Mesoscale and Nanoscale Physics
Superconducting edge contact to a buried Ge quantum well
Dina Sokolova, Davide Costa, Davide Degli Esposti et al. (+4 more)
Semiconductor-
superconductor hybrid systems form the basis for a broad range of
quantum phenomena, such as Andreev bound states, topological superconductivity, and hybrid
quantum Hall states. The performance of these hybrid systems can strongly benefit from the exceptional electronic properties of
heterostructures with deeply buried
quantum wells. However, establishing high-quality superconducting contacts to such
quantum wells remains challenging. Here, we address this problem with an edge-contact approach that combines anisotropic dry etching to access the
quantum well followed by wet etching for interface preparation. Josephson junctions fabricated using this approach consistently exhibit near-unity interface transparency, while
quantum point contacts show strong Andreev enhancement, with the first conductance plateau approaching $4e^2/h$, the limit of perfect Andreev reflection. In superconducting
quantum point contacts, the high contact quality is further manifested by quantized supercurrent. These results establish the combination of dry and wet etching as a robust strategy for integrating deeply buried
quantum wells into superconducting hybrid devices.
Read the full abstract ▾
All authors Dina Sokolova, Davide Costa, Davide Degli Esposti, Tommaso Costanzo, Giordano Scappucci, Georgios Katsaros, Anton V. Bubis
Artificial Intelligence
AdaGuard: Enhancing Safety and Policy Compliance with Reasoning-Enabled LLM-As-A-Judge Guardrails
Melissa Kazemi Rad, Sihui Dai, Isha Slavin et al. (+6 more)
Enterprise generative AI applications require robust safety mechanisms that can accommodate diverse risk postures, evolving policies, and varying latency constraints. Current guardrail solutions often suffer from
rigidity, relying on fixed policy sets and offering limited transparency or
reasoning flexibility. We present Adaguard, an adaptive LLM-as-a-Judge framework designed to address these challenges through dynamic policy enforcement and adaptive
reasoning-budget allocation. Built using supervised fine-tuning (SFT) and reinforcement learning (GRPO), AdaGuard generalizes to user-defined safety and compliance policies at runtime without requiring frequent model updates. A core innovation of our approach is the ability to dynamically infer the complexity of input-policy pairs, allowing the model to switch between high-speed black-box inference and explainable,
reasoning-enabled moderation. This flexibility enables developers to balance stringent latency requirements with the need for actionable transparency. This adaptive capability allows AdaGuard to rival other guardrail and frontier models several times its size, while its auto-
reasoning mode recovers the accuracy of always-on
reasoning at a fraction of the latency
Read the full abstract ▾
All authors Melissa Kazemi Rad, Sihui Dai, Isha Slavin, Kushal Chawla, Mann Patel, Jian Ni, William M. Campbell, Stephen Rawls, Sambit Sahu
Earth and Planetary Astrophysics
Searching for hot water world candidates with CHEOPS: II. A low-density super-Earth orbiting the K dwarf TOI-2211
J. A. Egger, P. Cortés-Zuleta, D. Kubyshkina et al. (+98 more)
Small
exoplanets on close-in orbits are exposed to high levels of
stellar irradiation, which can effectively remove low-mass
primordial H/He envelopes through atmospheric escape. Under these conditions,
planets whose masses and radii are incompatible with both bare rocky interiors and extended H-dominated atmospheres are therefore expected to host high mean molecular weight atmospheres, making them promising candidates for volatile-rich hot water worlds. We present the first detailed characterisation of the low-density super-Earth TOI-2211 b (P~3.09 d), a recently validated transiting
planet orbiting a K type
star (G mag: 9.1). Combining TESS and CHEOPS photometry with radial velocity measurements from HARPS, ESPRESSO, and KPF, we precisely constrain the
planet's radius and mass to $1.457 \pm 0.047$ R$_\oplus$ and $2.69 \pm 0.56$ M$_\oplus$. We also identify an outer, non-transiting planetary candidate (P~15.6 d, Msin(i) $ = 8.53 \pm 1.04$ M$_\oplus$) in the radial velocity data. We investigate the internal structure of the transiting
planet and assess the stability of the inferred volatile envelopes through a detailed evaporation analysis, thereby identifying TOI-2211 b as a promising hot water world candidate. TOI-2211 b therefore joins the small but growing sample of these close-in low-density super-Earths, which represent particularly interesting targets for testing current
planet formation and
evolution theories. Additionally, we investigate the full architecture of the planetary system using a generative model of
planet formation and
evolution, with the resulting posterior distributions showing a clear over-density consistent with the observed orbital period and minimal mass of the non-transiting planetary candidate identified in the radial velocity data.
Read the full abstract ▾
All authors J. A. Egger, P. Cortés-Zuleta, D. Kubyshkina, H. P. Osborn, S. Marques, I. Crossfield, A. Deline, S. Ulmer-Moll, A. E. Simon, G. Scandariato, A. C. M. Correia, Y. N. E. Eschen, D. Gandolfi, V. Adibekyan, A. Bonfanti, C. Pezzotti, P. H. Premnath, S. G. Sousa, T. G. Wilson, L. Fossati, R. Eltschinger, Y. Alibert, D. Degen, G. Lacedelli, T. Zingales, B. Akinsanmi, R. Alonso, T. Bárczy, D. Barrado, S. C. C. Barros, W. Baumjohann, W. Benz, N. Billot, L. Borsato, A. Brandeker, C. Broeg, M. Buder, A. Castro-González, A. Collier Cameron, C. Corral Van Damme, Sz. Csizmadia, P. E. Cubillos, M. B. Davies, M. Deleuil, O. D. S. Demangeon, B.-O. Demory, A. Derekas, B. Edwards, D. Ehrenreich, A. Erikson, A. Fortier, M. Fridlund, K. Gazeas, M. Gillon, M. Güdel, M. N. Günther, J. Hasiba, A. Heitzmann, Ch. Helling, K. G. Isaak, L. L. Kiss, D. Kitzmann, J. Korth, K. W. F. Lam, J. Laskar, A. Lecavelier des Etangs, A. Leleu, M. Lendl, M. Lund, D. Magrin, P. F. L. Maxted, B. Merín, C. Mordasini, M. Munari, V. Nascimbeni, G. Olofsson, R. Ottensamer, I. Pagano, E. Pallé, A. Peel, G. Peter, D. Piazza, G. Piotto, D. Pollacco, D. Queloz, R. Ragazzoni, N. Rando, H. Rauer, I. Ribas, N. C. Santos, D. Ségransan, A. M. S. Smith, M. Stalport, S. Sulis, Gy. M. Szabó, S. Udry, V. Van Grootel, J. Venturini, E. Villaver, N. A. Walton, S. Wolf
Strongly Correlated Electrons
Exact solutions for kinematics across interaction-deformed Fermi surfaces
Alexander Kruchkov and F. Duncan M. Haldane
Luttinger's
theorem fixes the volume of the
Fermi sea of a
Fermi liquid, but not how interactions redistribute occupation around its surface. For an interaction switched on and tuned at fixed density, we derive the exact rates of change of the occupation deficit inside the interacting
Fermi surface and of the surplus outside it. Each is the weight swept by the deforming surface plus a change of Luttinger's integrand at fixed momentum; the difference of the surface terms is flux of incoherent weight $1 - Z_{\vec k}$. These identities hold with or without the Luttinger
theorem.
Read the full abstract ▾
High Energy Astrophysical Phenomena
PEARLS: NuSTAR and XMM-Newton Extragalactic Survey of the JWST North Ecliptic Pole Time-Domain Field V: Unraveling X-ray Spectral Variability
Samantha Creech, Francesca Civano, Daniel Wik et al. (+13 more)
As one of the first deep time-domain surveys in the hard X-ray band, the NuSTAR campaign in the JWST North Ecliptic Pole Time Domain Field (NEP-TDF) offers a
prime opportunity to measure variability in a faint population of Active
Galactic Nuclei (AGN) in the medium-to-high
redshift universe. In the X-ray band, AGN variability is generally caused by changes in the brightness of the corona ($F_{int}$ variability) or changes in the column density of obscuring material near the SMBH ($N_H$ variability). Broad-band spectroscopy ($\sim$0.5-24 keV) is required to reliably disentangle these two phenomena. With simultaneous NuSTAR and XMM-Newton coverage—as well as 1.8 Ms of Chandra monitoring—the X-ray surveys in the NEP-TDF are ideally situated for this purpose. This work presents the first spectral modeling of the 52 NuSTAR-identified sources discovered in the latter half of the campaign, as well as a spectroscopic variability study of the entire 112-source NuSTAR NEP-TDF catalog. Seven variable candidates are identified. We find that $\gtrsim 1000$ total counts split across many ($\gtrsim 10$) epochs are needed in order to detect variability. Additionally, variable sources are compared to the survey sensitivity in order to inform population synthesis models of how variations can affect the detectability of faint AGN, which dominate the SMBH
accretion history. Lastly, the timescale of $N_H$ variability is investigated as a probe for the obscurer location. Significant variability is slightly more common on $> 100$ day timescales, suggesting that the
torus may have only slight dominance in driving $N_H$ variability in faint AGN.
Read the full abstract ▾
All authors Samantha Creech, Francesca Civano, Daniel Wik, Núria Torres-Alba, Ross Silver, Xiurui Zhao, Tonima Ananna, Gibson Bowling, Norman A. Grogin, Anton M. Koekemoer, Peter Maksym, Sylvia Mesicek, Rafael Ortiz III, Payaswini Saikia, Steven Willner, Rogier Windhorst
Cosmology and Nongalactic Astrophysics
The BUFFALO Survey : The Subhalo Mass Function for Abell 2744 with Strong+Weak Gravitational Lensing
Nency R. Patel, Mathilde Jauzac, Anna Niemiec et al. (+16 more)
We present a combined strong and weak gravitational
lensing mass model of the massive merging
galaxy cluster Abell 2744 (z = 0.308). The mass reconstruction combines strong-
lensing constraints in the cluster core with weak-
lensing measurements derived from deep imaging with the Hubble Space
Telescope (HST) and the James Webb Space
Telescope (JWST), enabling a reconstruction of the projected mass distribution from the core to the cluster outskirts. Using the resulting mass map, we identify cluster substructures and construct the subhalo mass function (SHMF) for Abell 2744. To interpret the observed substructure population, we apply identical detection methods to projected mass maps of massive
galaxy clusters from the
BAryons and
HAloes of MAssive Systems (BAHAMAS) simulations. We observe a systematic excess of detected substructures in Abell 2744 compared to BAHAMAS. We interpret this excess as likely reflecting Abell 2744's exceptionally disturbed dynamical state, characterised by complex ongoing mergers, rather than tension with $\Lambda$Cold Dark Matter ($\Lambda$CDM) or self-interacting dark matter (SIDM) predictions. Although confirming this interpretation will require extending the analysis to a larger sample of clusters spanning a range of dynamical states.
Read the full abstract ▾
All authors Nency R. Patel, Mathilde Jauzac, Anna Niemiec, Guillaume Mahler, David Lagattuta, David Harvey, Johannes Schwinn, Andrew Robertson, Lukas J. Furtak, Sownak Bose, Alastair Edge, Anton Koekemoer, Jose Diego, Andreas Faisst, Eric Jullo, Marceau Limousin, Richard Massey, Derek Perera, Sut-Ieng Tam
High Energy Astrophysical Phenomena
A Newly Identified Transient Ultraluminous X-ray Source in M31
Ferdinand, Murray Brightman, Jean-Marie Hameury et al. (+9 more)
Ultraluminous X-ray sources (ULXs) are off-
nuclear X-ray sources with luminosities exceeding the Eddington limit of typical
stellar-mass compact objects. While most known ULXs are relatively persistent sources, transient ULXs have emerged as a distinct population that offer unique insight into extreme, super-Eddington
accretion. We report the discovery and X-ray characterization of a newly identified transient ULX in M31. First detected serendipitously on 2021 June 9, the source reached a peak unabsorbed $0.3 - 10$ keV luminosity of $\sim10^{39}\ \rm erg\ s^{-1}$. Its spectral
evolution suggests increasing advection at later epochs, opposite to the expected transition from a super-Eddington slim disk toward a standard thin disk during the decay of the outburst. The observed X-ray light curve can be reproduced using the disk-instability model of Hameury & Lasota, with characteristic timescales of $\sim41-61$ days. The inferred mass-
accretion rate is super-Eddington for a $1.4\,M_{\odot}$
neutron star but approximately Eddington for a $10\,M_{\odot}$
stellar-mass black hole. Nearby optical sources from the Hubble Source Catalog suggest a possible
giant donor with an age of $\log_{10}({\rm age/yr})\sim8.5$ and a mass of $\sim3.2\,M_{\odot}$, when compared with
stellar isochrones with an assumed $A_V=1.58$. We estimate a transient ULX occurrence rate in M31 of $0.18\pm0.10\ {\rm yr^{-1}}$. This rate is lower than those reported for other
galaxies in previous studies. M31 ULX-3 expands the small population of known transient ULXs in M31 and highlights the importance of continued, regular X-ray monitoring of nearby
galaxies.
Read the full abstract ▾
All authors Ferdinand, Murray Brightman, Jean-Marie Hameury, Elias Kammoun, Dominic J. Walton, Matthew J. Middleton, Hannah Earnshaw, Felix Fürst, Wynn V. Jacobson-Galán, Jacob E. Jencson, Margaret Lazzarini, Mark T. Reynolds
Computation and Language
CARE: Certifying Acceleration for Vision-Language-Action Inference
Rui Liu, Tong Zheng, Jindong Gu et al. (+1 more)
While
vision-
language-action (VLA) models have advanced rapidly, running them at every control step remains expensive. Prior work accelerates VLA inference using techniques like action chunking and visual-
token pruning, typically evaluating based on latency and average task success. However, acceleration may discard information and break tasks the original policy would solve, a risk hidden by average metrics. Measuring these failures is challenging because action deviations compound over closed-loop trajectories, meaning task failure is only observable across full episodes. We therefore define an acceleration-induced failure via paired rollouts from identical initial conditions, tracking when the reference succeeds but the accelerated policy fails. To manage this, we introduce CARE, an approach for certified accelerator selection. CARE uses paired rollouts on a calibration set to provide finite-sample guarantees that acceleration-induced failure risk stays below a user-specified budget. It deploys the fastest certified candidate, falling back to the reference if none qualify. By relying only on terminal outcomes and measured compute, CARE applies unchanged across diverse acceleration mechanisms, while sequential testing and failure-triggered reference rollouts keep certification affordable. On four LIBERO suites with OpenVLA-OFT, CARE certifies $9.0$–$10.8\times$ speedups while guaranteeing (at $95\%$ confidence) that at least $85.8\%$ of reference-solved episodes are preserved. Under tight budgets, selectors without guarantees exceed the budget in up to $75\%$ of trials, whereas CARE stays within budget and its sequential form uses $78.9\%$ fewer rollouts than exhaustive evaluation. CARE further generalizes to flow-step reduction for $\pi_{0.5}$, and to Qwen3.5-9B and Llama-3.1-8B
agents in Crafter.
Read the full abstract ▾
All authors Rui Liu, Tong Zheng, Jindong Gu, Zhipeng Wang
Solar and Stellar Astrophysics
Debye cooling of ultramassive white dwarfs
David Barba-González and Sivan Ginzburg
Recent observations of cold and massive white
dwarfs are beginning to probe Debye cooling: an accelerated cooling phase driven by the vanishing heat capacity $C_V(T)$ of the crystallized interiors at low temperatures $T$, due to the
quantum nature of
phonon oscillations. We analyse this regime by calculating, both analytically and numerically, the instantaneous cooling time $\tau(L)\equiv -\textrm{d}t/\textrm{d}\ln L$, which is a proxy for the number density of white
dwarfs at a given luminosity $L$. We find that cooling accelerates ($\tau$ shortens) at $L\lesssim 10^{-4}\,\textrm{L}_{\odot}$ due to a combination of Debye cooling and the simultaneous convective coupling of the degenerate core to the surface, with neither process causing significant acceleration on its own. Specifically, strong
magnetic fields can inhibit convective coupling, leading to an almost flat $\tau\propto L^{1/7}$ instead. We parametrize the uncertainty in the solid microphysics by generalizing the Helmholtz free energy such that $C_V\propto T^\delta$ at low temperatures. The currently observed decreasing luminosity function in the Debye regime is tentatively consistent with $\tau(L)$ of non-
magnetic white
dwarfs with $\delta=3$, corresponding to the standard Debye model. Larger samples could probe, in addition to magnetism, deviations $\delta<3$, which diverge appreciably before the
Galactic disc's age $t\approx 10\textrm{ Gyr}$ for masses $M\gtrsim 1.0\,\textrm{M}_{\odot}$. For $\delta>3$, in contrast, the cooling curves converge because the heat capacity becomes dominated by the degenerate electrons $C_V\propto T$.
Read the full abstract ▾
High Energy Physics (Theory)
Cosmological Correlators from Generating Potentials and Projective Geometry
Aswini Bala, Sachin Jain and Deep Mazumdar
We introduce novel scalar generating potentials to compute three and four point cosmological correlators in orthogonal Grassmannian space. These correlators are obtained by applying appropriate directional derivatives on these potentials, combined with helicity-dependent minors. We emphasize that, this is not a
spin-raising analog, but a much more powerful construction. The four point generating potential is associated with the geometry of a projective triangle. First-order directional derivatives generate
gluon correlator, while the Hessian of the same scalar potential produces the four-graviton correlators, thus realizing a differential double copy. Thus, double copy is a natural consequence of the underlying scalar potential as the origin of the
gluon and graviton four point correlators. This framework extends naturally to any
spin-$s$. We exhibit a
geometric origin for the
gluon colour identities: the projective homogeneity of the scalar potential implies Kleiss-Kuijf identity, while the kinematic Jacobi identity follows as a consequence of the closure of the triangle under directional derivatives. We also uncover a connection between the BCFW bridge deformation of the scalar potential and directional derivatives that produces the
spin-s correlator. Thus the generating potential provides a universal geometrical interpretation for any
spin-$s$ four point Grassmannian correlator, unlike \cite{Bala:2026lvw} where
geometric interpretation was only shown for the
gluons.
Read the full abstract ▾
Combinatorics
Positivity for graph Laurent phenomenon algebras
Ross Glew
Laurent phenomenon
algebras were introduced by Lam and Pylyavskyy as a generalisation of cluster
algebras. For a family associated with directed graphs, they showed that every cluster variable is a Laurent
polynomial in the variables of any cluster and conjectured that these expansions have nonnegative coefficients. We prove this
conjecture for all directed graphs without loops or multiple edges with respect to an arbitrary seed cluster.
Read the full abstract ▾
High Energy Physics (Theory)
Worldsheets reconstructed: Duals to two-cut matrix models
Sounak Pal
We compute the higher-rank translation ($\mathcal{T}$) and rotation ($\mathcal{R}$)
matrices associated with a genus-one spectral curve. Using these data, we reconstruct the worldsheet correlators of vertex operators in the corresponding supersymmetric Landau–Ginzburg (LG) theory. This provides a generalization of the genus-zero, one-cut spectral-curve construction to higher-genus geometries. The
matrix-model correlators are mapped to integrals over the
moduli space of Riemann surfaces for the symmetric sextic double-well potential, which has a double-cut solution. We further identify the rational and
elliptic contributions to the worldsheet integrated vertex operators and show that each of the sector precisely matches the corresponding correlators of single-trace operators in the
matrix model. Altogether, our construction showcases an explicit dictionary between the higher-genus spectral-curve data, the associated Cohomological field theory (CohFT), and the worldsheet description of the corresponding
matrix-model observables.
Read the full abstract ▾
Analysis of PDEs
Optimal Linear Inviscid Damping and Vorticity Depletion for Non-monotonic shear flows
Rajendra Beekie, Shan Chen and Hao Jia
We give a proof of optimal pointwise linear inviscid damping and vorticity depletion for a large class of spectrally stable shear flows $b(y)$ on the rectangular
torus. This extends the linear inviscid estimates in \cite{CJWZ25} for Kolmogorov flow $b(y) =\cos y$ using completely different methods. Crucially, both methods remove an apparent logarithmic correction to the depletion and inviscid damping rates. Our method combines pointwise bounds for a parametrix with an explicit solution of a simpler inner problem which makes the removal of the logarithm clear.
Read the full abstract ▾
High Energy Physics (Phenomenology)
Sensitivity targets for dark photon searches in charged kaon decays
M. Fabbrichesi, E. Gabrielli and K. Müürsepp
We investigate the prospects for detecting invisible dark
photons in charged kaon decays through the two-pion channel $K^+\to\pi^+\pi^0\,\gamma_D$ and the one-pion radiative decay $K^+\to\pi^+\gamma\,\gamma_D$, where $\gamma$ and $\gamma_D$ denote the ordinary
photon and dark
photon, respectively. The former decay was previously proposed as a golden channel for massless dark
photon searches at NA62, while here we introduce the radiative process as a complementary probe and extend both analyses to massive dark
photons. The three-body modes are particularly relevant at low dark
photon masses, where the two-body decay $K^+\to\pi^+\, \gamma_D$ becomes strongly suppressed and vanishes in the massless limit. For massless dark
photons, renormalizable interactions induced by kinetic mixing are absent, and the dominant contribution arises from the effective non-renormalizable
flavor-changing
quark interaction $s\to d\, \gamma_D$ mediated by a dimension-five
magnetic-dipole operator. This is also true for massive dark
photons. Their kinetic mixing contributes negligibly to these channels, after considering the existing bounds on the mixing parameter. Only the same
flavor changing
magnetic-dipole operator as in the massless case is relevant. Using the
chiral quark model, we estimate the relevant hadronic
matrix elements of the dipole operator and account for constraints from $K^+\to\pi^+\,X$, where $X$ denotes resonant missing energy, $K^+\to\pi^+\nu\bar\nu$, and $K^0\!\!-\!\bar{K}^0$ mixing. Based on these calculations, we determine the allowed branching ratios for these decay modes and the sensitivities required at NA62 and future charged-kaon facilities.
Read the full abstract ▾
High Energy Physics (Phenomenology)
An early $Z'$ dominance: dark matter, gravitational waves & collider probes
Basabendu Barman, Nayan Das, Partha Kumar Paul et al. (+1 more)
In the minimal anomaly-free gauged B$-$L extension of the Standard Model (SM), we consider a scenario in which the energy density of the new Abelian gauge
boson dominates the Universe prior to big bang nucleosynthesis (BBN), resulting in an early matter-dominated (EMD) era. Taking one of the three right-handed
neutrinos to be a decaying dark matter (DM) candidate, we investigate its production during this EMD phase. We show that the region of parameter space consistent with the relic abundance lies within the projected sensitivity of several gravitational wave (GW) experiments, where the GW signal arises from
cosmic strings formed during the spontaneous breaking of the gauged B$-$L
symmetry. We further constrain the viable parameter space using bounds from
collider experiments, beam dump experiments, BBN, and ultra high energy
neutrino observations, demonstrating the complementarity of multi-messenger probes in exploring physics beyond the SM as well as non-standard cosmological histories prior to BBN.
Read the full abstract ▾
All authors Basabendu Barman, Nayan Das, Partha Kumar Paul, Narendra Sahu
General Relativity and Quantum Cosmology
The price of regularity: core gradient instability in nonsingular dyonic black holes
Lodovico Capuano and Alessandro Longo
Spacetime singularities displayed by classical black hole solutions are expected to be cured by high-energy effects, beyond the regime of validity of General Relativity. On the other hand, there are well-known examples of classical theories allowing for nonsingular black hole solutions. The most prominent is obtained by complementing the gravitational sector of General Relativity with a minimally coupled nonlinear-electrodynamics sector. It has recently been shown that purely electric/purely
magnetic nonsingular black holes in an Einstein-nonlinear-electrodynamics setup are always plagued by a gradient instability. In this work, we derive the quadratic action for linear perturbations of static, spherically symmetric dyonic solutions of Einstein-nonlinear-electrodynamics theories with a general dependence on both invariants. We show that regularity of the center, absence of ghosts and angular gradient stability cannot hold simultaneously: arbitrarily close to a regular center, at least one
photon polarization has a negative squared angular propagation speed. We further show that regularity drives one electromagnetic kinetic
eigenvalue to zero at the center, which may limit perturbative control near the core. Evading the obstruction requires extending the theory, for example by introducing additional degrees of freedom.
Read the full abstract ▾
Cosmology and Nongalactic Astrophysics
Full-Shape Analysis of the Redshift-Space Galaxy Trispectrum in the EFTofLSS
Beatriz Tucci and Oliver H. E. Philcox
The effective field theory of large-scale structure (EFTofLSS) program for spectroscopic
galaxy surveys has been remarkably successful for cosmological inference from the
galaxy power spectrum and bispectrum. However, a full-shape cosmological analysis including the
galaxy trispectrum has not yet been performed. In this work, we develop the tree-level EFTofLSS model of the
galaxy trispectrum, including all relevant ingredients:
galaxy bias, stochasticity, infrared resummation, Fingers-of-God (FoG) counterterms, Alcock-Paczynski distortions, binning effects, and covariance. We validate the model by performing cosmological inference using the one-loop
galaxy power spectrum, tree-level bispectrum, and tree-level trispectrum measured from dark-matter
halos in 25 AbacusSummit $N$-body simulations, whose large effective volume ($V=200h^{-3}\text{Gpc}^3$) allows us to test the precision of our model and determine scale cuts in both real-space and
redshift-space. We also study the cosmological information content of the
galaxy trispectrum in a DESI-like volume, finding that its inclusion can improve constraints e.g., on $\omega_{\rm cdm}$ by up to $15\%$. We further find that the maximum scale cut can be extended to $k_{\max}=0.12h\mathrm{Mpc}^{-1}$ depending on the tracer, thanks to our improved treatment of FoG effects. This provides a path towards using the three-dimensional large-scale structure to constrain new-physics scenarios with distinctive signatures in the trispectrum, in particular
primordial non-Gaussianity.
Read the full abstract ▾
Strongly Correlated Electrons
Junyi Cao and Eduardo Fradkin
We study zero modes bound to charge-conjugation defects in fractional
quantum Hall-
superconductor (FQH-SC)
heterostructures built from Abelian and non-Abelian FQH states. Although the associated defects are well described by $G$-crossed braided tensor categories, a local operator description of their zero modes is not generally known. We generalize the bulk-to-boundary construction of localized zero mode operators to FQH-SCs with non-Abelian parent topological orders, mapping bulk Wilson operators in the folded theory to localized zero mode operators in the one-dimensional FQH-SC theory. This construction distinguishes zero mode anyons from the local zero mode operators that act on the defect
Hilbert space and yields a generalized zero mode
algebra. It also determines the fusion-channel dependence of tunneling: tunneling of zero mode operators probes bulk monodromy, while direct zero mode anyon transfer probes $G$-crossed associativity data. Applying this framework, we recover the $\mathbb{Z}_{2m}$ parafermion zero mode structure of fermionic Laughlin FQH-SCs at filling $\nu=1/m$, obtain a $\mathbb{Z}_2^{(n)}\times\mathbb{Z}_N$ structure for generalized Moore-Read states at filling $\nu=1/m$, with $N=2m$ for fermionic states and $N=m$ for bosonic states, and find a $\mathrm{Fib}\times\mathbb{Z}_N$ structure for the $k=3$ Read-Rezayi state at filling $\nu=3/(3M+2)$, with $N=2(3M+2)$ for fermionic states and $N=(3M+2)/2$ for bosonic states. In the Read-Rezayi case, zero mode operator tunneling and direct zero mode anyon transfer yield distinct universal amplitude ratios between the neutral fusion channels, providing a characteristic signature of the underlying Fibonacci defect structure.
Read the full abstract ▾
Astrophysics of Galaxies
Modeling the nanohertz gravitational wave background with the t-process: Improved frequentist analysis of gravitational-wave power
B. E. Moreschi, K. A. Gersbach, G. Shaifullah et al. (+2 more)
Pulsar timing array (PTA) collaborations have recently reported evidence for a gravitational wave background (GWB) signal in the nanohertz band, which most likely originates from the incoherent superposition of GWs emitted by a population of supermassive black hole binaries (SMBHBs). Since the number of binaries emitting in each frequency bin is finite, and since individual loud sources can dominate the GWB, the power spectral density (PSD) of the background is expected to deviate significantly from the smooth power law that describes an idealized population. PTA searches usually describe the PSD either with a power law or with a free spectrum model, which fits the PSD in each frequency bin as a free parameter. In this paper, we explore the t-process as a third model; it assumes a power law as the underlying spectrum but leaves each frequency bin free to deviate from it. We compare the performance of these three models on simulated
datasets, injecting $100$ realizations of a realistic SMBHB population, to evaluate how well each model recovers a complex PSD. Using
Bayesian and frequentist tools, we find that the t-process model gives the most precise PSD results, with uncertainties $43\%$ and $26\%$ narrower than those obtained with the power law and the free spectrum, respectively, while the improvement in accuracy is smaller. Moreover, using the t-process model, we compute isotropic and anisotropic signal-to-noise ratios with median gains of about $46\%$ over the power law and $65\%$ over the free spectrum in the first nine frequency bins. The t-process also returns lower anisotropy p-values and among the frequency bins with $p<0.01$, it allows to localize the loudest binary within the angular resolution of the sky maps in $14.3\%$ of the cases, compared to the $10\%$ for the power law and the $6.2\%$ for the free spectrum.
Read the full abstract ▾
All authors B. E. Moreschi, K. A. Gersbach, G. Shaifullah, S. R. Taylor, A. Sesana
Astrophysics of Galaxies
KDG 162: A Nearby Isolated Star-forming Dwarf Galaxy
Gagandeep S. Anand, Igor D. Karachentsev, Meredith J. Durbin et al. (+6 more)
The nearby
dwarf galaxy KDG 162 (also known as KK 153) has recently attracted
attention as a potentially gas-rich, isolated, and low-mass
dwarf galaxy. However, many of these characterizations depend on the true distance to KDG 162. Here we present new Hubble Space
Telescope (HST) imaging that resolves individual
stars down to 3 magnitudes below the tip of the red
giant branch. Using this HST imaging, we determine a distance to KDG 162 of D = 3.48 $\pm$ 0.12 Mpc, somewhat higher than previous determinations. We find KDG 162 to be relatively isolated and lying at the edge of first infall onto the M81 group. We use its
stellar content to measure its recent
star formation history and find that only a modest amount of
star formation can be attributed to within the last $\sim$500 Myr. The present-day
stellar mass outweighs the H I mass by a factor of 2$\times$. The properties of KDG 162 add to an increasingly detailed picture of nearby and isolated low-mass
galaxies. In the future, deeper observations would allow for reaching more age-sensitive populations, enabling the recovery of its
star-formation history back to ancient times.
Read the full abstract ▾
All authors Gagandeep S. Anand, Igor D. Karachentsev, Meredith J. Durbin, R. Brent Tully, Rachael Beaton, Andrew Dolphin, Serafim S. Kaisin, Lidia N. Makarova, Edward J. Shaya
High Energy Physics (Theory)
The Cosmological $\delta$-shift
Carolina Figueiredo, Jiajie Mei, Yuyu Mo et al. (+1 more)
We consider the extension of the $\delta$-shift connecting Tr$(\phi^3)$ and pion amplitudes to observables in cosmology. We start in flat-space and explain how one can extract the pion wavefunction and correlators from those of Tr$(\phi^3)$ theory, via a simple kinematic deformation. As opposed to the amplitude, the $\delta$-shift for the wavefunction is no longer a linear deformation on the kinematics since the wavefunction depends on magnitudes rather than the squares of momenta. We "linearize" this deformation by using the frequency representation - allowing us to recast the contribution from any graph as a sum over residues. We then show how this construction extends to de Sitter, and find that, remarkably, in the frequency representation, the pion de Sitter wavefunction and correlators can be formulated in precisely the same way as in flat-space. Amongst other things, this unified formulation reveals a simple universal soft-limit for flat-space and de Sitter, where, in the soft-limit, the wavefunction turns into a sum of wavefunctions in a mixed theory of pions and scalars.
Read the full abstract ▾
All authors Carolina Figueiredo, Jiajie Mei, Yuyu Mo, Francisco Vazão
High Energy Astrophysical Phenomena
Extreme Stellar Death and Galaxy Feedback at z = 2: A Rest-frame Ultraviolet Characterization of the Strongly Lensed Superluminous Supernova 2025wny
Willem B. Hoogendam, David O. Jones, Christopher J. Storfer et al. (+26 more)
We present photometric and spectroscopic follow-up observations of the strongly-lensed H-poor superluminous
supernova (SLSN) 2025wny at $z = 2.0155$. We use integral-field Keck-II/KCWI spectroscopy to obtain rest-frame ultraviolet spectra spanning $11$ to $61$ days after maximum light. We use Pan-STARRS photometry to constrain the brightest image's magnification, finding $\mu_A=19.4\pm1.6$, a factor of $\sim$2-4 higher than mass-modeling estimates. We fit the Pan-STARRS photometry with a magnetar energy-injection model including $^{56}$Ni decay using the Modular Open Source Fitter for Transients (MOSFiT); the model can reproduce the near-maximum light curve with parameters typical of local universe SLSNe-I, including a magnetar
spin period $P_{\rm spin}=4.1^{+0.7}_{-1.0}\,\mathrm{ms}$ and
magnetic field $B_{\perp}=2.4^{+1.1}_{-0.8}\times10^{14}\,\mathrm{G}$; however, the late-time plateau suggests an additional power source. Spectroscopically, iron-group element line blanketing is virtually absent in the rest-frame UV spectra, and the absorption features match previously identified SLSN-I features from abundance
tomography. Based on narrow ISM line
metallicity estimates, SN 2025wny exploded in a sub-
solar-
metallicity galaxy. From host-
galaxy lines in our rest-frame UV-to-optical spectrum we estimate high-velocity
outflows of $\sim$575 km s$^{-1}$ and an exceptionally high Ly$\alpha$ escape fraction of $f_{esc}^{Ly\alpha}=0.23\pm0.03$. SN 2025wny matches lower-
redshift SLSNe reasonably well. Upcoming surveys like LSST and Roman will discover additional high-
redshift SLSNe, further testing whether SLSNe-I explosions remain consistent at the population level in the early universe.
Read the full abstract ▾
All authors Willem B. Hoogendam, David O. Jones, Christopher J. Storfer, Luca Izzo, Kyle W. Davis, Miranda Y. Kong, Katie Auchettl, Aadya Agrawal, David R. Aguilera-Dena, Kenneth C. Chambers, Siyuan Chen, Thomas de Boer, Ryan J. Foley, Matthew Grayling, Jason T. Hinkle, Jens Hjorth, Mark E. Huber, Chien-Cheng Lin, Thomas B. Lowe, Eugene A. Magnier, Kaisey S. Mandel, C. Tanner Murphey, Gautham Narayan, Gregory S. H. Paek, Kishore C. Patra, Haille M. L. Perkins, Armin Rest, Benjamin J. Shappee, Richard J. Wainscoat
Artificial Intelligence
Agent Plasticity: Measuring Self-Improvement Through Experience
Harman Singh, Anton Bakhtin, Rulin Shao et al. (+8 more)
AI
agents increasingly operate in environments where they can diagnose failures and improve through experience, yet existing evaluations largely measure what an
agent can do at a fixed point in time rather than how effectively it learns. Evaluating self-improvement requires answering three questions: does future performance improve and generalize beyond the interactions that enabled learning; how efficiently are new capabilities acquired; and where does the self-improvement process break down? To answer these questions, we study self-improvement in a controlled setting where
agents amortize past experience into reusable artifacts that are inherited by future instances. At each checkpoint, we measure performance on training and held-out environment interactions while accounting for learning cost. We introduce
agent plasticity, the efficiency with which an
agent converts experience into gains in future held-out performance. Across multiple environments, frontier models exhibit sharply different improvement trajectories despite comparable opportunities to learn. Some achieve substantial and persistent gains, while others remain near or below their initial performance, and gains within the training regime often transfer only partially to out-of-distribution conditions. Endpoint capability and acquisition efficiency also diverge: the
agent that ultimately performs best need not be the one that improves most efficiently. Tracing failures through the improvement loop further reveals different candidate bottlenecks.
Agents with low plasticity often fail to reuse relevant artifacts, whereas more plastic
agents may still fail despite reusing relevant artifacts, pointing to limitations in artifact quality, generalization, or application. Evaluating self-improving
agents requires measuring not only what they can do, but how effectively they become better through experience.
Read the full abstract ▾
All authors Harman Singh, Anton Bakhtin, Rulin Shao, Gabriel Synnaeve, Ilia Kulikov, Rob Fergus, Sanjeev Arora, Kurt Keutzer, Jason Weston, Anuj Mahajan, Anirudh Goyal
Artificial Intelligence
Sequential Probabilistic Uncertainty Estimation for Parallel Multi-Agent Reasoning Systems
Tunyu Zhang, Zihao Zhao, Yusong Zhao et al. (+5 more)
LLM-based multi-
agent systems (MAS) have attracted growing
attention for improving
reasoning through interaction among multiple
agents. In this work, we focus on parallel multi-
agent reasoning systems, where several
agents solve the same problem over multiple rounds and aggregate their outputs into a final answer. Despite their strong
reasoning performance, uncertainty estimation for such systems remains underexplored: the reliability of a MAS depends not only on individual generations, but also on how
agents interact and evolve across rounds. We propose SAUCE (Sequential
Agent Uncertainty through Consensus
Evolution), a lightweight, training-free uncertainty
estimator that formulates MAS uncertainty as sequential inference over a latent system-level belief. SAUCE aggregates round-level agreement and generation-uncertainty signals through a filtering-style update. Across five backbones, five
benchmarks, and two MAS protocols, SAUCE improves misclassification detection, selective prediction, and calibration over a broad set of uncertainty estimation baselines, including standard log-likelihood-based methods and MAS-specific
estimators.
Read the full abstract ▾
All authors Tunyu Zhang, Zihao Zhao, Yusong Zhao, Haizhou Shi, Zhuohang Li, Haoxian Chen, Hao Wang, Dimitris N. Metaxas
Artificial Intelligence
Humanize: Judgement Engineering for Agentic Coding
Sihao Liu, Ligeng Zhu, Zijian Zhang et al. (+6 more)
Agentic coding makes code generation cheap, but reliable completion remains difficult: the
agent that writes the code is a weak judge of whether it is done. We present Humanize, a multi-
agent orchestration workflow for agentic coding built around judgement engineering: explicit, mechanically enforced decisions at the boundaries between planning, implementation, review, and learning. A human approves a plan contract, a builder
agent implements it in rounds, and a reviewer
agent from another vendor decides completion; deterministic hooks, not a model, route work between these roles and enforce 72 mechanical gates. Viewed as a Markov chain over repository states, alternating builder and reviewer samples jointly from two models, so a defect survives only if both miss it. We study Humanize through its deployment, 118 public postmortems of real loops, and its applications. Over 68 versions in 108 days, it gathered 1,468 GitHub
stars. Applications include a 567-file gem5 build-system migration under upstream review; Kernel Design
Agents, which extend the loop with a kernel knowledge base and profiling feedback and placed in the top three of all three Full-
Agent tracks of the MLSys 2026 FlashInfer contest; and, through Humanize Olympiad
Agents (HOA), full scores in IOI 2026, IMO 2026, IPhO 2026, and IBO 2024, 418.5/437 in IChO 2026 (gold-medal). Humanize also achieves 672/672 on PutnamBench and ranks first (251/303) on Lean-Eval's leaderboard even competiting with professional mathematicians. The postmortems show that independent review catches unsupported builder claims, but stopping remains a key weakness. In reports that separate rounds by phase, two thirds of rounds occurred after implementation was accepted. This evidence is observational, not a controlled comparison of workflows.
Read the full abstract ▾
All authors Sihao Liu, Ligeng Zhu, Zijian Zhang, Dongyun Zou, Zhengyang Zhang, Changye Li, Song Bian, Song Han, Tony Nowatzki
Performance
CXDVirt: Low Latency Kernel Module Based CXL-SSD Emulation
Hyunsun Chung, Seongho Bong, Hong-Yeon Kim et al. (+1 more)
CXL-SSDs promise
memory-semantic access to NAND-scale capacity through a small in-device DRAM cache, but scarce hardware makes emulation central to CXL-SSD research. Cylon, a state-of-the-art emulator, runs workloads in a VM, intercepting DRAM misses to inject NAND latency. We show this VM machinery adds over 3$\mu s$ per miss, exceeding high-performance NAND read latency, while repeated VM exits inflate tail miss latencies several-fold. We present CXDVirt, a kernel-module-based CXL-SSD emulator that removes the VM entirely: the device is exposed through host devdax, hits run natively via the MMU, and a custom page-fault path models NAND latency on misses. CXDVirt preserves CXL-SSDs' bimodal latency profile while reducing per-miss latency 1.4-1.6$\times$ (4.1$\times$ at P99.9); with eight threads its miss latency only doubles, against 8-9$\times$ for Cylon behind QEMU's global lock. CXDVirt stays within 5% of remote DRAM without NAND traffic, where Cylon is 3.1-3.6$\times$ slower, and enables configurable eviction and prefetch policy studies.
Read the full abstract ▾
All authors Hyunsun Chung, Seongho Bong, Hong-Yeon Kim, Youngjae Kim
Logic in Computer Science
Proceedings Seventeenth International Symposium on Games, Automata, Logics, and Formal Verification
Giorgio Bacci and Mickaël Randour
This volume contains the proceedings of GandALF 2026, the Seventeenth International Symposium on Games, Automata, Logics, and Formal Verification. The symposium was held in Aalborg, Denmark, on September 15-17, 2026. The aim of GandALF 2026 is to bring together researchers from academia and industry who are actively working in the fields of Games, Automata, Logics, and Formal Verification. The idea is to cover an ample spectrum of themes, ranging from theory to applications, and stimulate cross-fertilisation.
Read the full abstract ▾
Quantum Physics
Visualizing two qubits
Simon Grosse-Holz
I propose a simple visualization for pure two-
qubit states, constructed from three components: two Bloch balls for the marginal
qubits (the mixed states obtained by ignoring the other
qubit, respectively) and an
entanglement component consisting of a rotoreflection and a contraction. I prove that this picture accurately captures the local and global structure of the relevant state space without gaps or degeneracies: the underlying parametrization is a global diffeomorphism onto the state space. Finally, I use this picture to visualize conditional states, the
entanglement degree(s) of freedom, and single
qubit rotations.
Read the full abstract ▾
Software Engineering
Psychological Safety in Software Engineering Teams: A Systematic Mapping Study of Team Processes and Performance
Adnan Manzoor, Muhammad Azeem Akbar, Jussi Kasurinen et al. (+3 more)
Context: Psychological safety is an important human and behavioral factor in software development teams, supporting collaboration, knowledge sharing, interpersonal risk-taking, and learning. However, evidence in software engineering remains fragmented across team processes, performance outcomes, contextual conditions, barriers. Objective: This study maps literature on psychological safety in software engineering, examining its conceptualization, associations with team performance and processes, contextual influences, barriers, and strengthening strategies. Method: We conducted a systematic mapping study following established software engineering guidelines. We analyzed 112 primary studies published between 2006 and June 2026 using descriptive statistics and thematic analysis across six research questions. Results: Psychological safety emerges as a multidimensional, context-dependent, team-level construct characterized by interpersonal risk-taking, open communication, collaboration, learning, trust, and supportive interaction. It is predominantly associated with productivity, software quality, innovation, team satisfaction, collaboration, communication, and decision-making. These relationships are shaped by leadership, organizational culture, work environment, trust, and sociocultural and individual conditions. Recurring barriers include communication problems, organizational constraints, leadership issues, diversity and inclusion challenges, and workload pressures. Although diverse practices are reported, evidence of their effectiveness and long-term sustainability remains limited. Conclusion: Psychological safety is a multidimensional, context-dependent team-level construct associated with team interaction and performance. This study highlights the need for longitudinal, context-sensitive research on psychological safety interventions in software development environments.
Read the full abstract ▾
All authors Adnan Manzoor, Muhammad Azeem Akbar, Jussi Kasurinen, Muhammad Hamza, Aakash Ahmad, Abdul Razzaq
Cryptography and Security
Contextualization of Third-Party Cloud Security Findings
Leon Goldberg and Gal Engelberg
Finding severity is the main driver of how security teams prioritize remediation. For third-party cloud security findings, that severity is static: the rule that raised the finding assigns it before the rule meets any environment, so it reflects the risk of the condition in general rather than the risk the finding poses to the concrete environment where it lives. Scoring standards define where environment-specific context belongs. How far that context changes finding severities in production, where the deciding evidence lies, and whether it holds against the live environment have not been measured. We address this gap with contextualization, re-deriving each finding's severity from evidence in the environment where the finding lives. A deep research
agent over a precomputed cross-signal asset graph investigates each finding against the resource's state, its graph neighborhood, and other products' signals, and returns an adjusted severity with an evidence trace. We evaluate it in a production field study of 9,967 vendor HIGH findings from two commercial cloud security platforms across eight real production environments, on three criteria: the faithfulness of the facts behind each verdict to the live environment, the dependence of each decision on context beyond the flagged resource, and the regularity of the
reasoning. Three in four findings are re-graded, mostly downward, and the same rule often moves in opposite directions inside a single environment. About half of the decisive evidence lies beyond the flagged resource, and read-only probes of live infrastructure confirm the decisive fact for 99.4% of decided findings.
Read the full abstract ▾