Astrophysics of Galaxies
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
[Abridged] Investigating substellar populations in young massive clusters offers crucial insights into the formation of brown dwarfs (BDs) and the role of environmental conditions in shaping their properties. Westerlund 2 (Wd2), as one of the nearest dense and massive clusters in the Milky Way, represents an ideal laboratory for studying the effects of high stellar density and ionizing radiation from massive stars on BD formation. This paper presents deep JWST/NIRCam observations of Wd2 in a large number of filters between 1.15 and 4.1 $\mu m$. Our analysis focuses on identifying and characterizing the BD population within the cluster. We carried out PSF photometry on deep JWST/NIRCam data obtained in 4 wide and 6 medium-band filters, using DOLPHOT. The resulting catalog was used to identify BD candidates in Wd2 through spectral energy distribution (SED) fitting with atmospheric models. The 50% detection limit of our NIRCam catalog is $\sim$0.015-0.02 $M_\odot$, at the distance, age, and extinction of Wd2, providing the deepest view of a supermassive star cluster to date. We identify 353 substellar candidates. Most candidates (301) lie above the 10 Myr isochrone in the Hertzsprung–Russell diagram consistent with cluster membership, which are defined as strong candidates. Comparison with a control field indicates a contamination level of $\lesssim$5% for the strong candidate sample down to masses of $\sim$0.01–0.015~$M_\odot$. We identify 73 candidates exhibiting infrared excess indicative of circumstellar disks. These objects occupy the expected infrared-excess locus in de-reddened color diagrams and represent $27.7^{+3.4}_{-3.2}$% of the candidates detected in F410M. The resulting catalog provides a well-characterized sample for future spectroscopic confirmation and subsequent studies of the substellar population of Wd2.
Materials Science
Morphology and Dynamics of Self-interstitial Clusters in Irradiated Nickel
Self-interstitial atom (SIA) clustering is a key early step in radiation damage evolution in face-centered cubic (FCC) metals, governing defect transport, recombination, and the long-term microstructural response of irradiated alloys. We combine molecular dynamics (MD) simulations and high-speed (>1000 frames/s) in situ transmission electron microscopy (TEM) to investigate the structure, energetics, and migration dynamics of SIA clusters in FCC Ni. MD simulations show that interstitials initially form disordered dumbbell clusters that evolve into either sessile Frank loops or glissile perfect (prismatic) loops; the latter progressively reorganize into compact ordered configurations with increasing mobility. Direct construction of both loop types over a wide size range, validated against MD-relaxed structures, shows that perfect loops are thermodynamically favored over Frank loops for cluster sizes N greater than or equal to 14, where N is the number of SIAs, with the energetic advantage increasing with cluster size. Nevertheless, substantial kinetic barriers allow Frank loops to persist as metastable defects. For perfect loops, diffusion coefficients computed over N = 16 to 400 reveal a nearly size-independent migration barrier of approximately 0.02 eV, while the diffusion prefactor decreases approximately as N^(-0.54). Trajectory analysis reveals a non-rigid, row-wise relay mechanism in which the number of participating atoms increases systematically with loop size, accounting for much of the observed prefactor scaling. Sub-millisecond in situ TEM observations reveal intermittent loop motion at velocities higher than previously observed but still several orders of magnitude below the intrinsic mobilities predicted by MD, indicating migration through a heterogeneous energy landscape of mobile and pinned states.
Read the full abstract
Computation and Language
GAPS: Dimension-Level Gates for Conditional Activation Steering
Activation steering suppresses undesired behaviors in language models by adding a steering vector to the hidden state during generation. Recent conditional methods such as CAST and DSAS improve the behavior-capability trade-off by deciding when to intervene, but once active, they apply the full dense vector to all hidden dimensions, regardless of whether a neuron carries concept information or already lies in the desired regime. We introduce dimension-level conditioning as a complementary axis of selectivity that also decides which neurons to intervene on. Our method, GAPS (Gated Activation steering via Posterior and Separability), combines two training-free gates: a static separability gate that restricts steering to neurons with statistically reliable concept information (via AUROC), and a dynamic posterior gate that steers a neuron only when its current activation is better explained by the undesired concept under a Gaussian model. The gates add O(D) overhead per token, and they plug into existing conditional methods. On toxicity mitigation (RealToxicityPrompts) and concept removal (OneSeC) with Gemma-3 (4B) and Qwen-3 (1.7B), GAPS consistently matches or improves the Pareto front of its token-level counterparts; under a fixed capability budget, DSAS+GAPS reduces Gemma-3's toxicity rate from 6.52% to 0.48%, versus 3.52% for DSAS alone. Ablations attribute most of the gain to the posterior gate.
Read the full abstract
Networking
Will there be a 7G?
The transition from 5G to 6G is becoming concrete: the ITU-R IMT-2030 framework has established the high-level vision and capability set for 6G, while 3GPP Release 21 has defined the path toward the first 6G specifications. This raises a deliberately provocative question for the research and standards communities: will there be a 7G, and if so, what would justify it? This paper argues that 7G should not be treated as an inevitable numbering exercise or as a catalogue of more ambitious radio targets. Instead, its justification should depend on whether post-6G systems introduce needs or coordination problems that cannot be met by 6G/6G-Advanced, Wi-Fi, NTN, private cellular, neutral-host deployments, edge-cloud platforms, or complementary wireless and software-based systems. To support this assessment, the paper develops a readiness framework covering demand-led need, system-level discontinuity, coordination value, sustainability and circularity, trust, and geopolitical viability. It then applies the framework to candidate 7G discontinuities, including agentic network operation, RF-native computing, quantum-enabled interworking, policy-aware spectrum governance, grid-interactive infrastructure, outcome-assured services, and regionalized standards. The contribution is not a prediction of a fixed 7G architecture, but a structured basis for deciding whether 7G should become a distinct mobile generation, an extension of 6G evolution, or a broader post-6G infrastructure fabric.
Read the full abstract
Computer Vision
RAFT-DVC: Resolution-Aware Machine Learning-Based Digital Volume Correlation
Digital volume correlation (DVC) provides three-dimensional full-field displacement measurements from volumetric images, but how the internal resolution of a machine-learning-based DVC model affects accuracy and operating range remains poorly understood. Here, we present RAFT-DVC, a resolution-aware family of recurrent all-pairs field transforms (RAFT)-based DVC solvers with encoder downsampling factors s = 2, 4, and 8. Using a matched design, we find that the three solvers localize displacement to approximately 0.017 feature-grid voxel, giving an empirical raw-volume error scaling of approximately 0.017s voxel. The solvers exhibit complementary operating regimes governed jointly by displacement reach and volumetric-texture compatibility. Synthetic benchmarks show that RAFT-DVC achieves errors of the same order as tuned classical DVC under fine-texture, small-to-moderate-displacement conditions and becomes competitive or advantageous under coarse-texture, large-displacement conditions. Frequency-swept tests quantify deformation spatial resolution, while tiled inference enables dense estimation on large volumes. Evaluation on confocal volumetric images acquired during indentation illustrates the importance of matching solver operating regime to deformation magnitude and image texture. Tests on micro-CT images of elastomeric foam, despite training only on particle-labeled synthetic data, provide evidence of cross-texture transfer. We also identify coordinate-order inconsistencies in three-dimensional RAFT correlation sampling and introduce a non-cubic impulse test to verify sampler geometry independently of network training. Correcting the sampler improves native-input accuracy and generalization to unseen volume dimensions. Together, these results establish RAFT-DVC as a fast, resolution-aware framework for dense DVC with characterized accuracy and operating regimes.
Read the full abstract
Combinatorics
A stronger upper bound on the D-chromatic index
For a graph $G$, a proper edge coloring of $G$ is called a D-coloring if every diamond subgraph of $G$ is rainbow. Let $\chi'_D(G)$ be the D-chromatic index of $G$, which is the smallest integer $k$ such that $G$ admits a D-coloring with $k$ colors. Let $\Delta$ be the maximum degree of $G$. The only known Brooks-type upper bound on $\chi'_D(G)$ is $\frac{9}{16}\Delta^2 + \frac{1}{2}\Delta$, given by a greedy coloring. In this paper, using a probabilistic method, we obtain the first improvement upon this upper bound by proving that $\chi'_D(G) \le (1-c)\frac{9}{16}\Delta^2$ for some $c > 0$ and sufficiently large $\Delta$.
Read the full abstract
Combinatorics
Permutation Wordle
We introduce a guessing game, “Permutation Wordle,” in which a guesser attempts to recover a setter's hidden permutation of the set $\{1, \ldots, n\}$. In each round, the guesser submits a word over the alphabet $\{1, \ldots, n\}$, and, as in the game Wordle, learns which entries are correct. We describe a natural strategy and prove that it is optimal in a strong sense: for every $r$, it solves at least as many secrets within $r$ rounds as any possible strategy. The number of permutations it solves in exactly $k+1$ rounds is the Eulerian number $A(n,k)$.
Read the full abstract
Artificial Intelligence
Epistemic Sybil Resistance: Multiplying AI Agents Without Multiplying Evidence
Multi-agent AI systems improve inference by spawning agents and synthesizing reports. But another agent is not another observation: apparently independent reports may descend from the same evidence, and genuinely independent evidence can produce nearly identical reports. We formalize this as an epistemic Sybil problem. A report Z is an epistemic Sybil extension relative to reports R when I(Theta; Z | R) = 0. No report-only aggregator can generally distinguish replication from independent corroboration: identical reports can warrant different posteriors under unobserved ancestry. A Gaussian shared-root model shows common ancestry does not imply complete redundancy. Repeated extraction adds information toward a source-level ceiling, and correlated extraction errors, which a shared base model can induce among independent agents, lower that ceiling further. We test these predictions with more than 20,000 controlled LLM-agent report and extraction calls on synthetic evidentiary documents. Holding one evidence root fixed while report multiplicity rises from 1 to 32 collapses naive posterior coverage from 0.940 to 0.263. Holding report count fixed while evidence-root multiplicity rises from 1 to 16 closes the gap, and the aggregators are statistically indistinguishable at k = 16. The agent's replicate extraction errors are correlated (gamma_cal = 0.719, estimated out of sample), and a correlated-extraction aggregator restores calibration accordingly. A controlled manipulation isolates representation similarity from evidential ancestry. It changes a report-space deduplication mechanism's mean inferred cluster count by 1.425 (95% CI [1.363, 1.485]), whereas a fourfold change in true ancestry changes it by only 0.040 ([-0.045, 0.120]). Collective inference should therefore track evidential ancestry and dependence, not agent or report multiplicity or similarity.
Read the full abstract
Materials Science
Molecular spin qubits in a van der Waals bottle
Advancing quantum information technologies requires qubits whose coherence can be precisely engineered. Among the qubit platforms in development, molecular spin qubits (MSQs) stand out for their atomic scale tunability and chemical specificity, making them powerful candidates for sensing, simulation, and information processing. However, integrating MSQs into solid-state architectures without degrading their coherence remains a central challenge. Here, we introduce van der Waals (vdW) confinement within two-dimensional materials as a strategy for stabilizing quantum states in MSQs by engineering their local electronic, vibrational, and symmetry environments. Using cobaltocene as a model system, we show that confinement within vdW SnS2 and CdPS3 single crystals reorganizes the single-ion energy landscape and slows spin-lattice relaxation by over two orders of magnitude relative to unconfined cobaltocene. The confined MSQs adopt deterministic orientations and self-assemble into ordered, atomically precise superlattices, establishing vdW confinement as a pathway for integrating MSQs into functional quantum devices.
Read the full abstract
Chemical Physics
Latent unified smooth Hamiltonians for excited state chemistry
We describe a neural network architecture and training procedure designed to model electronic ground and excited states of arbitrary molecular systems. By indirectly learning a latent, implicit basis representation of the electronic-state Hamiltonian, the model offers a unified treatment of multiple electronic states, conical intersections, and non-adiabatic couplings. The formalism can be further extended to learn consistent latent representations of additional operators such as transition dipole moments, for example. To demonstrate the general capabilities of our architecture, we train and evaluate networks on two realistic photochemical systems, thymine and azobenzene. The resulting models accurately reproduce energies and oscillator strengths for the ground- and low-lying excited states relevant to the photochemistry of these systems. We highlight the performance of the trained networks by studying critical molecular geometries, including conical intersections and excited state minima. By construction, the proposed framework also recovers the emergence of Berry phase accumulation around conical intersections. By pairing key mathematical structure from quantum chemistry with the representation learning power of transformers, the presented architecture offers a qualitatively new path toward fast and accurate ground- and excited-state simulations.
Read the full abstract
Multiagent Systems
ArcticSwarm: Deferring Early Consensus in Long-Horizon Multi-Agent Research
Multi-agent systems have shown strong performance in domains with reliable verifiers such as coding, where a verifier can select among candidates generated in parallel. Long-horizon web search has no such verifier. Majority voting over independent runs cannot add an answer that none of the runs reaches. Agents that instead read each other's findings adopt an early candidate before alternatives are tested, a failure we call early consensus. We present ArcticSwarm, a multi-agent research architecture that separates evidence gathering from evidence integration. A gathering gate lets selected search tasks post findings to a shared board without reading it. An integration gate checks a candidate at three steps, and a failed check reopens search instead of only rejecting the candidate. On all 830 BrowseComp-Plus questions with the open-weight Qwen 3.5-27B, ArcticSwarm reaches 82.6%. MiroFlow reaches 70.6% under the same model, tools, and limits, and majority voting reaches 62.5% at the same token budget. Without the gathering gate, teams keep a wrong first candidate in 32.2% of runs instead of 22.3%, and accuracy drops by 3.8 points. Removing the integration gate lowers accuracy by 5.9 points. With GPT-5 on live-web BrowseComp, ArcticSwarm reaches 73.6%, above the published scores of GPT-5 (54.9%) and MiroFlow (63.4%). These results show that multi-agent search without a verifier benefits from deferring consensus until candidates are checked, instead of voting after search ends.
Read the full abstract
Category Theory
(2-dep,$\Sigma$)-categories are not generalised categories with families
The notion of a generalised category with families, introduced by Coraglia and Emmenegger is one of the most general notions introduced to capture categorically the notion of dependent typing. It is shown that these generalised categories with families are biequivalent to comprehension categories. We will show that the notion of a (2-dep,$\Sigma$)-category, an extension of the notion of a (dep,$\Sigma$)-category, introduced by Petrakis, is not biequivalent to generalised categories with families, but instead is equivalent to a direct generalisation of that notion.
Read the full abstract
Group Theory
On the Self-Similarity of Permutational Wreath Products and Their Embedding into Finitely Presented Simple Groups
In this work, we study the self-similarity of permutational wreath products of the form \(A \wr_X G\), where \(A\) is a finitely generated abelian group and \(G\) is a self-similar group (the permutational wreath product \(A \wr_X G\) is also known as a lamplighter group). In the case where \(G\) is a non-torsion contracting group, we prove that, under certain conditions, the Scott–Röver–Nekrashevych group \(V_m(\mathbb{Z}^d \wr_X G)\) is finitely presented and virtually simple. Moreover, we prove that \(\mathbb{Z}^d \wr_X G\) embeds into a finitely presented simple group. Furthermore, this provides a new family of groups that satisfy the Boone–Higman conjecture.
Read the full abstract
Computation and Language
Thinking effort aligns between humans and reasoning models in abductive reasoning
A major question in cognitive modeling concerns the behavioral alignment between large language models and humans across linguistic and non-linguistic tasks. Unlike standard LLMs, large reasoning models (LRMs) are optimized with reinforcement learning from verifiable rewards, encouraging correct solutions to reasoning tasks rather than preference-aligned responses. Recent work (de Varda et al., 2025) investigates the cost of thinking in humans and LRMs by comparing human reaction times with model reasoning traces across a range of reasoning tasks. We isolate this alignment by turning to abductive reasoning: unlike deductive tasks, its difficulty cannot be inferred from formal structure and offers no shortcuts a model could exploit to mimic effort without genuine search, providing firmer ground for empirical claims of shared effort. We find further evidence of alignment between LRM and human reasoning effort, as well as evidence that models and humans tend to make similar errors. Finally, we show that decoding methods that let models explore multiple reasoning paths increase alignment in reasoning cost between humans and LRMs across the three models tested.
Read the full abstract
Combinatorics
The Multiorbital Bivariate Chromatic Polynomial: A Subgroup-Lattice Refinement
We introduce the multiorbital bivariate chromatic polynomial F_\Gamma(G;x,y) = \sum_{H\leq G}\frac{1}{|H|}\sum_{h\in H}P_{\Gamma/h}(x,y), which aggregates the orbital bivariate chromatic polynomials associated with all subgroups of a finite group acting on a graph. We derive the equivalent element-wise representation F_\Gamma(G;x,y) = \sum_{g\in G}c_G(g)P_{\Gamma/g}(x,y), where c_G(g) = \sum_{\langle g\rangle\leq H\leq G}\frac{1}{|H|}. The coefficient function depends only on the cyclic subgroup generated by the element and is constant on conjugacy classes. This yields decompositions by cyclic subgroups and conjugacy classes and an interpretation in terms of the incidence algebra of the subgroup lattice. After normalization, the coefficients define a probability distribution obtained by choosing a subgroup uniformly and then an element uniformly from that subgroup. We also establish diagonal multiplicativity for disjoint unions and a weighted cycle-index expression for edgeless graphs. A further contribution concerns the distinguishing power of the orbital bivariate chromatic polynomial. We answer a question of Dohmen and Lange-Geisler affirmatively by exhibiting two non-isomorphic graphs, P_3 and K_2 \mathbin{\dot\cup} K_1 under C_2-actions, with identical orbital bivariate chromatic polynomials. The two actions nevertheless have different multiorbital bivariate chromatic polynomials. Thus the multiorbital polynomial is not determined by the orbital bivariate polynomial, whereas the converse question remains open.
Read the full abstract
Software Engineering
ExecRetrieval: Measuring the Functional-Correctness Gap in Code-Embedding Retrieval
Embedding-based code retrieval is a core component of coding agents and retrieval-augmented code generation, where retrieving correct code matters more than retrieving lexically similar code. Existing code-retrieval benchmarks do not plant controlled, execution-verified single-edit variants of each query's canonical implementation in the search pool, leaving the question of whether embeddings can functionally discriminate correct from near-clone-but-incorrect code unanswered in a retrieval setting. Resolving this requires a benchmark whose search pool itself contains the relevant counterfactuals – execution-verified buggy variants near-identical to each canonical – so that a retriever's rank ordering can be directly tested for functional discrimination rather than topical or identity overlap. We introduce ExecRetrieval, 939 Python tasks each paired with one execution-verified canonical implementation and up to four execution-verified buggy distractors, each generated by a mechanical mutation making a single targeted edit, and evaluate 23 dense embedding configurations plus BM25 under provider-native invocation with paired McNemar tests and query-level bootstrap intervals. With near-clone counterfactuals in the pool, the top hosted system reaches exec@10 = 1.00 but only exec@1 = 0.331; rank-1 misses are paired buggy variants 91.5-99.4% of the time across the four leading systems, and the canonical scores below at least one of its four paired distractors in 67-78% of queries on the leading systems. The full dataset, execution oracle, embedding matrices, environment snapshot, and pairwise statistical tests are released at the URL in Appendix D.
Read the full abstract
Distributed Computing
CREDIT: Cost-guided Reduction-reuse with Efficient DSMEM Inter-CTA Tiling
NVIDIA distributed shared memory (DSMEM) enables direct shared-memory access within a thread block cluster. However, cluster synchronization, remote access, and resource costs make it difficult to determine when DSMEM improves performance. To fill this gap, we propose CREDIT, a cost-guided framework that identifies DSMEM-profitable workload patterns, predicts their profitability range, and delivers consistent speedups across diverse workloads. CREDIT combines three innovations: (1) a profiling-driven characterization that identifies workload patterns likely to benefit from DSMEM; (2) a transformation that applies DSMEM to reduction-reuse workloads; (3) a cost model based on profiling data, to determine its profitability range. Evaluations on diverse workloads show CREDIT achieves 91.7% prediction accuracy on profitability. CREDIT beats torch.compile, Triton, and optimized non-DSMEM CUDA baselines on all six workloads, with geometric-mean speedups of 1.466x on RTX 5090 and 1.318x on H100. CREDIT's source code is publicly available at https://github.com/zhengxiongli08/CREDIT.
Read the full abstract
Chemical Physics
Dodecahydrogen uranium: an icosahedral f-electron superatom with 26-electron shell structure
A new icosahedral superatom UH$_{12}$ is reported and analyzed using post-Hartree-Fock multiconfigurational methods. A RASSCF/CASPT2/RASSI/SINGLE_ANISO workflow is employed to determine the electronic energy levels and magnetic properties. Superatomic characteristics, such as orbital topologies and shell filling order, are discussed. The high hydrogen-to-uranium ratio and potential for optically stimulated decomposition make UH$_{12}$ worth investigation in the context of hydrogen storage. The electric-dipole transition oscillator strengths of the molecule vanish, while its lowest electronic transitions fall in the THz and infrared ranges. The lowest septet manifold exhibits crystal-field splitting into a 1+3+3 pattern, as expected from icosahedral symmetry. In turn, the interlevel gaps within these crystal-field triplets are 30-150 GHz (1-5 cm$^{-1}$). The molecule is magnetically isotropic at low fields, while its levels exhibit higher-order Zeeman anisotropy that becomes significant around 2 Tesla. These properties suggest potential relevance for quantum computing, quantum communication, and quantum memory, including qubit- and qudit-based architectures.
Read the full abstract
Mathematical Physics
Applying the convolution theorem to the two-dimensional magnetic inverse problem
A collection and clarification is given of the key mathematical details of the method of applying the convolution theorem as part of the solution to the two-dimensional magnetic inverse problem: reconstructing current density from magnetic field data. Particular emphasis is given to obtaining a closed-form solution for the Fourier transform of the convolution kernel that features in the Biot-Savart law. Having done this, the Fourier-transformed current density is obtained, via the convolution theorem, as a function of Fourier-transformed magnetic field data and other variables.
Read the full abstract
Artificial Intelligence
Belief-Calibrated Optimization: An Explicit World Model for Agentic Optimization
The performance of an LLM agent depends on the scaffold around a frozen model. A common way to improve that scaffold is to use a coding agent as an optimizer: it reads current scores and traces and iteratively edits the source, producing a new candidate each round. Each edit is chosen according to a belief about how the environment will respond: what went wrong, and which change should help. That belief is typically implicit. It lives in the coding agent's reasoning on the current call, or remains latent in its parameters, rather than as something written down. Later calls therefore see scores and traces, but they do not use that belief. We introduce Belief-Calibrated Optimization (BCO), a method that writes that belief down as a persistent in-context document and continually revises that document as new candidates are evaluated. The resulting document is a world model: the current account of how the environment responds to edits. Added to an otherwise standard loop, BCO reaches a higher train passrate than a matched control that lacks only the world model, on five benchmarks spanning memory QA, tool-use QA, code-as-action app agents, and terminal agents. The gap remains on every held-out split, which is not used to select the candidate. After a target-model swap, in which the frozen model is replaced and the scaffold is not, the selected BCO scaffold leads on the tasks we test, except where context-window overruns leave it unfinished. An offline ablation then asks whether that gap comes from what the world model says. A fresh predictor given the accumulated document forecasts how the environment will respond more accurately than predictors given either no document or a same-form copy whose content has been falsified. The comparison indicates that the document carries reusable information in its content, not only in its form.
Read the full abstract
Information Theory
Comments on the recent improvements of the MRRW bounds
The asymptotic McEliece–Rodemich–Rumsey–Welch bound (1977) limits the largest attainable rate of binary codes as a function of the relative distance. After a nearly half-century hiatus, this result was recently improved in two concurrent works, by OpenAI and by O. Alrabiah and V. Guruswami. The two arguments look entirely different, a Delsarte certificate on the one hand, a classical-quantum channel and the pretty good measurement on the other, and they yield the same bound. The purpose of this note is to explain why: in both proofs, a subspace is attached to every codeword and moved with it, and the bound counts how many such subspaces fit in the ambient space, exactly in the first case and in the probabilistic sense of typicality in the second. We also present the OpenAI proof in the language and context of coding theory, as an extension of the spectral method in which the single vector attached to a codeword is replaced by a subspace.
Read the full abstract
Nuclear Theory
Unveiling QCD Criticality with Cross-Rapidity Net-Baryon Cumulants
Fluctuations of conserved charges are a primary tool in the search for the QCD critical endpoint, but their beam-energy dependence is complicated by global baryon-number conservation, whose influence changes as the experimental acceptance covers different fractions of the collision system. We propose using correlations of net-baryon fluctuations between two separated rapidity windows to exploit the distinct signatures of conservation and critical dynamics. Global conservation produces a negative cross-window correlation, whereas a common long-wavelength critical fluctuation produces a positive one. We show that, within a canonical independent-source framework, the conservation-induced background and the critical signal enter additively at leading order, allowing the leading conservation term to be estimated and subtracted. The resulting yield-scaled correlator follows the nonmonotonic enhancement of an Ising-mapped equilibrium correlation length along a freeze-out trajectory passing near a hypothetical critical endpoint, while wider rapidity windows reduce the response through thermal smearing. Cross-rapidity cumulants therefore provide a rapidity-differential strategy for reducing the leading conservation background and sharpening fluctuation-based searches for QCD criticality in beam-energy-scan experiments.
Read the full abstract
Populations and Evolution
A greedy nearest-neighbor approach to quantify site revisitation: comparing two sympatric raven species
A central challenge in movement ecology is to describe ecologically meaningful residence sites from raw tracking data due to heterogeneous sampling frequency and uncertain site boundaries. Here, we develop a greedy nearest-neighbor clustering approach with local reassignment and polygon-based site construction that generates timestamped sequences of site visits for each tracked individual, and apply it to nine years of GPS data from two sympatric raven species in the Dead Sea region—the Fan-tailed Raven (Corvus rhipidurus) and the Brown-necked Raven (C. ruficollis). Using the resulting visitation sequences, we quantify recursion patterns using the non-Markovian individual mobility model (IMM), which captures the balance between novel-site discovery $\beta$ and preferential return $\alpha$. The inferred dynamics are consistent with IMM predictions and reveal clear interspecific differences: Brown-necked Ravens continue to discover new sites at a higher rate (lower $\beta$), whereas Fan-tailed Ravens show a steeper concentration of visits among top-ranked sites. Entropy analyses further separate the species, with Brown-necked Ravens exhibiting higher site entropy and higher conditional entropy of site-to-site transitions. Together, these results provide a robust framework for describing interspecific differences in movement strategies that can be used to infer memory from GPS data and suggest distinct space-use strategies in two closely related sympatric species.
Read the full abstract
Dynamical Systems
The Hessian of Planar Central Configurations in Pair Space: Decomposition, Morse Index and Symmetry Reduction
We give a variational derivation of the central configuration equations in pair space, where the relative position vectors between pairs of bodies serve as the primary variables. Vector Lagrange multipliers enforce the linear triangle relations required for these pair vectors to be realizable in the plane. For arbitrary $N$, we show that the constrained Hessian decomposes as $H_{\mathcal C}=L^\Delta+\widetilde L$, where the gap Laplacian $L^\Delta$ is positive semidefinite and the transverse Laplacian $\widetilde L$ is signed and contains all possible negative directions. For non-collinear planar four-body central configurations, the signed part has rank two. This yields a new proof of the known bound that the Morse index is at most two. We then characterize positive definiteness and degeneracy of the Hessian by the eigenvalues of an explicit $2\times2$ matrix or, equivalently, by two generalized eigenvalues of a pencil of symmetric matrices. We illustrate the general criterion using the equal-mass square central configuration. For reflection-symmetric configurations, the generalized eigenvalue problem decomposes into two independent smaller problems. This reduction applies to kite and isosceles trapezoidal configurations. For the trapezoid, however, the reflection acts non-orthogonally on the pair coordinates, giving a different, non-orthogonal reduction. Applying the reflection-symmetry reduction to the rhombus family, we prove that the Hessian is positive definite modulo the rotational zero mode; equivalently, every rhombus central configuration is nondegenerate modulo rotations and has Morse index zero.
Read the full abstract
Cryptography and Security
Adversarial Vulnerabilities of Neural Biomarker Identification Systems
There is growing interest in the proposed use of EEG signals as biometric credentials, but thus far there has been little research on the reliability and security of such biometrics. Prior adversarial tests have focused on deep-learning classifiers and assumed attackers have full access to the classifier model. This has left unexamined other, more popular categories of neural signature methods as well as the more realistic case of an adversary having only black-box access to a classifier. In this paper we develop a collection of adaptive attack algorithms which learn to fool an authentication system via targeted alterations of stolen EEG recordings, without requiring any knowledge of the authentication system itself. Tested on 6 public datasets spanning three recording conditions (reacting to visual stimuli, imagining hand movements, and resting), it reveals that different signaturing approaches vary significantly in their degrees of vulnerability to adversarial attacks. We show that vulnerability to spoofing attack is greatly impacted by the recording conditions, with significant variation depending on task at time of recording. Finally, we provide recommendations for improving neural signature biometrics based on the results of our adversarial testing.
Read the full abstract
Cosmology and Nongalactic Astrophysics
Blast.jl: Differentiable Non-Limber Power Spectra for Joint Clustering, Shear, and CMB lensing Analyses
Upcoming large-scale structure surveys require accurate theoretical predictions for angular power spectra on the largest angular scales, where the commonly used Limber approximation breaks down and multiple relativistic and observational effects become relevant. At the same time, modern cosmological analyses increasingly rely on gradient-based inference techniques, motivating the development of fast, fully differentiable algorithms. In this work, we present a major extension of Blast.jl, a numerical toolkit for the efficient computation of non-Limber angular power spectra. Building on the Chebyshev-polynomial decomposition of the original framework, the updated algorithm incorporates redshift-space distortions, magnification bias, primordial non-Gaussianity, intrinsic alignments, CMB lensing, and the integrated Sachs-Wolfe effect, while remaining fully differentiable through custom automatic differentiation rules. Despite the increased physical complexity, the algorithm retains favorable scaling by precomputing all cosmology-independent quantities. We validate the expanded framework against brute-force integration and established cosmological codes, and provide a detailed analysis of the trade-off between computational speed and precision as a function of the algorithm hyperparameters. We demonstrate the readiness of Blast.jl for gradient-based inference through a simulated likelihood analysis with LSST Y10-like mock data, recovering unbiased cosmological and nuisance parameter constraints from a 35-parameter model using gradient-based samplers.
Read the full abstract
Rings and Algebras
Orthogonal completion of algebraic systems
The paper considers the generalization of the construction of orthogonal completion to all semiprime linear algebras (systems), its implementation in maximal systems of quotients similar to the symmetric Martindale ring of quotients for known examples of their description in terms of algebras of quotients of graded Lie algebras and technical results for applying the orthogonal completion method to them.
Read the full abstract
Mesoscale and Nanoscale Physics
Magnons in Metallic Altermagnetic $\text{KV}_2\text{Se}_2\text{O}$
We investigate the spin excitation in altermagnetic vanadium oxyselenides, focusing on the metallic quasi-two-dimensional compound $\text{KV}_2\text{Se}_2\text{O}$. Using a fermionic Hamiltonian derived from ab initio calculations, we compute the transverse spin susceptibilities within the random phase approximation and extract the magnon dispersion relation. We show that the altermagnetic symmetry leads to characteristic degeneracies and directional splittings in both the electronic bands and the magnon spectra along high-symmetry paths of the Brillouin zone. The metallic $\text{KV}_2\text{Se}_2\text{O}$ exhibits finite magnon linewidths arising from the coupling to the particle-hole continuum. Furthermore, by fixing the magnitude of the wave vector and varying its in-plane direction, we uncover a pronounced angular dependence of the magnon energies and lifetimes, with complementary damping behavior between the two magnon branches. Upon including spin-orbit coupling, the system exhibits an out-of-plane easy c-axis anisotropy, which opens a magnon gap of approximately 6 meV at the $\Gamma$-point, possibly rendering the magnetic order stable at room temperature. Our results demonstrate that altermagnetism controls not only the dispersion but also the anisotropic decay of spin excitations, highlighting altermagnetic metals as promising platforms for directionally selective magnon transport.
Read the full abstract
Artificial Intelligence
The Memory Trust Gap: Reliance on Stale Evidence Across Model Scales in LLM Agents
Persistent-memory agents can follow stale information despite current authoritative evidence. We study stale-evidence reliance across Qwen3 0.6/1.7/4/8B checkpoints using two controlled, action-scored suites: a Benefit suite requiring stored information and a Safety suite providing current evidence as static prompt text. We distinguish stale-value reliance from net accuracy loss relative to no memory. Benefit-suite reliance is 0.92-1.00 across checkpoints. In the Safety suite, larger checkpoints resist a plain stale note but incur substantial harm when it appears newer. A full $2\times2\times2\times2$ factorial manipulates label, recency, source authority, and position. Direct cross-size contrasts identify stronger recency effects in the larger tested checkpoints and a position-effect sign reversal within Qwen3. Metadata improves accuracy at all four checkpoints, with larger gains at 4B/8B; pre-resolution yields near-ceiling accuracy across all four. Llama-Instruct comparisons provide cross-family evidence for selected effects, while RGB and MisBench corroborate stale-value reliance. A framing control shows that the failure extends beyond memory-labeled content. These results characterize cue-dependent vulnerability across tested checkpoints without isolating capability or evaluating a complete memory lifecycle.
Read the full abstract
Combinatorics
Minimum Weakly Saturated Graphs and Bootstrap Percolation in General Host Graphs
A graph $G$ is weakly $H$-saturated if one can obtain $K_n$ by adding one edge to $G$ at a time, where each additional edge creates at least one new copy of $H$. The minimum number of edges needed for a weakly $H$-saturated graph $G$ of order $n$ is known as the weak saturation number of $H$, written $wsat(n,H)$. A graph $G$ is minimum weakly saturated if $wsat(n,G)=|E(G)|-1$ for some value of $n$. We explore classes of minimum weakly saturated graphs and their connection to the $H$-bootstrap percolation process, as well as weak saturation in a more general setting than the complete graph.