Peer Revenue Play credits. Real papers.

Papers

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

All of arXiv

Tuesday 8 September

Computation and Language

Performance of Clinical AI System and Physicians and Frontier Language Models in primary care diagnostics

Clinical AI evaluation should encompass diagnosis and management after adaptive information gathering. We compared Doctorina, eight physicians and four standalone frontier language models in 150 synthetic Polish-language primary-care consultations. Doctorina achieved 82.0% Top-1 concordance versus 57.0% for physicians (difference, 25.0 percentage points; 95% confidence interval, 17.7-32.7) and 97.3% versus 85.0% primary-or-reference-differential concordance. Across 149 case pairs, normalized workup and treatment scores were 89.4 versus 66.9 and 83.7 versus 61.2. Doctorina had the highest diagnostic point estimates among all six groups; Kimi K3 ranked next, while Claude Opus 5 led the closely spaced management estimates of Opus, Doctorina and Kimi. A second Doctorina execution reproduced the advantages over physicians across all outcomes. Doctorina's advantage over physicians therefore extended from primary-diagnosis selection to higher-rated diagnostic workup and initial treatment after adaptive consultation.
Read the full abstract

All authors Andy Nkansah, Hanna Plotnitskaya, Stanislau Salavei, Anna Kozlova, Piotr Gibas, Julian Milek, Viktar Harbachou, Aleksey Ropan, Pavel Satalkin

Robotics

Rethinking Learned Occupancy in Autonomous Active Mapping with Observation-Gated Filtering

Autonomous 3D active mapping requires a space robot to choose where to sense while building the geometry needed for navigation. Learned occupancy completion extends spatial context beyond the current field of view, but one predicted map often serves two planning roles: it scores expected surface gain and constrains collision-free motion. Unsupported occupancy can therefore distort both where the robot looks and where it believes it can travel. We study this coupled interface in a controlled closed-loop benchmark by holding the active-mapping system fixed and varying only its planner-facing occupancy across observation-only, learned, oracle-corrected, and ground-truth conditions. Improving occupancy accuracy does not monotonically improve closed-loop coverage: across 25 starts, planning with ground-truth occupancy reaches 70% of the learned baseline's final coverage 12.7 steps earlier on average, while increasing final coverage by only 0.031. Guided by this diagnosis, we introduce an observation-gated filter that retains completion in insufficiently observed regions and suppresses predictions only after repeated frustum exposure without nearby RGB-D support. The filter improves both targeted failure-prone starts without retraining or ground truth. These results motivate online revision of planner-facing geometry during autonomous intervals between communication windows. The current study assumes benchmark RGB-D observations and sufficiently accurate pose estimates; planetary sensing conditions and accumulated localization drift remain to be evaluated.
Read the full abstract

All authors Jiahui Zhang, Bonian Han, Gongbo Liang, Yu Zhang

Quantum Gases

Renormalization of One-Dimensional Semirelativistic Bosons with Contact Interactions

We study one-dimensional spinless bosons with semirelativistic spinless-Salpeter dispersion and attractive pairwise contact interactions. Because the dispersion becomes linear at large momentum, the contact interaction is marginal by power counting and produces a logarithmic ultraviolet divergence. We construct the renormalized many-body theory using an enlarged Fock space and a Schur-complement representation of the resolvent, eliminating the bare coupling in favor of the physical zero-total-momentum two-body bound-state energy. The resulting cutoff-independent resolvent defines a self-adjoint Hamiltonian in each fixed particle-number sector. We treat the two-body problem explicitly and show, in the norm-resolvent sense, that the nonrelativistic limit reproduces the attractive Lieb–Liniger Hamiltonian. We also formulate a mean-field approximation directly within the renormalized theory. In the massless and deeply bound large-particle-number regimes, it predicts an exponentially increasing binding scale whose exponent is determined by a one-dimensional variational problem.
Read the full abstract

Strongly Correlated Electrons

On spectral properties of Generalized Kadanoff–Baym Ansatz

The Generalized Kadanoff–Baym Ansatz with mean-field propagators is increasingly used to simulate non-equilibrium quantum fermionic and bosonic systems. Compared with the full Kadanoff–Baym equations, its mean-field propagators substantially reduce computational cost, enabling simulations with computational effort that scales linearly with propagation time. However, the time-diagonal structure of the Generalized Kadanoff–Baym Ansatz obscures spectral properties within the collision integral of the transport equation. Here, we recover and investigate these hidden spectral properties using the extended Generalized Kadanoff–Baym Ansatz. For a Hubbard cluster, we compare ground-state spectral functions obtained with the extended Generalized Kadanoff–Baym Ansatz against those from standard Kadanoff–Baym equations. At moderate interaction strengths, the spectral functions show good agreement. At large interaction strengths, however, significant deviations emerge, and the spectral function obtained with the extended Generalized Kadanoff–Baym Ansatz becomes negative. These results demonstrate that, despite its computational advantages, the Generalized Kadanoff–Baym Ansatz has important limitations in strongly interacting regimes and may produce unphysical spectral properties when interactions become sufficiently large.
Read the full abstract

Robotics

A Distributed Consensus Particle Filter for Target Tracking using Autonomous Surface Vessels

Maritime target tracking over large distances often requires multi-agent teams without centralized coordination, and intermittent communication. Each agent must maintain an independent estimate that can take advantage of opportunistic communications availability when possible. This can lead to overly confident local estimates in the absence of external data. In this work, we propose an augmentation to a classical particle filter implementation that accounts for this potential source of error by forcing particles to spread strategically in the absence of informative updates from other sensor nodes. We demonstrate our method using Unmanned Surface Vessels (USVs) on a lake, and show that our augmentations do not deteriorate nominal performance, and provide an advantage in some specific edge cases.
Read the full abstract

All authors Carter Noh, Kyle Crandall, Connor Yates, Corbin Wilhelmi

Mesoscale and Nanoscale Physics

Divergent Orbital Diamagnetism from Chiral Edge States in Chern Insulators

Two-dimensional massless Dirac systems, exemplified by graphene, are known to exhibit a divergent orbital diamagnetic susceptibility that scales linearly with system size. Motivated by viewing a chiral edge state as one half of an enlarged analogue of a benzene ring, we study Chern insulators under open boundary conditions and find the same divergent scaling. This giant diamagnetism is robust against disorder, revealing its topological nature. Our results further suggest a profound connection to the divergent diamagnetism of massless Dirac systems.
Read the full abstract

Data Structures and Algorithms

Parity and Pattern Detection in Permutation Streams

Consider a permutation of $[n]$ whose values arrive one at a time. We resolve two questions about the space needed to decide natural properties of such input: First, computing the parity of the permutation requires $\Theta(n)$ bits, even with randomization and constant error, and a constant number of passes. Second, every permutation pattern of length three can be detected deterministically in one pass using $O(\log n)$ bits. Together with the 2026 lower bounds of Berendsohn, this completes the classification of fixed permutation patterns; The optimal space complexity is $\Theta(\log n)$ for monotone patterns and patterns of length at most three, and $\Theta(n)$ for every other pattern. As a consequence, we observe that we can verify BST traversals in streaming with logarithmic memory.
Read the full abstract

Earth and Planetary Astrophysics

The size and mass distribution of cold classical TNOs for $5<H<13$

The cold classical trans-Neptunian objects (CCs) are the only observable in situ population of planetesimal remnants believed to have escaped collisional grinding. Recent JWST observations make it possible to fit the differential absolute magnitude distribution $dN/dH$ of the CCs from $5<H_r<13,$ and infer the differential mass distribution $dN/dM$ across $\ge5$ orders of magnitude in mass. We find $dN/dM$ well fit by a lognormal distribution, and equally well by a generalized $\Gamma$ distribution or a double power law. The maximum fraction of the total CC mass per log interval in $M$ is in bodies near 75~km diameter, or $\approx10^{-7.5}M_\oplus.$ Extrapolation of the fitted $dN/dH$ functions to $H_r>13$ is unwise, as the different analytic forms diverge. It remains unclear if $dN/dH$ turns over at faint $H.$ The uncertainty in the total mass of the CC belt is dominated by uncertainty in the relation between $M$ and $H$. A calibration using CC binaries suggests a total CC mass of 1.7–2.7$\times10^{-3}\,M_\oplus.$ A trend toward lower density and/or higher albedo for smaller bodies may be present in the data, and would lower the estimated total CC mass. Qualitative comparison of the derived mass distribution to the results of numerical simulations of the streaming instability (SI) suggest the simulations produce $dN/dM$ distributions that are more sharply peaked, and steeper at the bright end, than the CCs. Such differences could be ascribed to inhomogeneous formation conditions in the classical belt that are not yet included in modeling. The variety and uncertainty of $dN/dM$ derived from state-of-the-art SI simulations currently preclude any definitive test of the SI hypothesis.
Read the full abstract

All authors Gary M. Bernstein, Wesley C. Fraser, Marielle R. Eduardo, William M. Grundy, Bryan Hilbert, Matthew J. Holman, Anastasia N. Morgan, Kevin J. Napier, John A. Stansberry, David E. Trilling

Machine Learning

Multi-Task Learning for Sparsely-Labeled Time Series: A Case Study on Cold-Hardiness Modeling

We present a real-world case study of multi-task learning (MTL) for temporal process modeling from limited data with temporally sparse labels. Specifically, we investigate multi-task learning for the important agricultural problem of predicting grape cold hardiness, which is the temperature at which lethal freezing occurs. Cold hardiness changes in response to weather and is difficult to measure directly in the field. Thus, growers rely on predictions to decide when to apply costly frost mitigation measures. We apply recurrent neural networks (RNNs) for daily cold-hardiness prediction from time series weather data. A major challenge is that the cold hardiness response varies across plant cultivars and ground-truth data for each cultivar is temporally sparse and limited. To address this challenge, we investigate multi-task learning (MTL) approaches for combining data, where different tasks correspond to different cultivars. We develop a variety of MTL architectures and evaluate them in both MTL and transfer learning settings. Our results show significant differences between architectures and that certain architectures are able to consistently outperform single-task learning and state-of-the-art scientific models. Additionally, we show similar results for the qualitatively different, but related, task of budbreak prediction. Further, improved accuracy for budbreak and cold hardiness is achieved by a single MTL model that simultaneously learns both tasks.
Read the full abstract

All authors Aseem Saxena, Paola Pesántez-Cabrera, Markus Keller, Alan Fern

Human-Computer Interaction

Location-Independent Robot-Assisted Finishing Using Digital Twins and Extended Reality

This paper presents a cyber-physical system (CPS) for location-independent programming, supervision, training, and teleoperation of a Robot-Assisted Finishing (RAF) system used to post-process metal additive-manufactured (AM) components. A digital twin (DT) built in Unity is delivered to the operator as a WebGL application that supports both desktop and immersive modes through WebXR-compatible devices. Moreover, it exchanges robot state and pose commands with a collaborative robot through a Message Queuing Telemetry Transport (MQTT) broker. The DT enforces kinematic and collision constraints before a pose is released to the physical robot, and augments the virtual component with a color map of the surface topography that supports operator decisions on part repositioning or process termination. The architecture was validated on a specially designed physical RAF system. A steady-state joint synchronization error of 0.12 deg and a mean round-trip latency of 563 ms were measured, which is adequate for supervisory programming and intermittent teleoperation.
Read the full abstract

All authors Jose Outeiro, Jia Holt, Tero Kaarlela, Khalil Chakal

Systems and Control

Efficient Quantile-Resolved Hosting Capacity Assessment on Nodal Level for Low-Voltage Grids

Hosting capacity - the maximum additional capacity a network can accommodate without violating operational limits - is a key metric in distribution system planning and operation. Decisions on grid reinforcements and the deployment of flexibility management require not only the worst-case HC but also an understanding of the distribution of HC under different likelihoods in load and generation patterns. Quantile-resolved HC distributions provide this view by expressing HC as a function of an acceptable operational limit exceedance likelihood. Monte Carlo sampling is the established approach for computing such distributions but demands large computational resources. Approximations sacrifice either accuracy, the ability to capture uncertainty correlations, or scalability when assessing real-world networks. This paper introduces a computationally efficient method for calculating distributions for quantile-resolved HC. It uses a representation of load samples as multivariate normal distribution, propagated through a linearized power flow model. This allows for leveraging a re-parametrized AC-OPF problem for each hosting capacity quantile. Benchmarking against Monte Carlo-based methods on realistic LV networks demonstrates that the proposed method achieves comparable accuracy with a mean deviation of approx. 3%, while reducing computational time by orders of magnitude. For the exemplary networks the computational time decreases from 11 min to 2 s, and 38 h to 50 s, respectively. The method's scalability is also suitable for recalculation in 15-minute cycles encountered in DSO practice for e.g., real-time grid management.
Read the full abstract

All authors Maximilian Köhler, Edwin Mora, Mathias Duckheim, Stefan Niessen

Machine Learning

PlayTrain: An Efficient Reinforcement Learning Framework for LLM-Generated Adaptable JavaScript Games

While many video-game environments (VGEs) have played crucial roles in advancing reinforcement learning (RL), developing novel VGEs or modifying existing ones to support new features, has been a laborious process requiring extensive hand-coding. Here we present PlayTrain, an RL framework that combines the abilities of large language models (LLMs) to robustly generate JavaScript (JS) games from a minimal human prompt, and an efficient pipeline that can run any JS game in a standard 'gym' environment. Not only are recent LLMs particularly good at writing JS code, but the JS format also allows users to easily play generated VGEs, while PlayTrain enables us to train RL agents on the exact same games. We demonstrate multiple use cases of PlayTrain, including cloning well-known Atari and ProcGen games in simple JS, where PlayTrain trains pixel-based agents end-to-end at over 1M agent-decisions per second on a single GPU node; and creating modified versions thereof (e.g., that support novel test sets, procedural generation logics, or game dynamics). Through PlayTrain, we reimagine RL VGE development: all we need is a single JS file, generated and modified through an LLM. We discuss promising future RL research directions that PlayTrain unlocks.
Read the full abstract

All authors Ryan Truong, Lance Ying, Samuel J. Gershman, Kazuki Irie

Group Theory

Images of word maps with constants on algebraic groups

We study word maps with constants on quasisimple algebraic groups over a local field $L$. We prove that, for such a group $G=\mathbf{G}(L)$ and a word $w\in (G\ast\mathbf{F}_r)\setminus G$, either $w$ has a so-called Tomanov-small critical constant or the minimal dimension of the word image $w(G^r)\subseteq G$ of such a word is bounded from below by a function $c(G)>1$. We compute the optimal value of $c(G)$ for most quasisimple linear algebraic groups.
Read the full abstract

Soft Condensed Matter

Discretization strategies for colloidal particles in multiparticle collision dynamics simulations

Discrete models are frequently used in multiparticle dynamics simulations to capture hydrodynamic interactions between colloidal particles and the solvent as well as to represent anisotropic pairwise interactions between colloidal particles; however, there is currently limited guidance on how to reliably parameterize these models. Here, we first compare strategies for selecting the density and mass of discrete surface sites used to couple colloidal particles to the solvent, finding that using a minimum of 2 sites per unit area with a scheme that matches the total mass and moment of inertia for a neutrally buoyant solid particle produces reliable and accurate results for the transport properties of colloidal particles at both infinite dilution and in suspension. We then compare strategies for representing the excluded volume of nearly-hard shape-anisotropic colloidal particles using a collection of discrete interaction sites with isotropic repulsion, finding that having a discrepancy between the nominal volume and the excluded volume of the particle can significantly affect suspension transport properties. This discrepancy can be mitigated by placing the interaction sites inside and tangent to the surface of the colloidal particle. We expect these findings to help construct discrete models for colloidal particles with less sensitivity to parameterization.
Read the full abstract

Category Theory

Time-Varying Data as Sheaves: an Invitation to Narratives

Modern science and engineering increasingly rely on time-varying data, yet the mathematical tools used to model temporal phenomena are often developed within separate disciplines, obscuring common principles and limiting the transfer of ideas across fields. This chapter presents the theory of narratives, an abstract framework for time-varying objects of any mathematical kind that supports both theoretical investigations and applications. To illustrate this perspective, the chapter develops three vignettes, each illustrating a different research direction. The first addresses a general concern: What information loss can occur when switching between different representations of temporal data? The second concerns structural and algorithmic approaches: How can we systematically decompose time-varying data into simple pieces and obtain invariants describing its structural complexity? The third is an application to control theory: How can we model multi-agent systems with switching communication topologies? More important than any individual vignette, the central message of this invitation is that a suitable abstract perspective can organize and guide research across remarkably diverse mathematical and scientific domains.
Read the full abstract

All authors Wilmer Leal, Benjamin Merlin Bumpus, Jana K. Nickel, Johan García, James Fairbanks, Warren Dixon

Computer Vision

"World Knowledge" in the Weights: Reading Concept Circuits of Vision Transformers

Vision transformers (ViTs) have achieved remarkable generalization across visual domains, yet little is known about how they internally represent the structure of the world. To address this gap, we use Cross-Layer Transcoders (CLTs) to read concept circuits from ViTs: directed graphs whose nodes correspond to sparse, interpretable concepts and edges capture concept interactions across layers. Our method yields two complementary views of model behavior. The global concept circuit is input-invariant and can be recovered directly from learned cross-layer weights, exposing the reusable "world knowledge" encoded in the model. The instance concept circuit is input-dependent and identifies the concepts and pathways actually used for a specific prediction, enabling faithful example-level explanations. We demonstrate the utility of concept circuits in three ways: (1) Automatic spurious correlation discovery: leveraging the statistics of our global concept circuits to identify shortcut dependencies within the model. (2) Spurious correlation removal: intervening on the instance concept circuit to steer the model towards correct predictions. Empirical results show that our method outperforms existing counterparts by 11.0% on the Waterbird dataset. (3) Model comparison: contrasting the global concept circuits of different foundation models (e.g., CLIP vs. DINO) to reveal how supervision paradigms shape representational structure. Our code is available at https://github.com/deep-real/VisionCLT
Read the full abstract

Machine Learning

Training-Free Task Vectors for LLM Behavioral Control

Task vectors enable post-training model editing by identifying semantically meaningful directions in weight space, typically computed as the difference between a fine-tuned model and its pretrained initialization. However, this reliance on fine-tuning makes discovering such directions costly and limits the practicality of post-training model editing. To address this limitation, we introduce Training-Free Task Vectors (TFTVs), a novel method to compute task-vector-like directions without requiring fine-tuning. Our method maps activation steering vectors to rank-one weight-space edits using only forward-pass statistics, while satisfying arithmetic properties that directly support learning via addition, forgetting via subtraction, and the composition of multiple edits. Empirically, we evaluate TFTVs on large language model behavioral control tasks and show that they consistently amplify, suppress, and compose target behaviors while preserving general knowledge and problem-solving skills. We also validate our method against other editing and steering baselines, experimentally demonstrating that TFTVs achieve stronger trait control with better or competitive utility preservation. We hope our work opens new directions for the community in post-training model editing and broader training-free model control. Code is available on the project website: tftv-llm.github.io.
Read the full abstract

All authors Gabriel J. Perin, Lucas Boscaini, André Araujo, Nina S. T. Hirata

Numerical Analysis

Quadratic Point Estimate Method for Uncertainty Quantification with Dependent Non-Gaussian Inputs

As an extension of the Point Estimate Method (PEM) to evaluate probabilistic moments of quantities of interest (QoI) in general $n$-dimensional spaces, the Quadratic Point Estimate Method (QPEM) has been recently developed. This new method is defined to fully represent up to fifth-order input moments in the Gaussian space, providing general analytical expressions for sample locations and weights, without requiring any numerical optimization. The QPEM can significantly improve the estimation accuracy of the output QoI moments, in relation to PEM-based methods whose numbers of sigma points grow linearly with the problem dimension, while at the same time having an affordable and competitive computational cost up to a considerable number of dimensions. The QPEM is further enhanced in this work by enabling copula integration into the framework, which enables effective modeling of the joint input probability density function by estimating marginals and the dependence structure of the involved random variables. The validity and efficient performance of the copula-based QPEM are showcased against numerous other sampling methods in various examples considering two practical scenarios: (i) when the joint dependence structure can be inferred from data, and (ii) when only marginal distributions and correlation matrices are known.
Read the full abstract

Group Theory

On the complement of a union of cosets

Let $g_1H_1,\ldots,g_nH_n$ be left cosets of subgroups of a group $G$ whose union $U$ is a proper subset of $G$. We prove that $G$ is the union of at most $2^n$ left translates of $A:=G\setminus U$. For finite $G$ this yields $|A|\ge|G|/2^n$, which settles a conjecture of Tărnăuceanu and the author. Equality $|A|=|G|/2^n$ can only hold when $A$ is a left coset of $K:=H_1\cap\ldots\cap H_n$. Moreover, every subgroup occurring in an irredundant cover of an arbitrary group by $n$ cosets has index at most $2^{n-1}$. This improves a lemma of Neumann.
Read the full abstract

Algebraic Geometry

Automorphisms of toric varieties and Gale duality

We classify complete toric threefolds $X$ such that the automorphism group $\text{Aut}(X)$ acts on $X$ with an open orbit whose complement does not contain a divisor. The latter condition means that for any ray $\rho$ of the fan $\Sigma_X$ there is a Demazure root of $\Sigma_X$ associated with $\rho$. We also find among these varieties those $X$ for which the group $\text{Aut}(X)$ is transitive on the smooth locus $X^{\text{reg}}$. The classifications are based on Gale-dual interpretations of these properties.
Read the full abstract

Pattern Formation and Solitons

Reduction Based Dynamical Systems Analysis of Nonlinear Wave Equations: A Review

Nonlinear wave equations exhibit rich interplay among dispersion, nonlinearity, coupling, dissipation, and external forcing, producing diverse coherent and complex wave structures. Although exact travelling-wave solutions provide analytical benchmarks, their construction alone does not reveal underlying phase-space, bifurcations, stability, or responses to perturbations. This review presents a reduction-based methodological framework connecting nonlinear partial differential equations(PDEs) to reduced dynamical systems, invariant phase-space structures, exact waveform reconstruction, stability analysis, and verification in PDEs. Its applicability and limitations are examined across Schrödinger-type, coupled, nonparaxial, dissipative, magnetic, shallow-water, and fractional wave equations. Particular emphasis is placed on correspondence between analytical waveforms and invariant orbits. Equilibria, periodic trajectories, homoclinic orbits, and heteroclinic connections geometrically represent constant, periodic, localized, and front-like structures, respectively. This review distinguishes existence and stability of invariant structures from onset of chaos, emphasizing complementary diagnostics rather than reliance on phase-portraits or finite-time indicators alone. Major methodological gaps include incomplete parameter-space characterization, weak correspondence between reduced models and full-PDE dynamics, ambiguities in generalized and fractional formulations, limited robustness analysis, inadequate numerical reproducibility, and insufficient links to experimental observables. The framework prioritizes physical admissibility, stability, robustness, and predictive relevance over generation of additional formal solutions. It supports reliable nonlinear wave prediction, stability assessment, control, and system design in fluid, optical, plasma, and other nonlinear physical systems.
Read the full abstract

Astrophysics of Galaxies

Active galactic nucleus activity in brightest cluster galaxies at intermediate redshifts in Sunyaev-Zel'dovich–selected clusters

Context. Brightest cluster galaxies (BCGs) host powerful active galactic nuclei (AGNs) and help regulate the heating–cooling balance of the intracluster medium. While radio-mode feedback is well established locally, its prevalence at intermediate redshifts remains poorly constrained. Aims. We investigate the incidence, accretion state, and host-galaxy properties of AGNs in BCGs in Sunyaev–Zel'dovich (SZ)-selected clusters at $0.3 < z < 0.8$, testing dependences on cosmic epoch, cluster mass, and dynamical state. Methods. We analyse 171 BCGs from the Atacama Cosmology Telescope SZ catalogue, combining Southern African Large Telescope spectroscopy, WISE mid-infrared photometry, and Rapid ASKAP Continuum Survey radio data with X-CIGALE modelling to derive stellar masses, star formation rates, and Eddington-scaled accretion rates. Results. Thirty-eight BCGs (22%) host radio-loud AGNs. Only two ($\sim$5%) are high-excitation radio galaxies; the remainder are low-excitation systems. Most AGNs have low Eddington ratios ($-3 \lesssim \log_{10}\lambda_{\mathrm{Edd}} \lesssim -1$), consistent with radiatively inefficient accretion. Accretion efficiency increases with redshift, while correlations with cluster mass and dynamical state are weak. Radio luminosity is independent of cluster mass. Compared to field radio AGNs, BCG radio-loud AGNs have similar radio powers but are hosted by galaxies $\sim0.8$ dex more massive. Conclusions. BCG radio-loud AGNs at $0.3 < z < 0.8$ are predominantly low-excitation, radiatively inefficient systems, indicating that maintenance-mode accretion was widespread in massive clusters ($M_{500c} \simeq 3.2 \times 10^{14},M_\odot$) by $z \sim 1$. Accretion appears regulated primarily by fuel availability and cosmic evolution rather than global cluster properties, while the cluster environment mainly sets the host-galaxy mass scale.
Read the full abstract

All authors N. Hatamkhani, R.E. Skelton, J. Delhaize, M. Hilton, S.I. Loubser, N. Gupta, K. M. Mogotsi

Computation and Language

The Audit Decides the Verdict: Instrument Effects Rival Demographic Bias in LLM Decision Audits

Whether a language model looks demographically biased can depend on how the audit asks its question. A charitable-aid benchmark reports that the same models favor minority applicants when rating requests one at a time and penalize some when ranking side by side. We test whether that reversal generalizes to hiring, lending, and medical triage: 40,726 requests to five models, applications differing only in the applicant's name, and a primary test fixed before collection. It does not. None of 36 planned contrasts survives correction. The rating advantage keeps its sign at roughly half the published size, and a precision extension bounds any hiring ranking penalty below the published effect, though the lending and triage ranking floors sit above that margin, so the exclusion is conclusive for hiring ranking and for rating in all three domains only. Planted disparities tracking their injected sizes and a directional replication on the original aid materials bound these nulls. The audit is livelier than the demographics: models recognize transparent audits nearly always, tie every identical-content comparison whether the varying detail is race or a hobby, and reward first-listed candidates as much as any demographic effect we measure. Audit verdicts reflect audit construction more than demographic bias.
Read the full abstract

High Energy Physics (Phenomenology)

Bottomonium suppression with a machine-learning-informed Debye mass

Motivated by recent progress in data-driven approaches, we introduce a machine-learning (ML)-informed Debye mass, extracted from lattice-informed inputs, exclusively in the complex-valued heavy-quark Kent State University (KSU) potential. The resulting complex potential is used to solve the real-time Schrödinger equation within the quantum trajectories (QTraj) framework for the evolution of bottomonium in the quark-gluon plasma. We then compute the nuclear modification factors and double ratios for bottomonium $\Upsilon(1S)$, $\Upsilon(2S)$, and $\Upsilon(3S)$ states in Pb-Pb collisions at $\sqrt{s_{NN}} = 5.02$ TeV. We compare our ML-induced results with those from the original KSU model and with experimental measurements from ALICE, ATLAS, and CMS collaborations. We find that the machine-learned Debye mass leads to improved agreement with data, particularly for excited states, highlighting the utility of machine learning in modeling in-medium QCD effects.
Read the full abstract

All authors Ajaharul Islam, Shibo Chen, Fu-Peng Li, Long-Gang Pang, Guang-You Qin

General Economics

Going Beyond the In-/Out-Group Dichotomy: Investigating Altruism towards Middle-Groups

In-group favouritism and out-group hostility are well-documented, but real-world group settings rarely fit a simple dichotomy. Often, a "middle-group" shares some identity markers with the in-group without fully belonging to it. How do people treat such intermediate groups? We address this question using a formal identity marker framework and a multi-lab online experiment (N = 376) with a private allocation task immune to reputation effects and demand characteristics. We find that a middle-group can be treated neutrally, i.e., distinct from both in-group favouritism and out-group hostility, but only under specific conditions. When the middle-group shares two identity markers with the in-group, including university affiliation, a complete hierarchy emerges: the in-group is favoured, the middle-group is treated according to the allocation rule, and the out-group bears the loss. However, when the middle-group shares only one marker, it is treated indistinguishably from the out-group, regardless of which marker is shared. Furthermore, differentiated behaviour evolves only after participants have re-encountered the experimental instructions and group constellation in a second phase, not gradually across rounds. These findings demonstrate that participants can perceive beyond binary in-/out-group categorisation, but such perception requires both sufficient identity overlap and repeated exposure to the group setting. Our identity marker framework provides a tool for systematically studying group relations in more complex settings.
Read the full abstract

General Physics

Sector-Resolved Bayesian Model Averaging for DESI-Era Cosmology

We present a quotient-space Bayesian formulation for DESI-era anomaly interpretation. Given a pattern-labeled catalog with map \(i\mapsto \Act(i)\), the induced posterior \(p(\Act\mid D)\), sector inclusion probabilities \(P_\alpha\), co-activation probabilities \(P_{\alpha\beta}\), and grouped Bayes factors \(B_\alpha(D)\) are exact summaries over predeclared physical activation events. Pairwise comparisons such as \(\lcdm\) versus \(\wacdm\) remain ordinary Bayes-factor tests between specified families; the quotient construction addresses the coarser question of which physical sector carries posterior support when different sectors are represented by unequal numbers of catalog elements. We derive a sector-resolved DESI-CMB-SN likelihood specification for late-time background, early-time ruler, supernova calibration, perturbation, and gravitational-wave propagation sectors. The construction includes an Alcock-Paczynski/isotropic-scale BAO decomposition, a pure-ruler projection, analytic marginalization of low-rank supernova calibration modes, Fisher-normalized sector priors, inactive-sector leakage tests, log-evidence uncertainty propagation, and prior/sector-partition diagnostics. The result is a quantitative procedure for reporting model-comparison support at the level of physically interpretable sectors.
Read the full abstract

Earth and Planetary Astrophysics

The Luminosity Function of Ultra-Faint Trans-Neptunian Objects Detected by JWST

We present a definitive discovery of 27 trans-Neptunian objects (TNOs) using the Near-Infrared Camera (NIRCam) aboard the James Webb Space Telescope (JWST). By employing a shift-and-stack technique and a machine learning network geared specifically to identifying false-positive detections in JWST images produced through the shift-and-stack process, we achieved a 40% detection threshold of $m_{F150W2}=28.8$ mag (corresponding to $m_r\sim29.8$ mag) across a sky area of $0.05 \ \text{deg}^2$. This marks the deepest Solar System survey to date, reaching magnitudes that allow us to explore never-before-seen regions of the TNO size distribution. Our faintest detection has $m_{F150W2}=29.3$ mag and diameter of $\sim10$ km (assuming 15% albedo). Within our sample, we find that both the Cold and Hot TNO subpopulations exhibit a power-law slope. The distribution of apparent magnitudes of our nominal sample (detections at all epochs) are well fit by a single power law $dN/dm \propto 10^{\alpha m}$ with $\alpha=0.29^{+0.08}_{-0.07}$. The dynamically hot and cold subsamples in our discovery set are consistent with the same power law, suggesting that the planetesimal formation process yields similar slopes despite the differing disk conditions at the presumed $\sim25$ and $\sim45$~au formation regions of the two populations.
Read the full abstract

All authors Marielle R. Eduardo, Anastasia N. Morgan, Wesley C. Fraser, David E. Trilling, Gary M. Bernstein, John A. Stansberry, Bryan Hilbert, Matthew J. Holman, William M. Grundy, Thomas L. Storer, Kevin J. Napier

Analysis of PDEs

Speed and stability of segregated waves in a pressure-based model of heterogeneous cell populations

We consider a minimal pressure-based model of heterogeneous cell populations consisting of proliferative and non-proliferative cells with different mobilities. The model is formulated as a system of reaction–cross–diffusion equations describing the spatio-temporal dynamics of the cell densities. The model is known to admit one-dimensional travelling wave solutions with strictly segregated components: non-proliferative cells occupy a finite region at the leading edge, while proliferative cells remain at the rear. However, the speed, parameter dependence, and stability of these waves remain poorly understood. In this work, we derive an almost explicit variational bound on the wave speed by reformulating the problem as a free-boundary problem for a generalised porous–Fisher equation. The estimates we obtain apply to general pressure laws and growth kinetics, agree closely with the results of numerical simulations, and become sharp in the incompressible limit, where we formally recover a fully explicit characterisation of the wave speed. We then analyse the stability of the waves to show that segregated waves are stable only when non-proliferative cells are more mobile than proliferative cells. Finally, motivated by numerical observations of finger-like protrusions, we investigate the stability of incompressible segregated circular waves through asymptotic shape-perturbation analysis. This yields explicit expressions for the pressure, interface velocity, and growth rates of angular modes, thereby making evident the destabilisation mechanisms that may lead to the emergence of fingering instability. Interestingly, we find that, in contrast with the one-dimensional case, the stability of such circular waves is not determined solely by the relative value of the mobility coefficients, and thus instabilities may arise irrespective of which cell type has the larger mobility.
Read the full abstract

All authors Carles Falcó, Rebecca M. Crossley, Martina Conte, Tommaso Lorenzi

Combinatorics

Rook characters as symmetric functions

We give several characterizations of an inhomogeneous basis of the ring of symmetric functions whose evaluations are the character values of the irreducible representations of the rook monoid (symmetric inverse semigroup). Using Schur–Weyl duality, we show that the transition coefficients of this basis with the power symmetric basis are the characters of the propagating partition algebra (dual symmetric inverse monoid algebra). In addition, the structure coefficients of this basis are equal to the coefficients in the smash (or Heisenberg) product of Schur functions.
Read the full abstract

High Energy Physics (Theory)

M5 Instantons in ABJM

We study the existence of supersymmetric M5-brane instantons in asymptotically AdS$_4 \times S^7/\mathbf{Z}_k$ geometries of M-theory, dual to the ABJM theory. The supersymmetric embeddings of these instantons are determined by the M5-brane $\kappa$-symmetry, which forces them to localise to the equivariant fixed points in the asymptotically AdS$_4$ geometry, and wrap $(S^3\times S^3)/\mathbf{Z}_k $. At the quantum level we find that the one-loop partition function of these instantons vanishes, suggesting that the bulk partition function is instead to be expanded in terms of supergravity modes and M2-brane instantons, in line with known results for the $S^3$ partition function of the dual ABJM theory.
Read the full abstract