-world applications, best represented by graphs. Learning on graphs relies on message-passing mechanisms to propagate information between connected nodes, making it conceptually well-suited for collaborative environments where agents must exchange information. Yet, the opportunities and challenges of learning on graph-structured data in collaborative settings remain largely underexplored. This survey provides a comprehensive investigation of collaborative learning from Euclidean to graph-structured data, aiming to consolidate this emerging field. We begin by reviewing its foundational principles for Euclidean data, organizing them along three core dimensions: learning effectiveness, efficiency, and privacy preservation. We then extend the discussion to graph-structured data, introducing a taxonomy of graph distribution scenarios, characterizing associated statistical heterogeneities, and developing standardized problem formulations and algorithmic frameworks. Finally, we systematically identify open challenges and promising research directions.
Read the full abstract

Cryptography and Security

Boundary-Mutation Testing for Pattern-Based Secret Detection: A Rule-Level Method and Cross-Scanner Evaluation

Pattern-based secret scanners are commonly validated with example-based fixtures that fix one variable: the text surrounding a credential. We introduce boundary-mutation testing to vary that context, generating credentials from each rule's own regular expression, embedding them in realistic source contexts, and classifying outcomes at the rule level rather than the tool level, yielding three detection metrics. Applied to three scanners - a 43-rule open-source scanner, Gitleaks 8.21.2, and TruffleHog 3.82.13 - detection in the primary subject holds at >=0.9976 across ten contexts but collapses to 0.5233 when a credential ends in a hyphen. Five rules are affected: two, with fixed-count quantifiers, fail totally and deterministically; three, with variable-count quantifiers, backtrack and match a truncated credential; an entropy fallback rescues some failures but downgrades their severity. We validate a repair restoring full robustness with no new false positives, and report a caution: a plausible first attempt silently regressed two rules, caught only by re-running the same battery. Gitleaks has an unrelated, source-confirmed defect - a hard-coded terminator allowlist causing total misses for most credential types - while TruffleHog shows no boundary fragility but narrowest coverage. We report marginal, not conditional, failure probabilities: one common token format is structurally immune, another fails once in 64; on 292,527 lines of real code, the false-positive ordering inverts relative to the synthetic corpus. Because per-type detection is deterministic, the comparison unit is ten credential types, not hundreds of samples; no recall difference reaches significance, so we report the null result, not a ranking.
Read the full abstract