Peer Revenue Impact factor: volatile

Papers

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

All of arXiv

Monday 9 November

Earth and Planetary Astrophysics

Searching for Planets around Altair with JWST/NIRcam

Altair (alpha Aql) is one of the nearest stars to the Sun at 5.13 pc and its age has recently been revised to 88+/-10 Myr, making it one of the youngest stars in the Solar neighborhood. This makes Altair a compelling target for high-contrast direct imaging searches for giant planets. We present two epochs of simultaneous F444W and F200W MASK430R JWST/NIRCam coronagraphic imaging observations. After post-processing, these data achieve F444W contrast levels of 2 x 10^-6 at 1" (5.1 au), 1.5 x 10^-7 at 2" (10.3 au), and a few x 10^-8 beyond 4"(20 au). Using the Sonora Flame Skimmer models, this corresponds to sub-Jupiter mass sensitivity beyond 5 au and sub-Saturn mass sensitivity beyond 10 au. We identified several candidate point sources in the field in the first epoch data. Second epoch observations confirmed that none of the candidates were co-moving with Altair and their colors are consistent with background stars. Altair has no known companions, no detected debris disk, and is not associated with any stellar association. Despite this, its youth and proximity ensure it will remain a high-priority target for future high-contrast imaging observations. We additionally discuss Altair as a potential reference star for future observations, presenting measurements of background sources in the field and showing that Altair will remain at least a few arcsec from background stars over the next 10-20 years, making it a reliable reference target for upcoming imaging campaigns including the Roman Space Telescope Coronagraph Instrument.
Read the full abstract

All authors Tiffany Meshkat, Charles Beichman, Jorge Llop-Sayson, Geoffrey Bryden, Alexandra Z. Greenbaum, Eric E. Mamajek, Marie Ygouf, Julien Girard, Jarron Leisenring, Aniket Sanghi, Marshall Perrin, Michael Meyer, Laurent Pueyo, William Balmer, András Gáspár, Klaus Hodapp

Cryptography and Security

Useful Work Can Increase Energy Use: Demand Elasticity and Proof of Useful Work

A serious criticism of Bitcoin's Proof-of-Work is its enormous electricity consumption. Proof-of-Useful-Work (PoUW), introduced by Ball, Rosen, Sabin, and Vasudevan (2017), in which mining performs useful computations, was suggested as an alternative to reduce electricity expenditure, with matrix multiplication (MatMul) as a recent concrete instantiation by Komargodski and Weinstein (2026). Perhaps surprisingly, we show that switching to PoUW may increase the total electricity consumption, echoing Jevons' paradox. Intuitively, the PoUW mechanism acts as a subsidy that lowers the effective price of the useful work. Recall that demand is elastic if a discount of x percent in the price leads to an increase of more than x percent in the quantity demanded. If demand is elastic, the resulting increase in quantity demanded leads to a net increase in total electricity consumption; if inelastic, total electricity consumption decreases. Even then, the electricity consumption due to PoUW never vanishes, since the mining subsidy is spent on useful work that would not have been produced without the subsidy. This result is proved in a simple model with electricity-only costs, no PoUW overhead, and an unchanged mining subsidy. The analysis extends to less restrictive settings with adjusted thresholds.
Read the full abstract

High Energy Physics (Theory)

Scale separation and partial supersymmetry breaking

We propose new candidate string theory vacua in which partial supersymmetry breaking occurs within the lower-dimensional effective theory in anti-de Sitter space, parametrically below the Kaluza–Klein scale. These vacua solve the equations of motion of massless type IIA supergravity in the smeared approximation and exhibit parametric scale separation, at both weak and strong string coupling. They are reproduced as $\mathcal N=1$ supersymmetric phases of a three-dimensional $\mathcal N=2$ gauged supergravity. Should these vacua be solutions of string theory, they would provide explicit setups in which partial supersymmetry breaking can be studied at energies parametrically below the compactification scale. In addition, we further comment on our previously proposed solution with scale separation and extended supersymmetry.
Read the full abstract

Number Theory

The uncertainty operator, the Weil ground state and the zeros of zeta: a numerical study and conditional positivity of the Weil quadratic form

Applying the Dirichlet principle to the Weil quadratic form $QW_\lambda$ on $L^2([\lambda^{-1},\lambda],d^*u)$–using only the primes $p\le\lambda^2$ of the Euler product–yields canonical finite-dimensional Hermitian matrices whose ground state has a Mellin transform with zeros proved real in full generality when the lowest eigenvalue is simple and even. While $\lambda^2=13$ gave $50$ zeros of $\zeta$, $\lambda^2=250$ ($\lambda\sim16$) gives the first $1000$ zeros with $>500$ correct decimals (the first with $1341$). We explain this via a correspondence between $QW_\lambda$ and Slepian's time-frequency uncertainty operator $U_\lambda=2-P_\lambda-\widehat P_\lambda$, diagonalized by prolate wave functions with eigenvalues $\delta_n(\lambda)\sim c_n\lambda^{2n+1}e^{-4\pi\lambda^2}$ (Fuchs). Via $\mathcal{E}(f)(u)=u^{1/2}\sum_{n\ge1}f(nu)$ and Gevrey smoothing, $\epsilon_\lambda\le C\delta_4(\lambda)(\log\lambda)^3$ is a theorem (Connes-Kastoris). Computations up to $\lambda^2=300$ ($\epsilon_\lambda(1)\sim10^{-1618}$) show a striking logarithmic coincidence with the Fuchs ladder, confirming this simplicity and evenness. Prolate vectors reproduce all minuscule eigenvectors, and conditioning by them and the pole term removes the minuscule spectrum: $QW_\lambda$ is a direct sum of a finite-dimensional near-radical part (Fuchs ladder) and an infinite-dimensional positive part from the Sonin space ($U_\lambda=2$). Under this ladder hypothesis and $QW_\lambda\ge0$, we bound the ground-state distance to the target with Mellin transform $\frac14\xi(s+\frac12)$ (Hurwitz' theorem). Finally, the semilocal trace formula gives the uncertainty principle via Coburn's lemma and proves, under a hypothesis valid at $\infty$, that $QW_\lambda\ge c>0$ outside a finite-dimensional near-self-dual space, reducing Weil positivity to a Schur complement at the ladder's edge.
Read the full abstract

Solar and Stellar Astrophysics

Planetary material is a major source of hydrogen in helium-atmosphere white dwarfs

Many white dwarfs with helium-dominated atmospheres show traces of hydrogen in their spectra. On average, the total mass of hydrogen within their convective envelopes increases with age, but there is no consensus on the origin of such hydrogen. Hypotheses include upward diffusion from hidden reservoirs deep within the envelope, or accretion, alongside heavier elements, of planetary material. We construct a simple model that includes both pathways, and test it using two independent samples of helium-atmosphere DB white dwarfs from the DESI and SDSS spectroscopic surveys, selected to have Gaia colours BP-RP > -0.25, corresponding roughly to 10000 < $T_{\rm eff}$ < 19000K. We find that at least $22 \pm 2$ per cent of the helium-atmosphere white dwarfs with detectable hydrogen must have accreted some of it from their planetary systems, via ices, hydrated minerals or atmospheres. We thus demonstrate that planetary hydrogen is a necessary ingredient to explain the spectral evolution of white dwarfs as a population, rather than only extreme hydrogen-enriched cases.
Read the full abstract

All authors Paula Izquierdo, Andrew Swan, Christopher J. Manser, Boris T. Gänsicke, Joan Najita, Jessica N. Aguilar, Steven Ahlen, Leandro Beraldo e Silva, Davide Bianchi, David Brooks, Todd Claybaugh, Kyle Dawson, Axel de la Macorra, Peter Doel, Andreu Font-Ribera, Jaime E. Forero-Romero, Enrique Gaztañaga, Satya Gontcho A Gontcho, Gaston Gutierrez, Klaus Honscheid, Robert Kehoe, Sergey E. Koposov, Andrew Lambert, Martin Landriau, Laurent Le Guillou, Ting S. Li, Aaron Meisner, Ramon Miquel, John Moustakas, Francisco Prada, Ignasi Pérez-Ràfols, Alexander H. Riley, Graziano Rossi, Eusebio Sanchez, Michael Schubnell, David Sprayberry, Gregory Tarlé, Benjamin A. Weaver, Hu Zou

Artificial Intelligence

Activation-Guided Graph Retrieval for Cognitive Memory Reconstruction in Language Model Systems

Retrieval-augmented language model systems generally select context by query-conditioned relevance. We argue that long-lived AI systems require a second operator: activation-conditioned memory availability. We specify an architecture in which request and goal cues seed an authorized memory graph, activation and fan attenuation allocate a bounded traversal frontier, retrieved entities recursively emit cues, and a typed evidence graph is reconstructed before generation. The specification yields five architectural consequences about candidate selection, traversal control, graph regularization, evidence reconstruction, and reinforcement. We derive implementation-level cost estimates and define ablations, failure conditions, and a reproducible evaluation protocol. The central claim is narrowed or falsified if a simpler graph reranker matches full traversal on causal and temporal recovery under equal budgets. This architecture-and-analysis preprint reports no benchmark measurements. If the claim holds, retrieval returns the evidence path that makes a memory available, not only a ranked list.
Read the full abstract

Algebraic Geometry

Reciprocal mirror map derivatives, euler-twisted Gromov-Witten invariants, and instanton counts via quantum Serre duality

We study mirror map derivatives as a means of relating the genus zero Gromov-Witten theories of an anticanonical Calabi-Yau hypersurface in a smooth, almost minimal Fano variety and the local Calabi-Yau formed from the total space of the Fano's canonical bundle with the euler-twisted theory of the ambient Fano. A precise relationship between reciprocal mirror map derivatives is obtained from the non-equivariant limit of quantum Serre duality for the euler-twisted theory of the ambient Fano as formulated by Iritani, Mann, & Mignon. This allows us to express the generating series of 3-point euler-twisted invariants as the derivative of the local mirror map with respect to the closed one. In low dimensions, this series is shown to be a modular form of Eisenstein type which has arisen before in this area. We prove integrality of the twisted invariants from this identification. The regularized small quantum $D$-module of the Fano is reducible, and we show it is equivalent to an inhomogeneous Picard-Fuchs equation governed by this generating series. The reciprocal mirror map derivative is then shown to be a generating function of local 3-point invariants, which recovers some known results for local Calabi-Yau threefolds and proves a conjecture of the second named author and A. Malmendier in previous work on local fourfolds. Explicit calculations for $\mathbb{P}^2,\mathbb{P}^3$ and $\mathbb{P}^4$ illustrate these relationships, which includes the vanishing of positive-degree ordinary genus-zero invariants for anticanonical elliptic curves and K3 surfaces, the recovery of the quintic threefold's instanton numbers, and finally some formulas for genus-zero invariants of local projective spaces.
Read the full abstract

Software Engineering

MiniZoo: An Evolving Dataset of Mini Apps for the Research Community

Mini apps run within host applications (apps) such as WeChat, Alipay, Baidu, and Telegram, providing services without requiring users to install separate native apps. Their growing ecosystem has attracted research on software quality, security, and privacy, creating a need for reusable datasets. However, building and maintaining such datasets is challenging, requiring platform-specific collection infrastructure and mechanisms to handle incomplete packages, changing availability, and heterogeneous metadata. A shared, continuously maintained collection is still lacking, therefore motivating the development of a common research resource, whose value has been demonstrated by the experience of AndroZoo in the Android ecosystem. In this paper, we present MiniZoo, an evolving corpus of mini apps and associated metadata. The initial release comprises 16,273 mini-app packages, with multiple versions of the same mini apps. The dataset is managed and distributed through AndroZoo under its existing access conditions. To facilitate its reuse, we describe the collection process, data organization, provenance, and limitations, and explain how researchers can select and access the data. MiniZoo is designed to grow continuously through the addition of newly discovered mini apps, the inclusion of further host ecosystems, and the collection of successive versions of previously collected apps. This ongoing expansion will enable both comparisons across ecosystems and longitudinal studies of how mini apps evolve over time.
Read the full abstract

All authors Luca Ferrari, Marco Alecci, Jordan Samhi, Tegawendé F. Bissyandé, Jacques Klein, Luca Verderame

Mathematical Physics

Microscopic derivation of the classical KMS condition for one-dimensional Bose gases

The Kubo–Martin–Schwinger (KMS) condition provides an intrinsic characterization of thermal equilibrium, both in quantum and classical statistical mechanics. In this work, we establish stability of the KMS condition through the quantum-to-classical limit for one-dimensional defocusing Bose gases. More precisely, we consider scaled quantum KMS states in the (mean-field) semiclassical regime and prove that every associated Wigner measure satisfies the classical KMS condition corresponding to the nonlinear Schrödinger Hamiltonian. This result extends to an infinite-dimensional setting an earlier derivation of the classical KMS condition for finite-dimensional Bose–Hubbard systems. It also provides a complementary perspective on the derivation of nonlinear Gibbs measures from many-body quantum statistical mechanics.
Read the full abstract

All authors Zied Ammari, Shahnaz Farhat, Sören Petrat, Andrew Rout

Solar and Stellar Astrophysics

Signatures of mass transfer in open-cluster blue straggler stars. I. Collinder 261: Evidence from rotation, binarity, and chemical abundances

We present the first dedicated $v\sin i$ survey of blue straggler stars (BSS) in the old open cluster Collinder 261 ($t=6.7$ Gyr), based on VLT/FLAMES-GIRAFFE spectroscopy of 40 BSS and 42 turn-off/subgiant (TO/SGB) reference stars. Seventeen of the 39 BSS entering the statistics (44%) rotate faster than 40 km/s, whereas all TO/SGB reference stars are slow rotators; fast rotation is strongly linked to spectroscopic binarity. TESS light curves identify seven eclipsing or ellipsoidal binaries with $P_{orb}=0.5$-2.1 d, all rotating at the velocity set by tidal synchronisation, so in these systems the fast rotation is maintained by a close post-mass-transfer orbit and does not date the formation event. Differential C, O, Ba, and Fe abundances for the 22 slow rotators reveal that 62% carry the surface-pollution signature of mass transfer, in two distinct patterns separated by an empty ~1 dex gap in [Ba/Fe]: a Ba-rich group tracing AGB donors and a C/O-enhanced, Ba-normal group tracing donors below the $s$-process activation mass. The BSS iron abundance is consistent with the cluster value; the few apparent Fe excesses are partly explained by differential reddening and do not affect the abundance pattern. The correlated C-O enrichment is opposite to the depletion seen in globular-cluster BSS, indicating binaries wide enough for donors to reach the AGB. With $A^+_{r_h}=0.171\pm0.034$ and $f_{FR-BSS}=44\pm8$%, Collinder 261 extends the globular-cluster rotation-density trends into the open-cluster regime: its BSS were assembled by a continuously operating binary channel, and their chemistry records the dynamical environment of the host cluster.
Read the full abstract

All authors M.J. Rain, L. Monaco, S. Villanova, G. Carraro

High Energy Physics (Theory)

Rotational instability and quadrupolar order in BFSS matrix theory

We study the effect of angular momentum on thermal states of the BFSS matrix theory in its gravitational and high-temperature classical regimes. We focus on ensembles with only one non-zero angular momentum. On the gravity side, the rotating ensemble is described by the singly-spinning D$0$ black hole. Beyond a critical rotation, we find that this solution becomes unstable to a non-axisymmetric quadrupolar perturbation. Following this mode non-linearly, we find a new branch of non-axisymmetric rotating black holes that is thermodynamically favoured over the axisymmetric branch in both the microcanonical and canonical ensembles. On the other hand, in the classical matrix limit, we find no evidence for such collective quadrupolar order over the matrix sizes and angular momenta studied. Instead, rotation accelerates the separation of a single D$0$-brane from the clump of branes initially prepared. The separated brane carries a large fraction of the angular momentum by orbiting the rest of the clump at a large in-plane radius. Both the angular momentum share and the orbital radius relative to the clump size increase with the matrix size. These results distinguish a collective quadrupolar deformation in the gravity limit from angular-momentum transfer to a single brane in the classical matrix limit.
Read the full abstract

Space Physics

Electric Propulsion Database Development and Integrated Modelling Analysis

Electric propulsion (EP) systems require significant subsystem integration and data driven mod- elling for accurate sizing. This paper is an overview of a high-fidelity database development and analysis aiming to fill the gap in electric propulsion sizing. This is a preliminary study to develop a tool that considers all thruster technologies that exist within Electric Propulsion. Since thrust is set by the product of efficiency and available electrical power, this paper showcases that thruster selection directly constrains the spacecraft power bus. We showcase and hypothesize within the database analysis and in the future work that the correct way to choose electric propulsion systems is through budgeting the EP system through Mission constraints. Comparing technologies for propulsion parameters is important for spacecraft optimisation and no such dataset exists. The literature offers either deep surveys of one thruster type, which cannot be compared across technologies, or broad catalogues that quote manufacturer figures at a single operating condition without checking whether those figures agree with one another. This paper reports the development, performance levels and preliminary analysis of a unified Electric Propulsion technology database. Every device is reduced to a common set of parameters including power, thrust, specific impulse, exhaust velocity and total efficiency linked by governing analytical equations and state space models. The database holds 390 distinct models from 22 countries across twelve thruster types, from 0.4 W to 1 MW, with first flights from 1964 to 2025.
Read the full abstract

Computer Vision

Just Image Transformers with Late-Stage Pixel Refinement

Recent work on latent-free diffusion transformers has demonstrated the feasibility of training image generators without a latent VAE. This avoids compression artifacts from the VAE and simplifies the training pipeline. In particular, JiT showed that even a standard transformer can be trained to generate images when the model is trained to predict the final image ($x$-prediction). Even though JiT unlocks training efficiency comparable to latent models, it still suffers from reduced image quality likely arising from modeling the image as a collection of large patches. In this work, we revisit JiT and show that adding simple convolutional pixel-pathway layers in a post-training phase significantly improves generation quality. Moreover, we show that the added pixel-pathway does not need to be active in the early denoising stages, and can be activated as late as in the last $20\%$ of the denoising process. Our experiments on ImageNet@256 show improvements compared to the original JiT results, achieving an FID of $1.78$ (from $2.34$ with corrected sampling), without using additional losses.
Read the full abstract

Symplectic Geometry

Linearized contact homology and closed Reeb orbits on prequantization bundles

We study closed Reeb orbits on prequantization circle bundles $(M,\xi)\to(B,\omega)$ with $c_1(\xi)|_{\pi_2(M)}=0$. When $\omega|_{\pi_2(B)}\ne0$, we construct augmentations and compute contractible linearized contact homology as a direct sum of shifted copies of $H_*(B;\mathbb Q)$, under a condition on the rational Betti numbers of $B$. The construction uses Morse-Bott methods and requires neither a filling nor index-admissibility. When $c_1(TB)|_{\pi_2(B)}\ne0$, index recurrence and Morse inequalities yield multiplicity and non-hyperbolicity bounds for nondegenerate contact forms under additional index assumptions. If $c_1(TB)|_{\pi_2(B)}=0$ and either $H_{\mathrm{odd}}(B;\mathbb Q)=0$ or $\omega|_{\pi_2(B)}=0$, every nondegenerate supporting contact form has infinitely many simple closed Reeb orbits. We also establish multiplicity and non-hyperbolicity results for smooth prequantizations of symplectic orbifolds admitting Hamiltonian circle actions with isolated fixed points.
Read the full abstract

Instrumentation and Methods for Astrophysics

Extracting Published Discussions of 13,616 Galaxies with Language Models

Astronomers often have difficulty determining whether an object, such as a star or galaxy of interest, has been discussed in the literature. Existing tools enable object identification and linking in papers but do not record whether an object was genuinely discussed in a paper or just appears in a table or catalog. We address this by using language model agents to identify galaxies discussed in more than two sentences across a corpus of 156,006 arXiv astro-ph papers and extract those discussions. We extract 44,332 discussions of individual galaxies containing 85,314 quotes from 18,641 papers. With sky coordinates for 98.2% of discussions, we group them into 13,616 galaxies. We bring these results together into GalaxyWiki, where users can find literature discussing galaxies, read what was said about them, and follow links to the source papers. The code, dataset, and webpage are available at https://github.com/NolanKoblischke/galaxywiki-paper
Read the full abstract

Lattice Field Theory

Reconstruction of pion polarizabilities in lattice QCD with machine learning

We apply the denoising diffusion probabilistic model (DDPM) from generative machine learning to a model-independent reconstruction of electric and magnetic polarizabilities of both charged and neutral pions. The inputs are the sparse momentum-space four-point-function data in lattice QCD for the connected quark-line contributions obtained in previous simulations at four pion masses between approximately 370 and 1100 MeV on quenched Wilson lattices. For charged pions, the elastic contribution is obtained from the pion form-factor slope, and the complete electric and magnetic inelastic terms are reconstructed separately. The neutral-pion electric and magnetic channels are reconstructed without an elastic charge-radius contribution. Four distinct families of synthetic training functions are employed, so that no single analytic fitting function is prescribed for the lattice data. We present low-momentum reconstructions to the limit of zero momentum transfer where the polarizabilities are defined, and mass extrapolations to the physical pion mass of 138 MeV.
Read the full abstract

Astrophysics of Galaxies

Revealing Hidden Excitation in the Core of NGC 4696: New Infrared Diagnostic Diagrams with JWST

The warm molecular and ionized gas in filamentary nebulae of cool-core BCGs provide a sensitive probe of the poorly understood exchange of energy between the ICM, cold molecular gas, and AGN feedback. Optical line ratios commonly yield similar LINER-like classifications across these systems, but do not resolve how the excitation of warm molecular gas varies between compact nuclear structures and surrounding filaments. We present JWST/NIRSpec integral-field spectroscopy of the central $3''\times3''$ ($620\times620~{\rm pc^2}$) of the Centaurus cluster BCG, NGC~4696. The NIR data resolve H$I$ and ro-vibrational H$_2$ emission across compact nuclear apertures, a circumnuclear swirl, and filamentary structures. H$_2$ excitation diagrams yield characteristic low-excitation temperatures of $T_{\rm ex}\simeq1.8-2.2\times10^3$~K, while curvature at higher energies requires hotter or non-thermal components. The H$_2$/Pa$\alpha$ balance varies strongly: filamentary regions have H$_2$ 1-0 S(1)/Pa$\alpha$ ratios 4-8 times higher than the compact nuclear apertures. In contrast, matched-aperture optical diagnostic diagrams show substantially less regional separation, with all regions occupying a low-ionization locus. We thus introduce new spatially resolved NIR diagnostic diagrams that reveal distinct molecular excitation conditions across the nucleus, swirl, and filaments. Ratios are broadly consistent with contributions from AGN-related photoionization, ICM-related cooling radiation, and non-radiative heating, although the present data do not uniquely determine their relative importance. These results demonstrate how spatially resolved NIR spectroscopy reveals excitation variations hidden in conventional optical classifications, providing a new route to constrain the multiphase cooling-feedback cycle in cluster cores.
Read the full abstract

All authors Olivia R Pereira, Julie Hlavacek-Larrondo, Michael Reefe, Gary J Ferland, Hyunseop Choi, Mathieu Marquis, Jorge Barrera-Ballesteros, Benjamin Vigneron, Ming Sun, Rebecca E A Canning, Gregory Taylor, Loic Albert, Francesco D'Eugenio, Megan Donahue, Andrew C Fabian, John S Gallagher, Marie-Lou Gendron-Marsolais, Pierre Guillard, Minghao Guo, Nina Hatch, Ralf Kotulla, Yuan Li, Roberto Maiolino, Allison Man, Michael A McDonald, Brian R McNamara, Valeria Olivares, Marine Prunier, Christopher S Reynolds, Carter Rhea, Annabelle Richard-Laferriere, Helen R Russell, Philippe Salome, Prathamesh Tamhane, Auriane Thilloy, Grant R Tremblay, G Mark Voit, Stephen A Walker

Solar and Stellar Astrophysics

High-precision abundance analysis of a unique CEMP-no star in Sculptor

The Carbon-Enhanced Metal-Poor stars (CEMP-no, low barium) are commonly accepted as the direct descendants of Population III (Pop III) stars which either ended their lives as faint supernovae (SNe; E_SN < 10^51 erg) or were fast-rotating stars (spinstars). While CEMP-no stars are common in the Milky Way (MW) and ultra-faint dwarf galaxies (UFDs), they are surprisingly rare in dwarf spheroidal galaxies (dSphs). This could hide an environmental dependence of the properties of Pop III stars, highlighting the importance of high-resolution studies of CEMP-no stars in dSphs. Here we present the first detailed study of a CEMP-no star with [Fe/H] < -3 in the Sculptor dSph. The high-resolution VLT/UVES spectra allowed us to derive one of the most precise abundance patterns obtained for a star belonging to a dwarf galaxy, providing chemical abundances for 17 elements and 4 upper limits. This star, DR20080, is extremely metal-poor, [Fe/H] = -3.26, and clearly enhanced in carbon, [C/Fe] = +1.60 (after correcting for internal mixing). Furthermore, it has low [Ba/Fe] = -1.28. The oxygen abundance of DR20080 ([O/Fe] = +1.53) is significantly higher than the average found in CEMP-no stars in the MW, while the iron-peak elements are under-abundant ([Mn/Fe] = -1.02; [Ni/Fe] = -0.45). The abundance pattern of DR20080 was compared to theoretical yield predictions of Pop III and Pop II SNe to infer the properties of its progenitor. The best result points toward a double Pop III progenitor scenario, with the major contributor to the abundance pattern being a Pop III faint SN with mass m_PopIII = 43 M_0 and explosion energy E_SN = 0.6 x 10^51 erg, while the minor contributor is a Pop III hypernova of m_PopIII = 19.4 M_0 and E_SN = 10 x 10^51 erg. The unique properties of DR20080 have never before been observed in Sculptor or in any other environment within the MW and its satellite system.
Read the full abstract

All authors Francisco Cuadra, Ása Skúladóttir, Romain Lucchesi, Ioanna Koutsouridou, Irene Vanni, Stefania Salvadori, Eline Tolstoy, Raghubar Singh

Robotics

ASPIRE: An Agentic System-Identification Platform for Iterative Reasoning and Exploration

System identification requires expert decisions on the identification method, incorporation of prior physical knowledge, and model validation. We propose ASPIRE, a framework for nonlinear system identification in which a large language model agent selects and executes the identification method. ASPIRE constructs closed-form, nonlinear static or dynamic models from measured data, optionally supplemented by a system description, the intended use of the model, and physical requirements expressed formally or in natural language. Each model includes its parameters, an estimation record, an interpretation of its terms, and an error evaluated on test data inaccessible to the agent. Formal physical requirements are verified independently. ASPIRE operates autonomously or interactively, allowing the agent to ask questions and the user to request further analyses and identification rounds. We evaluate ASPIRE on three measured systems: a mobile robot, a rotational system with dry friction, and a laboratory fermenter. Each system is tested with and without a description, in five runs per task. All 30 runs delivered valid models, at a cost of 0.96 to 3.38 US dollars per run. An assessment by the authors found that, for the 30 models recommended by the agent, one per run, the interpretation of the model and the documentation of its structure selection and parameter estimation fully satisfied 132 of 150 assessed items and partly satisfied 18. The most frequent discrepancy, found in nine models, concerned inaccurate or unsupported claims about model structure selection.
Read the full abstract

Artificial Intelligence

Building World Models from Intrinsic Motivation

In the absence of extrinsic rewards or task supervision, human behavior is guided by intrinsic motivation—neural signals that drive curiosity, creativity, play, and other essential facets of intelligence. Intrinsic motivation plays an important developmental role in building robust world models that enable generalization and adaptation. Yet existing studies leave open how a broad range of intrinsic motivations interacts with world-model representation learning to support downstream control. This limits our understanding of which intrinsically collected experiences build reusable models. To address these gaps, we integrate 24 intrinsic objectives, including established methods and explicit variants spanning curiosity, entropy, and competency, into a common model-based reinforcement learning framework. We evaluate the world models they build through zero-shot, few-shot, and multi-task control in locomotion, manipulation, and open-ended gaming environments, using three representation-learning strategies: observation reconstruction, latent prediction, and redundancy reduction. Curiosity-driven exploration with observation reconstruction yields the strongest aggregate family profile, although the best individual pairing depends on the experiment. In compositional vision-based tasks, intrinsic motivation derived from both episodic and lifelong novelty—providing a complementary memory to the world model's generic recurrent state—achieves the best zero-shot transfer and few-shot adaptation. These findings motivate designing intrinsic motivation and representation learning jointly. Our results demonstrate world models as a powerful tool for measuring the quality of intrinsically motivated experience, and positions intrinsic motivation as a key ingredient for building world models autonomously.
Read the full abstract

All authors Reece Keller, Declan Kutscher, Xaq Pitkow, Aran Nayebi

High Energy Physics (Theory)

Resummations In Higher-Curvature Modifications Of Starobinsky Inflation

We study higher-curvature corrections to Starobinsky inflation, which are expected to be generated at the quantum level. When treated perturbatively in the couplings $\gamma$ of the higher-dimensional operators, these terms produce inverse powers of the slow-roll parameter $\alpha$, casting doubt on the reliability of the slow-roll expansion. If instead $\gamma$ is kept exact, the expansion fails to reproduce the Starobinsky model smoothly as $\gamma\to 0$. We show that this problem is solved by resumming certain series in the ratio $\alpha/\gamma$. The resulting “improved” slow-roll expansion is regular in the limit $\gamma\to 0$ and gives the same result whether one expands first in $\gamma$ or in $\alpha$. We illustrate this explicitly for an $R^3$ term, obtain the power spectra, spectral indices and tensor-to-scalar ratio at next-to-leading order in the improved expansion, and show that results in the recent literature are recovered as the first-order Taylor expansion of ours. Comparing with the latest ACT, Planck and SPT data, we find that the resummed and first-order results differ appreciably only for values of $n_s$ outside the favored regions. Finally, we comment on the origin of higher-curvature operators in UV-complete theories, such as renormalizable quantum gravity, where the predictions of the Starobinsky model are stable.
Read the full abstract

Quantum Physics

FusionQ: A Generic and Optimal Decoding Framework for Ultra-Low-Latency Quantum Error Correction

Quantum error correction (QEC) decoding has traditionally been organized around fixed decoding instances: a monolithic space-time detector error model (DEM), or a predetermined temporal window, is handed to a backend decoder after enough syndrome data has arrived. This paper presents FusionQ, a fusion decoding framework that changes this paradigm by deriving local decoding tasks from the observed syndrome and then fusing their results into a single global correction. Instead of forcing every shot through the same global or windowed problem, FusionQ builds open-boundary local DEMs around fired detector clusters and expands the corresponding footprints only when a local optimum uses virtual boundary errors. FusionQ is generic in the decoder backend: the scheduler requires only a local solution and a boundary-use flag, so exact solvers, matching-style decoders on graphlike DEMs, and heuristic decoders can share the same runtime. We prove a compositional optimality theorem: with an exact minimum-weight local backend and boundary-free publication, fusing the published local decoding results yields a global most-likely error (MLE) solution of the original DEM. Our evaluation shows that, with an exact backend, FusionQ matches the monolithic MLE result on every shot. Across decoder backends, FusionQ substantially reduces decoding latency by overlapping local decoding with hardware syndrome acquisition.
Read the full abstract

All authors Junyu Zhou, Sean Garner, Aarthi Sundaram, Gushu Li, Yue Wu

Quantum Physics

Finite Measured Entropy Production Does Not Imply Finite Microscopic Irreversibility

Coarse-grained measurements provide a lower bound on entropy production, but how much microscopic irreversibility can they conceal? We construct a driven quantum system in which an imperfect counter registers a finite, strictly positive arrow of time while the microscopic entropy production is infinite. The system is a degenerate parametric oscillator below threshold, coupled to a thermal reservoir through repeated interactions with bosonic ancillas. The oscillator sustains a finite, positive heat current into the vacuum bath, where absorption is forbidden and forward and reverse trajectory laws have different supports. A photon counter with finite collection efficiency, gated time bins, and independent dark counts defines a common coarse-graining of the aligned trajectory laws. For fixed observation duration $\mathcal T$ and detector parameters, we prove that the observed relative entropy remains uniformly bounded as the bath temperature $T_b$ vanishes and is strictly positive at vacuum. The hidden contribution obeys $\Sigma_{\mathrm{hid}}(\mathcal T,T_b)=\mathcal T\dot Q_0/(k_BT_b)+O(1)$, where $\dot Q_0>0$ is the stationary vacuum heat current. False positives conceal the support mismatch while preserving measurable time asymmetry. Thus, finite and informative measurements of entropy production can coexist with infinite microscopic irreversibility.
Read the full abstract

High Energy Astrophysical Phenomena

Accretion variability as a probe of black holes and their environments: Light curve signatures of a single hot spot

The compact radio source at the centre of the Milky Way, Sagittarius A*, exhibits temporal broadband brightness variability. The observed timescales of such fluctuations, ranging from minutes to hours on horizon scales, arise from the complex interplay of gravitation, plasma, and light propagation processes in the immediate vicinity of the supermassive black hole. Kinetic and fluid simulations associate this variability with turbulence and the formation of magnetised, localised plasma structures. Focusing on a single fluctuating compact emission region, a 'hot spot', we study how intrinsic source fluctuations map onto different regions of the observer's screen. Our aim is to identify distinct imprints of light propagation effects, which encode information both about the source and the underlying space-time geometry. We simulate image time series by propagating the emitted light to a distant observer's sky plane in Minkowski, Schwarzschild, and Kerr space-times. We then compute the corresponding observed light curves and power spectra, both for the full observed images and for selected angular sectors of the observer's sky plane. We derive a semi-analytical model for the power spectrum of a fluctuating hot spot in flat space-time, reproducing all qualitative features of curved-space-time spectra at moderate inclinations. Moreover, we describe how such power spectra are affected by space-time geometry, inclination, and source velocity, and show that orthogonal two-sector angular decompositions of the sky plane provide information on system orientation and lensing effects. The physical effects highlighted in this paper are expected to be present in the timing signature of accretion-flow variability, as observed by current instruments and future photometric and interferometric facilities. Further work is needed to investigate if our conclusions hold for realistic disc configurations.
Read the full abstract

All authors I. Urso, C. Deffayet, A. Held, F. H. Vincent

Algebraic Geometry

Waldschmidt constants and invariant geometry for exceptional rank-four complex reflection groups

We compute the Waldschmidt constants of the point configurations dual to the reflection arrangements of the exceptional complex reflection groups $G_{29}$, $G_{31}$ and $G_{32}$. They equal $3$ for the forty points of $G_{29}$, $36/11$ for the sixty-point Klein configuration, and $10/3$ for the forty-point Penrose configuration. Together with the known cases $W(F_4)$ and $W(H_4)$, this completes the computation for all exceptional irreducible complex reflection groups of rank four. An invariant dodecic and its sixfold orbit product give extremal surfaces for the first two configurations, and we identify the six dodecics as pullbacks of the coordinate hyperplane sections of the Segre cubic under the Igusa-Segre duality. The matching lower bounds follow from restriction to quadrics and distinguished planes, including a weighted Bézout estimate arising from two related Hesse arrangements. As a consequence, we determine the asymptotic resurgences of the three ideals.
Read the full abstract

All authors Krishna Hanumanthu, Jack Huizenga, Jake Kettinger, Lukasz Merta, Tomasz Szemberg, Justyna Szpond

2611.13108 · 9 Nov · 14 pages, This work grew out of a working group at the Mini-Workshop on Asymptotic Invariants of Linear Systems and Homogeneous Ideals with Symmetries, held at the Mathematisches Forschungsinstitut Oberwolfach in September 2026

Computation and Language

Catch It Where It Happens: Real Time Error Correction During LLM Reasoning

Large language models (LLMs) are increasingly trusted with mathematical, scientific and engineering work, yet their mistakes are often confidently written and hard to spot. A single wrong step tends to propagate, making the model waste inference tokens following a false premise. Existing safeguards either check only the final answer and restart from scratch, or rely on a verifier LLM that competes for GPU capacity and slows down response time. We propose online, deterministic verification: many kinds of reasoning steps can be validated by code, so a step can be verified in real time, as it is written, without any model call or GPU use. Whenever a check fails, we point out the mistake and let the model continue from the same spot. We demonstrate this approach on symbolic mathematics, where derivations already contain explicit expressions that a computer algebra system can check in milliseconds. Tested on the ASyMoB benchmark with three open-weights models, our online validation improves accuracy by 1.4-5.6 percentage points, corresponding to a 1.8-14.9% net reduction in incorrect answers, while increasing token usage by only 6-7%. Because the verifier is independent of the model, the approach complements progress in LLMs rather than competing with it: every new model can benefit immediately, and the benefit grows as deterministic checkers extend to more kinds of reasoning steps.
Read the full abstract

Probability

Sharp and explicit Berry–Esseen inequalities over convex sets

We establish finite-sample Berry–Esseen bounds over all convex subsets of $\R^{d}$ for sums of independent random vectors. The upper bounds give explicit dependence on the sample size, dimension, and moments of the summands, with numerical constants. They allow arbitrary dependence among coordinates and cover every real moment order from three onward. For i.i.d. summands, we determine the optimal dependence of the worst-case error on the sample size, the dimension, and the prescribed moment bound. We obtain matching lower bounds using bounded distributions that can be chosen independently of the sample size. The results reveal a fourth-moment threshold for normal approximation over convex sets. Four is the first moment order for which uniform bounds on the corresponding coordinate moments yield a worst-case error bound linear in the dimension. Bounds on higher moments give the same worst-case rate. The proof combines convexity with the addition of independent Gaussian noise to establish a normal approximation bound for Rademacher sums on Parseval frames. Random swaps with independent copies then allow us to apply this bound conditionally to general sums of independent random vectors.
Read the full abstract

High Energy Physics (Phenomenology)

Galaxy clusters as cosmic lenses for high frequency gravitational-waves

High-frequency Gravitational Waves (HFGW)($f \gtrsim 1$ MHz) offer a distinctive probe of exotic physics. In the absence of dedicated detectors targeting this frequency range, indirect probes based on graviton-photon conversions in astrophysical magnetic fields become particularly relevant. In this work we show that the extended magnetic fields permeating Galaxy Clusters may act as cosmic lenses, leading to a strong enhancement of graviton-photon conversions. In this context, X- and $\gamma$-ray observations of the Virgo Cluster through large field-of-view instruments such as eROSITA and COMPTEL yield novel limits on the amplitude of stochastic HFGW backgrounds in the frequency range $10^{17}-10^{22}$ Hz, improving upon constraints derived from the Galactic magnetic field by one to two orders of magnitude.
Read the full abstract

All authors Francesca Lecce, Francisco R. Candòn, Alessandro Lella, Alessandro Mirizzi, Julia K. Vogel, Tiziano Zanzarella

Solar and Stellar Astrophysics

The chemical DNA of the Magellanic Clouds – VII. The lithium abundance in the Small Magellanic Cloud and the discovery of 4 lithium-rich stars

Measuring lithium abundance in galaxies beyond the Milky Way (MW) remains a challenging task. In these systems, only stars close to the tip of the red giant branch (RGB) are sufficiently bright to be observed spectroscopically.Their A(Li) are expected to be altered by mixing episodes that occur during the stellar evolution, making the interpretation of measured abundances more complex. In this context, we present A(Li) of 209 RGB stars in the Small Magellanic Cloud (SMC),observed with FLAMES@VLT. Approximately one-third of the sample shows a Li line detected at a significance level above 5sigma, while about half of the sample has detections above 3sigma; for the remaining stars, only upper limits can be derived. This sample exhibits a significant star-to-star scatter in A(Li). Using theoretical models, we find that the observed scatter can be likely explained by the combination of 2 effects.First,our stars span a wide range of ages, masses, [Fe/H]. Thermohaline mixing, one of the leading mechanisms proposed to explain the extra mixing occurring at the RGB bump, predicts different levels of A(Li) depletion depending on stellar mass and [Fe/H], naturally contributing to the observed scatter in A(Li). Second, in the SMC, the enrichment in A(Li) driven by novae is expected to occur at lower [Fe/H] than in the MW, because the SMC has experienced a lower star formation efficiency than the MW. This is also supported by the rise of A(Li) with decreasing age, as measured in 3 SMC globular clusters. Furthermore, we report the first identification of 4 low-mass Li-rich giant stars in the SMC, reinforcing the existence of Li-rich stars in extra-galactic systems. All these results demonstrate that A(Li) measured in stars close to the RGB tip, still provide valuable constraints on both stellar/galactic chemical evolution of A(Li) in external galaxies.
Read the full abstract

All authors A. Mucciarelli, M. Palla, C. T. Nguyen, L. Monaco, E. Pancino, D. A. Alvarez Garay, S. Anoardo, C. Lardo, F. Matteucci, D. Romano

Quantum Physics

Which data samples matter when training a quantum model?

The expressivity, loss landscapes, and generalization behavior of quantum machine learning models have been studied extensively, yet mainly from a model-centric point of view. Little is known about the compatibility of particular quantum models with specific datasets. In this work, we study quantum machine learning models with a data-dependent focus by establishing a quantum-compatible form of the V-usable information. This metric quantifies the information a model can extract from a dataset, enabling a direct, data-centric comparison of different models. We employ this metric to study which data points contribute most significantly to a model's learning process. This enables us to develop an informed preselection strategy for data subsets that can significantly improve the data efficiency of the training pipeline. This work establishes the V-usable information as a powerful diagnostic tool for learning dynamics and a practical route to data-efficient quantum machine learning.
Read the full abstract

All authors M. Emre Sahin, Cenk Tüysüz, David Galvin, Oscar Wallis, Alex Gregory, Paul Edwards, Christa Zoufal