High Energy Physics (Phenomenology)
Two-loop $gg\to AA$ amplitudes near threshold and the double parton scattering singularity
Arunima Bhattacharya
The two-loop amplitudes for pseudoscalar pair production in gluon fusion, $gg\to AA$, with each pseudoscalar emitted from its own effective gluon vertex, are known in closed form, but large cancellations near threshold obscure their analytic structure. We expand the IR-finite amplitudes in the pseudoscalar velocity $\beta_A$ through $\beta_A^6$ at fixed scattering angle. All inverse powers of $\beta_A$ and the odd powers through $\beta_A^5$ cancel, and at leading power the angular dependence enters only through $\ln(2p_T/m_A)$. We identify this logarithm with the double parton scattering (DPS) singularity: with the overall constant fixed at threshold, the factorisation formula of Gaunt and Stirling reproduces the velocity dependence of its one-loop coefficient in the forward limit to $3\times10^{-15}$, and a one-loop $gg\to HH$ analysis confirms the corresponding CP-even weight. At two loops the equal-helicity amplitude grows as $\ln^3p_T$, with the leading logarithmic ratio of planar $\mathcal N=4$ super Yang-Mills theory, while the opposite helicity amplitude, which has no one-loop DPS logarithm, grows as $\ln^2p_T$. In the finite virtual correction, relative to the squared tree amplitude, the highest logarithm at order $\alpha_s^2$ is $-\tfrac{7}{64}\alpha_s^2N_c^2\ln^2(2p_T/m_A)$. For $\beta_A\leq0.2$ the series reproduce the unexpanded amplitudes with relative errors of at most $6.77\times10^{-5}$ at the tested points; they provide this two-loop contribution to the hard function for soft-gluon resummation of the pair invariant-mass distribution near $m_{AA}=2m_A$ and are released with code for their numerical evaluation.
Read the full abstract ▾
Operator Algebras
Invariant von Neumann subalgebras of crossed products and tensor products
Tattwamasi Amrutam, Artem Dudko, Yongle Jiang et al. (+1 more)
Let $\Gamma$ be a countable discrete group acting on a tracial von Neumann
algebra $(Q,\tau_Q)$ in a trace-preserving way. We study invariant von Neumann subalgebras of $Q\rtimes\Gamma$ and of $L(\Gamma)\bar\otimes Q$, where the latter is equipped with the diagonal action. We show that, under the pointwise weak mixing assumption, an invariant subalgebra whose image under the canonical conditional expectation onto $Q$ is scalar is necessarily contained in $L(\Gamma)$. Further, under relative weak mixing and central freeness assumptions, we determine the structure of invariant subalgebras of the crossed product with non-trivial intersection with $Q$. For p.m.p. actions of groups with trivial amenable radical and the invariant subalgebras rigidity property, pointwise relative weak mixing gives a complete classification of invariant subalgebras in the tensor product case. We construct essentially free,
prime, pointwise weakly mixing p.m.p. actions of $\mathbb{F}_n$ for every $2\leq n\leq\infty$, as well as non-
prime examples with explicitly described factors.
Read the full abstract ▾
All authors Tattwamasi Amrutam, Artem Dudko, Yongle Jiang, Adam Skalski
Systems and Control
A Performative Control Approach to Day-to-Day Traffic Signal Retiming
Shuchang Jiang, Xinwu Qian, Santiago Segarra et al. (+1 more)
Conventional fixed-time traffic signal optimization and recent day-to-day signal retiming generally model route adaptation under prescribed route-choice behavior and origin-destination (OD) demand, while neglecting the fact that travelers' behavior can be performative to the signal control policy. A signal timing policy learned from historical traffic observations must therefore account for how its own deployment changes the distribution on which it will be evaluated in the future. We formulate this problem as performative signal control under observed traveler responses and characterize performative stability of the coupled policy-demand system. We propose incremental day-to-day retiming using the local policy gradient at the current operating point, estimated offline through simulation-based zeroth-order optimization without an explicit day-to-day demand-response model. This gradual approach provides a feasible path for policy and demand to co-evolve toward stability. Experiments on a downtown Denver network show that our approach improves congestion, throughput, and stability, generalizes across different demand levels, and remains effective when the offline simulator contains mismatched travel demand and routing behaviors.
Read the full abstract ▾
All authors Shuchang Jiang, Xinwu Qian, Santiago Segarra, Zhongyuan Zhao
Computation and Language
Can LLMs Access their World Knowledge for Event Prediction?
Zijun Wu, Zhitian Zhang, Hossein Hajimirsadeghi et al. (+2 more)
Recent work has questioned whether large
language models (LLMs) can be useful for temporal event prediction. Despite vast pretrained world knowledge, LLMs often underperform simpler specialized models. We ask whether this knowledge is truly inaccessible, or whether current methods fail to present event prediction in a form LLMs can use. We argue for the latter and show that LLMs become effective event forecasters when the interface is designed around two choices. The first grounds each event in a natural-
language description, and the second separates event semantics from the prediction space to prevent semantic bias from label associations. To test this, we introduce DescribeEvent, a
benchmark of ~150K events across seven
datasets and six domains with event-level descriptions. We find that capable LLMs substantially benefit from descriptions and symbolic labels, outperforming description-free variants and non-LLM baselines, while weaker models gain little. These results suggest that accessing LLM world knowledge for event prediction is an interface problem, and that event descriptions offer a practical way to elicit the knowledge LLMs already encode.
Read the full abstract ▾
All authors Zijun Wu, Zhitian Zhang, Hossein Hajimirsadeghi, Nikita Dvornik, Alexander Pashevich
Machine Learning (Statistics)
Discrete Diffusion Models are Minimax Optimal Under Sparsity
Luca Blunda, Tyler Farghly, Umut Simsekli et al. (+1 more)
Discrete
diffusion models have emerged as a leading alternative to autoregressive models for generating text and other categorical data, yet their theoretical foundations remain less developed than those of their continuous counterparts. For continuous
diffusion models, a prominent line of theory suggests that these models can adapt to low-dimensional structure in the data, thereby improving statistical efficiency and generalization. In this work, we investigate the discrete analogue of this phenomenon by studying the statistical rates of discrete
diffusion models under sparsity assumptions. By analyzing a nonparametric
estimator derived from a regularized score-
entropy objective, we establish minimax-optimal rates in the sparse regime. The proposed regularization can be viewed as an adaptation of label smoothing techniques introduced for classification, and does not require the knowledge of the sparsity parameter. This shows that discrete
diffusion models can adapt intrinsically to sparsity.
Read the full abstract ▾
All authors Luca Blunda, Tyler Farghly, Umut Simsekli, Alain Durmus
Dynamical Systems
Mixing Rates for Analytic Observables: Toral Endomorphisms and Anosov Blaschke Deformations
Ekaterina Shchetka
We analyze exponential mixing for analytic observables and show that mixing at a double exponential rate occurs for genuinely nonlinear systems. Neither its presence nor its sharp exponent is stable under arbitrarily small area-preserving perturbations in the real-analytic
topology. We construct two families of area-preserving Anosov Blaschke product diffeomorphisms that converge in the real-analytic
topology to the same
hyperbolic toral automorphism. For nonzero parameters, both deformations are topologically conjugate to the linear model but not $C^1$-conjugate. For all nonzero parameter values, the maps in the first family are mixing at a double exponential rate with the same sharp exponent, strictly smaller than that of the linear map. The sharp exponent is therefore not lower semicontinuous at the linear map. In the second family, mixing at a double exponential rate is lost altogether, although the maps remain exponentially mixing. We also establish double exponential bounds for every ergodic toral endomorphism and uniform bounds of the same type for totally ergodic higher-rank toral actions. Our method combines the
Fourier escape argument with Koopman operator estimates from a
Hilbert space of analytic functions to its dual space of hyperfunctions.
Read the full abstract ▾
Solar and Stellar Astrophysics
SPICY 30021: an embedded bona fide FUor with a mid-infrared precursor
Zhen Guo, Calum Morris, Vitor Fermiano et al. (+37 more)
Unstable mass
accretion in young stellar objects (YSOs) often leads to observable outbursts, which play a key role in stellar mass assembly during pre-main-sequence
evolution. The highest-amplitude, longest-duration events are classified as FUors. Until recently, only 50 FUor events had been confirmed. Many occur in embedded protostars, which are difficult to monitor in optical surveys. We present photometric and spectroscopic observations of SPICY 30021, an eruptive young
star discovered in the mid-infrared (MIR) NEOWISE time series. We aim to construct its spectral energy distribution and characterise its eruptive nature. We compiled multi-wavelength photometric time series from NEOWISE, VVV and VVV-RENEW, and obtained photometric and spectroscopic observations during the outburst with SOAR, Magellan and VLT, including data from the ESO public survey VVVX-GalCen. We identified the outbursting source, separated it from nearby companions, and estimated the line-of-sight extinction from spectroscopic features. Stellar parameters were constrained using radial velocity, proper motions and stellar
evolution models. SPICY 30021 is an optically obscured FUor with a total interstellar and circumstellar extinction of Av = 12.5 mag. Its spectrum closely resembles those of classical FUors. Multiple methods yield a distance of 1.7+_0.4 kpc. Its
accretion luminosity is Lacc = 32 L_sol, moderate among FUors. Quiescent near-infrared photometry is consistent with a young 0.9 M_sol progenitor, although the mass depends on the adopted model and distance. We confirm SPICY 30021 as a bona fide FUor event. Its MIR precursor suggests an outside-in triggering mechanism.
Read the full abstract ▾
All authors Zhen Guo, Calum Morris, Vitor Fermiano, Carolina Lizana-Vidal, Aldo Aliaga, Javier Osses, Felipe Navarete, Philip Lucas, Roberto K. Saito, Adolfo Carvalho, Vardan Elbakyan, Carlos Contreras Peña, Koshvendra Singh, Min Fang, Philipp Weber, He Zhao, Tigran Yu. Magakian, Tigran A. Movsessian, Matias Gomez, Dante Minniti, Valentin Ivanov, Maria Gabriela Navarro, Sara Federle, Ilaria Petralia, Alonso Luna, Jesus Corral-Santana, Elisa R. Garro, Sean D. Points, Cesar Briceño, Zenghua Zhang, Radostin Kurtev, Jura Borissova, Jose G. Fernández-Trincado, Javier Alonso-García, Claudio Cáceres, Casmir O. Obasi, Daniela Galdeano, Laura Baravalle, Teresa Giannini, Zhikai Zhu
Mesoscale and Nanoscale Physics
Topological superconductivity in layered d-wave altermagnetic metals
Yijie Mo and Zhongbo Yan
We investigate multilayer d-wave altermagnetic metals with odd-parity helical p-wave pairing. In the minimal bilayer system, interlayer tunneling reconstructs the
spin-split Fermi surfaces and generates two layer-hybridized sectors with distinct effective chemical potentials. Without Rashba
spin-orbit coupling (RSOC), the emergent mirror
symmetry and subsystem chiral
symmetry stabilize phases with different numbers of helical Majorana edge-mode pairs and additional counterpropagating pairs, jointly characterized by the mirror Chern number and the winding numbers defined along high-
symmetry lines. The combination of RSOC and the altermagnetic exchange field gaps out the helical Majorana pairs while, in a certain parameter regime, leaving another pair of counterpropagating Majorana edge modes gapless, giving rise to second-order topological phases with isolated or twofold Majorana corner modes, as well as a hybrid-order phase in which gapless Majorana edge modes coexist with Majorana corner modes. We identify a layer
symmetry that elucidates the physical origin of these phases. For an arbitrary N-layer system, such a
symmetry naturally arises from the interlayer-coupling
matrix, allowing N-fold Majorana corner modes to emerge when all layer-hybridized sectors enter the nontrivial second-order regime. Our results establish the interplay among the layer degree of freedom, odd-parity pairing, and Fermi-surface configurations as a controllable route to engineering topological superconductivity in layered altermagnetic systems.
Read the full abstract ▾
Machine Learning
IR-PRM+ with Previous-Utility Prediction: Constant Regret and Best-Iterate Convergence
Michael Sullins and Ian A. Kash
Regret matching (and its variants) and counterfactual regret minimization are central
algorithms in practice for normal-form and extensive-form game solving. Yet these have historically lacked the stronger theoretical convergence guarantees available to optimistic and extra-gradient methods. Recent work introduced IREG-PRM+, a regret matching variant which does have strong theoretical convergence guarantees. However, it relies on a simultaneous, extra-gradient update. IREG-PRM+ is an instance of the more general IR-PRM framework, where empirically optimistic and alternating updates often have better performance. However, the prior extra-gradient analysis does not extend to these cases. We introduce a new approach that covers both settings. For simultaneous and alternating IR-PRM+ in finite two-player zero-sum
matrix games, previous-utility prediction recovers the constant-regret and convergence-rate guarantees previously obtained with extra-gradient prediction: external regret bounded independently of T, an O(T^{-1}) average duality gap, and an O(T^{-1/2}) best-current-iterate gap. Moreover, the current duality gap vanishes and the current profile converges to the Nash
equilibrium set. The key step bounds the component of consecutive strategy movement not controlled by the existing IR-PRM+ regret inequality using the relative growth of the nondecreasing regret-vector norm rho_t. Coupling this bound across the two players yields recurrences whose effective coefficient scales as 1/(rho_t^X rho_t^Y), preventing unbounded norm growth. We extend the analysis to finitely many simplex learners under a weighted Minty condition and, through the corresponding CFR variant IR-PCFR+, recover analogous guarantees for finite two-player zero-sum perfect-recall extensive-form games under the same full-reach condition as prior work.
Read the full abstract ▾
Machine Learning
Scaling Cheap Synthetic Pretraining Improves Indoor Radio-Map Prediction
Khoren Petrosyan, Artashes Mkrtchyan, Rafayel Mkrtchyan et al. (+2 more)
Real measurements are too limited to train an accurate indoor pathloss predictor that works across a wide range of buildings. Large-scale synthetic pretraining offers a way forward, but real building layouts are also scarce, and accurate ray-tracing simulators are slow, expensive, and often proprietary. We address both problems by introducing a generator that automatically creates diverse indoor layouts and a fast, open-source approximate ray tracer that computes
radio maps for them. Using this pipeline, we generate 128M synthetic samples, pretrain ResNet-101 and Swin-T encoder-decoders, and fine-tune them on high-quality
benchmark data. Despite simulator mismatch, our ResNet-101 model reduces error by 15% on average relative to the best known methods across five indoor
radio-map prediction tasks and improves transfer on a small real-measurement
dataset. Both architectures show predictable data-scaling trends: power laws fitted on 2M-16M samples predict the mean normalized downstream RMSE at 32M-128M with 3.33% mean absolute percentage error for ResNet-101 and 7.10% for Swin-T. These results show that pretraining on large amounts of cheap simulated data can improve indoor
radio-map prediction, even when real measurements and high-quality simulations are limited.
Read the full abstract ▾
All authors Khoren Petrosyan, Artashes Mkrtchyan, Rafayel Mkrtchyan, Hrant Khachatrian, Theofanis P. Raptis
Combinatorics
Internal partitions of $5$-regular and $8$-regular graphs
Siwei Lin and Qinghou Zeng
An internal partition of a graph is a partition of its vertex set into two nonempty parts such that every vertex has at least half of its neighbors in its own part. DeVos asked whether, for every fixed positive integer $r$, all but finitely many $r$-regular graphs admit an internal partition. This was previously known to hold only for $r\in\{3,4,6\}$. In this paper, we confirm this for $r\in\{5,8\}$ by proving that every $r$-regular graph on at least $10^7$ vertices has an internal partition.
Read the full abstract ▾
Machine Learning
Data Density Scaling Laws for Image Self-Distillation
Alvard Barseghyan, Ani Vanyan, Hakob Tamazyan et al. (+1 more)
Current scaling law formulations suggest that increasing the number of unique samples in a
dataset improves model performance, while repeated exposure to the same samples provides diminishing value. In this work, we examine this assumption for image self-
distillation models. Through systematic experiments across diverse
datasets, namely Maxar, SATLAS, Waymo, nuPlan, Ego4D, ImageNet, and YFCC15M, we demonstrate that the effects of sample repetition on downstream performance are non-trivial and
dataset-dependent. We identify settings where repeatedly training on a smaller random subset can outperform training on a larger set of unique samples. This finding highlights a limitation of existing scaling laws. To address this gap, we propose a modified version of a Chinchilla-style scaling law that explicitly accounts for two factors: the number of unique samples and repetition per sample. This formulation provides a framework for understanding data-compute trade-offs and improves prediction in some regimes where repetition plays a central role, although its benefits are not consistent across
datasets. Finally, we examine how subset selection and augmentation affect downstream performance.
Read the full abstract ▾
All authors Alvard Barseghyan, Ani Vanyan, Hakob Tamazyan, Hrant Khachatrian
Algebraic Topology
On the work of Graeme Segal
Daniel S. Freed
This is my written Laudatio for Graeme Segal on receiving the Chern Medal at the 2026 International Congress of Mathematicians. I make some general remarks about Segal's qualities as a mathematician. Then I discuss the origins of his work on classifying spaces of categories and Gamma-spaces. In the last section I describe his axiom system for Wick-rotated
quantum field theory and some of his subsequent work in the area. Of course, this is necessarily only a small slice of Segal's total research output.
Read the full abstract ▾
Machine Learning
PMO-Dock: Benchmarking Docking, Specificity, and Generalization in Molecular Optimization
Gor Simonyan, Tatevik Abrahamyan, Narek Abelyan et al. (+2 more)
The Practical Molecular Optimization (PMO)
benchmark standardized evaluation in molecular optimization, but it is built on simple property-based oracles that do not capture structure-based drug design. As the field has moved to docking-based objectives, evaluation practice has become unstandardized, and different methods report results on different docking tasks, often without a shared protocol for fair comparison or sample-efficiency constraints. To address this, we introduce Practical Molecular Optimization for Docking (PMO-Dock), a
benchmark and protocol for docking-based optimization consisting of 25 tasks covering hit generation, lead optimization, and a new specificity task requiring strong on-target binding while penalizing off-target interactions. The protocol separates model development from final evaluation: validation tasks are used for hyperparameter tuning, and test tasks are held out for final evaluation and are not used during development. Evaluation on the test tasks is done under a strict oracle budget to reflect realistic drug-discovery settings. We
benchmark four diverse high-performing methods spanning different optimization paradigms, Saturn (reinforcement learning), GenMol (discrete
diffusion), Genetic-guided GFlowNet, and ChemLlama (LLM-based evolutionary search). Our results show no universally best method across tasks, additional analyses of hyperparameter transfer (cross-task and cross-category) and sensitivity show that transfer is often limited and varies by method, motivating method-dependent selection strategies. This
benchmark provides a concrete standard for evaluating generalization to unseen tasks and sample efficiency under a limited test-time oracle budget.
Read the full abstract ▾
All authors Gor Simonyan, Tatevik Abrahamyan, Narek Abelyan, Tigran Fahradyan, Hrant Khachatrian
Functional Analysis
A sharp discrete Sobolev inequality and left inverses of higher order differences
Raul Hindov, Shahaf Nitzan, Jan-Fredrik Olsen et al. (+1 more)
For integers $\alpha,d \geqslant1$, we determine the best constants in the inequalities $||(a_j)_{j=1}^{d}||_{\ell^1} \leqslant C_{\alpha,d} ||((\Delta^\alpha a)_j)_{j=1}^{d+\alpha}||_{\ell^2},$ where $\Delta$ is the difference operator. We find that these constants correspond to the $\ell^2$ norm of the discrete Chebyshev
polynomials, and show that the extremal sequences may be obtained by appropriate sampling of the discrete Landau
polynomials. From beginning to end, this project was carried out without use of AI.
Read the full abstract ▾
All authors Raul Hindov, Shahaf Nitzan, Jan-Fredrik Olsen, Eskil Rydhe
Solar and Stellar Astrophysics
First L-type substellar object discovered in the vicinity of the embedded cluster NGC 2023
Styliani Tsilia, Carlos Dominguez-Tagle, Maruša Žerjal et al. (+8 more)
We present the first substellar object in a
star-forming region spectroscopically confirmed with Euclid, namely J054158.70-021031.4. This young and faint source is located close to the embedded open cluster NGC 2023. It has a spectral type of L3, a reddening of Av = 3 mag, proper motion consistent with the nearby cluster NGC 2024, and a magnitude of $\sim24.5$ in the Euclid optical band. At the distance of 403 pc, its inferred mass lies close to the planetary-mass boundary. We discovered the object by searching for ultracool dwarf candidates in the Euclid Early Release Observations. The search was based on photometric criteria, followed by the analysis of slitless 2D spectro-images in order to distinguish features that indicate the nature of these objects. This work demonstrates the potential of Euclid to unveil previously hidden populations of very low-mass objects, located in obscured regions where
star formation is ongoing.
Read the full abstract ▾
All authors Styliani Tsilia, Carlos Dominguez-Tagle, Maruša Žerjal, Eduardo L. Martín, Hervé Bouy, Nafise Sedighi, Sara Muñoz Torres, Nikola Vitas, Khalid Barkaoui, Nicolas Lodieu, Jerry J.-Y. Zhang
Numerical Analysis
Exponential Stability of Slow–Fast Stochastic Systems with a Small Delay
Yudong Wang, Kele Sun, Shijie Zhou et al. (+1 more)
In this paper, we investigate the exponential stability of coupled slow–fast
stochastic systems with a small delay. Under an averaged Lyapunov condition, we establish practical $p$th-moment exponential stability for $0<p<2$ using a cutoff-function approach to handle singular Lyapunov derivatives near the origin. Exponential stability in the $p$th moment and almost surely is further obtained through two distinct mechanisms: homogeneous dissipation of the fast dynamics or vanishing sensitivity of the slow coefficients to the fast state near the origin. We extend these results to locally Lipschitz systems, establishing exponential stability with arbitrarily high probability for bounded sets of initial data, thereby broadening the applicability of stability analysis for
stochastic systems with delayed feedback. Numerical examples illustrate our results.
Read the full abstract ▾
All authors Yudong Wang, Kele Sun, Shijie Zhou, Xuerong Mao
Computer Vision
Action Jets: Learning the Local Shape of the Action-conditioned Data Manifold
Anton Arapin, Chahyon Ku and Bernadette Bucher
World models are a powerful means of
reasoning and planning, but many of these action-conditioned predictive models are high dimensional and computationally expensive. These challenges limit the number of action candidates that can be evaluated by a planner. We show that this cost is avoidable for planners over short horizons, which only require local information with respect to the current state. For a fixed state, the frames reachable by varying a low-dimensional action (a camera movement or an end-effector command) lie on a low-dimensional
manifold. We estimate the local shape of that
manifold directly via a network trained to estimate how each image region moves per unit of each action axis supervised by data collected from samples by the generator. A correction head fit to trajectory data estimates the shape of this
manifold region up to a third-order Taylor
polynomial. We call this network estimating the Taylor
polynomial describing the
manifold shape an Action Jet and compute its radius of validity in closed form. By traversing this approximation of the local
manifold shape, we only have to perform inference once per state for any number of candidate actions. Scoring sixteen candidates from a single extraction is over 300x faster than with either source generator, and this gap widens with the number of candidates a planner evaluates. We measure against state-of-the-art action-conditioned generators, IRASim and WorldGym on BridgeData V2, a real 7-DoF manipulation
benchmark, and FrameCrafter on DL3DV scenes, where we chart from the novel-view generators DepthSplat and LagerNVS. Because only the tangent comes from the generator and every other term from the data, the Action Jet also predicts held-out frames more accurately than IRASim, the generator that supplied its tangent, at short horizons and at over 100x less compute.
Read the full abstract ▾
Mesoscale and Nanoscale Physics
Nanoscale Mapping of Strain-engineered Excitons in Monolayer WSe$_2$
Anna S. Roche, Nicholas V. Proscia, Paul D. Cunningham et al. (+12 more)
Excitons in monolayer transition metal dichalcogenides (TMDs) are highly sensitive to external stimuli, allowing control of resonance energies and lifetimes through electric and
magnetic fields,
heterostructures, moire potentials and mechanical strain. Observations of single-
photon emission (SPE) near sample edges, wrinkles, and strained nanobubbles have motivated strain-engineering aimed at realizing localized emitters in 2D materials. The origin of these emitters remains unclear and prior studies have relied on far-field spectroscopy with strain profiles inferred separately through topography, limiting direct correlation between local strain and excitonic response. We directly probe strain-modified excitonic spectra in monolayer WSe$_2$ using cryogenic scattering-type scanning near-field optical microscopy (s-SNOM). We compare two strain-engineering techniques, nano-indentation and patterned substrates, used to produce SPEs. At 9 K, we spatially resolve
exciton resonance shifts with sub-diffraction limit resolution, simultaneously measuring near-field optical response and topography. We observe that strain transferred from nano-indentation into a deformable
polymer substrate differs from that produced by draping TMDs over patterned dielectric pillars. Adhesion and tearing during plastic deformation drive a compressive
exciton blueshift, while draping produces dominant tensile strain. Comparing the strain model results to the experimentally observed energy shifts shows that the nano-indentation strain is over an order of magnitude smaller than previously inferred, due to puncturing of the fragile monolayer, whereas pillar-strained devices remain intact under more modest strain. These differences suggest SPE arises from complex, strain-mediated interactions rather than simple confinement potentials, establishing s-SNOM as a powerful probe of
quantum emission in strained 2D materials.
Read the full abstract ▾
All authors Anna S. Roche, Nicholas V. Proscia, Paul D. Cunningham, Kathleen M. McCreary, Dante J. O'Hara, Hsun Jen Chuang, Cory Cress, Collin Sanborn, Samuel W. LaGasse, Jose J. Fonseca, Jeremy T. Robinson, Takashi Taniguchi, Kenji Watanabe, John R. Schaibley, Brian J. LeRoy
Cryptography and Security
CyberArena: Benchmarking AI Agents in Live Symmetric Attack-and-Defense Combat
Bangya Liu, Jianhong Tu, Yueyang Pan et al. (+18 more)
AI
agents are demonstrating increasingly strong offensive and defensive cybersecurity capabilities in the real world. Current
benchmarks mostly evaluate
agents in isolation against static targets. Real cyber operations, however, are dynamic and coupled: security postures constantly change as attackers and defenders adapt to each other in real time. Static
benchmarks quickly saturate and miss an
agent's ability to attack and defend simultaneously under live pressure. To address this gap, we present CyberArena, a dynamic, symmetric
adversarial benchmark built around multi-round attack-and-defense combat. In isolated sandboxes, two
agents compete under identical rules, acting concurrently as both attackers and defenders over continuously changing services. This converts static targets into evolving, semi-open challenges, directly testing an
agent's speed and precision in discovering, exploiting, and patching vulnerabilities. To sustain the
benchmark's longevity, an automated agentic pipeline continuously generates, reviews, and validates new environments. Furthermore, because pairwise contests are notoriously hard to quantify, we introduce a dynamical model that maps match dynamics into unified, multidimensional scores across offensive, defensive, and domain-specific capabilities. Evaluating 12 frontier LLM
agents across 40 challenges in 466 matches, CyberArena separates model capabilities where single-
agent baselines hit a ceiling. The trajectories read like a fight rather than an exam: defenders patch one hole and miss the next and certify repairs on stale evidence, while capable
agents think a move ahead, planting backdoors for later rounds and, in one case, writing their own packet sniffer to watch the opponent. Beyond a
benchmark, CyberArena serves as an adaptable, scalable experimental platform for future research on cybersecurity
agents.
Read the full abstract ▾
All authors Bangya Liu, Jianhong Tu, Yueyang Pan, Hanhong Zhao, Zhe Ye, Hongwei Li, Saastha Vasan, Katherine Lambert, Junmin Zhu, Hao Wang, Jienan Chen, Yunze Xiao, Minglai Yang, Zihan Wang, Hairong Yin, Hao Wang, Zhiwen Fan, Tianneng Shi, Zhun Wang, Chenguang Wang, Dawn Song
Mathematical Physics
Anomalies, higher symmetries, and homotopy theory I: crossed precosheaves and a nonabelian Cech construction
Anton Kapustin and Shixiong Xu
Symmetries of a
quantum field theory or a
quantum lattice model are expected to be described not by a group but by a homotopy type, whose homotopy groups are the higher
symmetry groups and whose Postnikov invariants encode 't Hooft anomalies. We propose a homotopy-theoretic definition of 't Hooft anomalies based on the structure of localized
symmetries in the operator-algebraic approach. The input is a crossed precosheaf: a collection of groups attached to regions of space, equipped with pairings recording that the commutator of two transformations localized in two regions is localized in their intersection. To a crossed precosheaf and a finite family of regions we attach a pointed connected homotopy type, the Cech-Loday classifying space. As the family of regions ranges over finite covers of space, the classifying spaces form a pro-object in the homotopy category. There is a canonical map from the ordinary classifying space of the
symmetry group to this pro-type. Composing this map with a group action gives an anomaly whose nontriviality obstructs gauging. In favorable cases, the Cech-Loday space is essentially independent of the cover. In such cases, the homotopy fiber of the canonical map is a candidate for the classifying space of the higher group of
symmetries. We give examples of crossed precosheaves arising from gauge transformations in continuum field theory and from
quantum cellular automata and finite-depth
circuits. For
quantum spin systems in one and two dimensions, our construction recovers the Else-Nayak index and its generalizations to two dimensions.
Read the full abstract ▾
Computation and Language
Evaluating Large Language Models for Multilingual Abstractive Summarization: A Comprehensive Study of English and Bengali
Khan Mahi Al Atahar, Nazifa Tasneem and Ajwad Abrar
Abstractive summarization remains challenging across
languages, where linguistic quality and factual consistency may diverge, particularly in low-resource settings. We present a standardized zero-shot
benchmark of eight closed-source and open-weight LLMs on size-matched English and Bengali XL-Sum test sets. Our evaluation combines lexical and semantic metrics with reference-free factuality verification and multidimensional LLM-as-a-Judge assessment, providing a comprehensive analysis of both surface-level quality and factual reliability in multilingual summarization. Results reveal substantial cross-lingual variation: GPT-5.6 Luna leads several English lexical metrics but does not maintain the same advantage in Bengali, where LLaMA 3.3 70B achieves the highest ROUGE and BLEU scores. Gemini 2.5 Flash obtains the highest observed Bengali factuality score. Furthermore, evaluation methods produce different model rankings, highlighting the limitations of relying on a single metric for multilingual summarization assessment. These findings demonstrate that strong English summarization performance does not necessarily transfer to Bengali and emphasize the need for comprehensive evaluation frameworks for low-resource
languages.
Read the full abstract ▾
General Relativity and Quantum Cosmology
The WDM time-frequency transform in gravitational-wave data analysis II: Evolutionary spectrum and noise covariance
Aaron D. Johnson, Katerina Chatziioannou and Robert Rosati
Noise cannot be treated as stationary over the timescales required for upcoming gravitational-wave detectors, minutes for ground-based and months for space-based detectors. Nonstationarity leads to a non-diagonal frequency-domain noise covariance
matrix, thus motivating a shift to the time-frequency domain. We introduce evolutionary power spectra for describing nonstationary noise in the Wilson-Daubechies-Meyer time-frequency basis. Under the framework of oscillatory processes, we show that, to leading order in the ratio of the Meyer filter and the nonstationarity timescales, the evolutionary spectrum is a power-frequency distribution that can be estimated from a single noise realization. We derive the corresponding noise covariance
matrix and show that it is band tri-diagonal, like its stationary counterpart. The exact covariance to all orders is positive semi-definite by construction and dense depending on the noise. We further derive the optimal tiling of the time-frequency plane that minimizes the error of a diagonal covariance approximation and show that coarse-graining the plane suppresses correlations quadratically. Using simulated data with a LISA Galactic foreground and a drifting powerlaw spectrum, we show that the stationarity approximation leads to signal-to-noise ratio loss and severely biases inference. For realistic LISA data, approximating the covariance as diagonal in the Wilson-Daubechies-Meyer domain is accurate for the narrow-band analysis of a single resolved Galactic binary, but insufficient for a broad-band analysis. Finally, we provide mitigation strategies for the wraparound correlations induced across the data time edges whenever the spectrum differs between the start and the end of the data.
Read the full abstract ▾
Cosmology and Nongalactic Astrophysics
Kinetic Theory of the EFTofLSS : A Halo-Based Approach
Om Kishor Hebbar, Caio B. de S. Nascimento and Arka Banerjee
The Effective Field Theory of Large Scale Structure (EFTofLSS) provides a coarse-grained description of the non-linear gravitational clustering of dark matter, in which the effects of short-scale physics are encoded in an effective stress tensor and a set of counterterms. In this work, we develop a framework to investigate the physical origin of these counterterms by relating the effective stress tensor to the dynamics of individual dark matter
halos in cosmological N-body simulations. We find that
halo accretion contributes to the trace of the
halo-based effective stress tensor, contributing to the background pressure and effective sound speed, while anisotropic
accretion generates the traceless part, consequently contributing to the shear viscosity. For the
halo population with $M_\mathrm{vir}>10^{12}\,M_\odot/h$, we find that
halos account for $\sim\!95\%$, $\sim\!85\%$, and $\sim\!70\%$ of the full box background pressure, effective sound speed, and shear viscosity, respectively, depending on the EFT cutoff scale. The contributions to the counterterms are dominated by cluster-scale
halos, $M_\mathrm{vir} \sim 10^{14} - 10^{15}\,M_\odot/h$ and
halos with larger inflow rates, supporting an interpretation in terms of
accretion-driven, out-of-
equilibrium dynamics. Our results provide a short-scale, kinetic-theory interpretation of the EFTofLSS counterterms and establish a framework for connecting the properties and dynamics of short-scale structures to physics-informed priors for large-scale cosmological observables.
Read the full abstract ▾
Instrumentation and Methods for Astrophysics
jax-SNANA: A GPU-accelerated Supernova Framework for the SBI Era
Matthew Grayling
We present jax-SNANA, a JAX reimplementation of the
supernova analysis framework SNANA. jax-SNANA enables GPU-accelerated transient simulations and light curve fitting,
supernova standardisation and cosmological inference. Running in single precision on an NVIDIA A100 GPU, jax-SNANA simulates more than 100,000 LSST-observed
supernovae per second and fits 5,000 light curves per second after JIT compilation, a two orders-of-magnitude increase over current tools, essential for analysing samples at the scale expected from LSST. jax-SNANA currently supports the SALT and BayeSN spectral energy distribution (SED) models as well as physical models from redback-jax and a variety of SN SED templates; light curve fitting is supported with both photometric and spectroscopic
redshifts. We fully validate the accuracy of jax-SNANA by comparing to SNANA, finding excellent consistency between the two across a variety of tasks. The code provides a modular JAX end-to-end
supernova cosmology pipeline, user-extensible to support new features including different SED templates and photometric classification
algorithms. jax-SNANA has been designed with future applications of simulation-based inference in mind, providing fast simulations and enabling fully compiled on-the-fly simulations to be generated within the training loop of downstream SBI tasks. jax-SNANA extends beyond SNANA by supporting simulations of generic SN-host
galaxy relationships and multiple sub-populations of events, for example originating from young and old progenitors. jax-SNANA is publicly available at https://github.com/mattgrayling/jax-snana and installable through pip.
Read the full abstract ▾
Materials Science
Electron-Mediated Phonon Polaritons in Quantum Cavities
Mattia Udina, David Hagenmüller and Indranil Paul
Phonon-
photon coupling in
quantum materials is commonly treated as an effective direct interaction, neglecting the crucial role played by electrons. Here we develop a gauge-invariant theory of cavity
phonon polaritons that captures electronic screening, dissipation, and hybridization. We show that screening can suppress polariton formation in metals, severely constraining cavity control of
phonon-mediated superconductivity. By contrast, interband electronic resonances in time-reversal-symmetric gapped systems can strongly enhance
phonon-
photon coupling, enabling robust polariton formation down to the monolayer limit despite electronic absorption. We illustrate these effects in NbSe2, hBN, and gapped
graphene. Finally, we show that gauge invariance enforces
photon decoupling from a critical soft
phonon, establishing a general no-go mechanism for
photon condensation.
Read the full abstract ▾
Astrophysics of Galaxies
The convective-transition gap in 47 Tucanae with JWST
M. Scalco, L. R. Bedin, M. Libralato et al. (+3 more)
We report the detection of the convective-transition gap along the lower main sequence (MS) of the Galactic globular cluster (GC) 47 Tucanae using deep James Webb Space
Telescope (JWST) observations. We analyse the stellar luminosity function (LF) in the vicinity of the gap location using two independent colour-magnitude diagrams (CMDs). In both cases, the LF is characterised by a narrow gap followed by a discontinuous increase in the stellar number counts, and presents an overall morphology qualitatively consistent with that reported in previous studies. By modelling the observed LFs with an empirical likelihood model, we estimate the statistical significance of the observed feature. Additional JWST passbands exhibit similar signatures, providing further evidence for the presence of a discontinuity.
Read the full abstract ▾
All authors M. Scalco, L. R. Bedin, M. Libralato, M. Griggio, E. Vesperini, R. Gerasimov
General Relativity and Quantum Cosmology
The Forgetful Black Hole: Memory Burden and Hysteresis in Hawking Evaporation
Yen Chin Ong
The
memory-burden scenario of black hole evaporation proposes that Hawking emission may become increasingly suppressed as maintaining stored
quantum information becomes dynamically costly. In this work, we revisit this scenario and investigate the conditions under which information retention can produce a genuine dynamical burden. We derive general scaling conditions for incremental
memory accumulation at a fixed fractional mass loss and establish a conditional no-go result:
entropy-suppressed, zero-mean deposits with subextensive integrated correlations cannot produce an order-one luminosity modification through unamplified accumulation and bounded feedback. We identify three essential requirements for dynamical
memory: a record of the past must be written, retained, and coupled back to subsequent evaporation. We illustrate these requirements with concrete toy models that exhibit either a prolonged but finite black hole lifetime or an asymptotically approached finite-mass state (an effective remnant). We further distinguish kinetic-
memory, thermodynamic-
memory, and spectral-
memory sub-categories of possible
memory-burden mechanisms. In particular, we demonstrate that an independently evolving
memory variable can generate evaporation-
accretion hysteresis, such that a black hole returning to the same mass may possess a different retained
memory and consequently a different luminosity. This history dependence may have implications for primordial black hole phenomenology. Finally, we briefly discuss the possibility of "merger rejuvenation", in which black hole mergers reorganize or reset the effective
memory state.
Read the full abstract ▾
Solar and Stellar Astrophysics
The mass functions of multiple stellar populations in 47 Tucanae
M. Scalco, J. Bruce, E. Vesperini et al. (+1 more)
We used a combination of James Webb Space
Telescope (JWST) and Hubble Space
Telescope (HST) observations to investigate the mass functions (MFs) of the different stellar populations in the globular cluster (GC) 47 Tucanae (NGC 104) over a wide stellar-mass interval. We identified the multiple populations along the lower main sequence using only JWST photometry and, in particular, colour combinations sensitive to variations in oxygen abundance. We derived the MFs of the first- and second-population (1P and 2P)
stars separately and found that in the low-mass regime, the present-day MF of the 1P population is significantly steeper than that of the 2P population, which appears nearly flat. We also present the results of a dynamical model showing that this behaviour requires a different initial mass function (IMF) for the 2P and the 1P populations in the 0.1-0.4 solar masses mass range, thus providing a new fundamental constraint for the study of the
star formation of multiple populations in GCs.
Read the full abstract ▾
All authors M. Scalco, J. Bruce, E. Vesperini, L. R. Bedin
Solar and Stellar Astrophysics
JWST observations of 47 Tucanae: I. The low-mass main sequence down to the hydrogen-burning limit
M. Scalco, R. Gerasimov, L. R. Bedin et al. (+4 more)
We present the first results of a series of studies based on JWST observations of the globular cluster 47 Tucanae (47 Tuc, or NGC 104) obtained with NIRCam. By combining new data with earlier JWST observations, we derive high-precision proper motions (PMs) over a multi-year baseline (of ~3 years) and construct a clean sample of cluster members extending to very faint magnitudes (F444W~24.25 mag, corresponding to ~0.077 solar masses, slightly above the estimated hydrogen-burning limit of 47 Tuc, ~0.074 solar masses) along the main sequence (MS). We investigated the presence of the low-mass MS kink previously identified in 47 Tuc using the new F090W and F444W photometry. No significant signature of the feature is found using these filters, consistent with its proposed physical interpretation as a consequence of the transition in carbon chemistry toward a methane-dominated
equilibrium in very low-mass
stars. A comparison with theoretical isochrones indicates that the stellar population near the bottom of the MS is likely dominated by oxygen-poor
stars, in agreement with previous findings. Using artificial-
star tests, we quantify the completeness of our sample and show that the requirement of PM-based membership significantly limits the detection of the faintest objects, indicating that current JWST observations of 47 Tuc are not sufficiently deep to reliably trace the cluster's brown dwarf cooling sequence. The mass function (MF) of 47 Tuc shows a structured behaviour, with a break at ~0.15 solar masses not associated with the onset of convection, and a mild but systematic radial variation consistent with mass segregation. Our analysis indicates that the MF shape in this mass regime is more complex than a simple power-law description. This feature in the MF may point to a presently unrecognised mechanism affecting the formation or dynamical
evolution of low-mass
stars in globular clusters.
Read the full abstract ▾
All authors M. Scalco, R. Gerasimov, L. R. Bedin, E. Vesperini, M. Griggio, D. Nardiello, J. Anderson