Peer Revenue Equity in science

Papers

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

All of arXiv

Friday 28 August

High Energy Physics (Experiment)

A High- and Variable-Dimensional Measurement of the $Z$+jets Differential Cross Section with the ATLAS Experiment and Artificial Intelligence

Proton-proton collisions at the Large Hadron Collider (LHC) offer the opportunity to observe the interactions of fundamental particles at very high energy scales. The high collision rate and large number of final state particles produced per collision imply that the datasets produced by detectors such as the ATLAS experiment are large and high dimensional. The complexity of these datasets calls for the use of novel data analysis techniques which can exploit all of the available information to illuminate known interactions and search for new ones. This thesis presents applications of artificial intelligence (AI) techniques to improve the physics results of the ATLAS experiment. It centers on a full-phase-space measurement of the Z+jets production cross section at the LHC, where the cross section is measured differential in the kinematics of every final state charged particle through the use of an AI-based unfolding algorithm. This is the first such measurement performed at the LHC, which provides a complete experimental characterization of the Z+jets production process and a proof-of-principle for the further pursuit of such measurements on a host of LHC processes.
Read the full abstract

High Energy Astrophysical Phenomena

Reassessing the contribution of long-GRB high-energy afterglows to the isotropic gamma-ray background after GRB 221009A

Very-high-energy photons from gamma-ray bursts (GRBs) are absorbed by the extragalactic background light and partly reprocessed into the GeV band. We update previous estimates of the long-GRB contribution to the Fermi-LAT isotropic gamma-ray background (IGRB), using a time-integrated high-energy afterglow template of GRB 221009A, anchored by GeV-TeV data, together with an explicitly specified scaling with prompt energy and beaming. Despite the observed multi-TeV afterglow, the cascade component reaches only 1.6e-4 of the IGRB. Including direct GeV emission raises the maximum fraction to 1.1e-3, but makes the result more dependent on the population-averaged broadband template. Variations of the extragalactic background light, propagation implementation, intrinsic spectral cutoff, and population prescription do not change the conclusion: the multi-TeV detection of GRB 221009A does not make long GRBs a significant source of the IGRB.
Read the full abstract

Artificial Intelligence

Learning to Use Tools: Reinforcement Learning for Tool-Integrated Mathematical Reasoning

Current large language models (LLMs) increasingly benefit from external tool integration, especially for tasks requiring reliable computation and verification. Motivated by this, we study calculator tool calling for improving mathematical reasoning on the Countdown task. We first analyze reasoning failures and find that calculation errors account for a substantial portion of incorrect responses. We then construct supervised fine-tuning datasets to teach the model useful tool-use patterns and how to interpret returned outputs. Building on this tool-formatted policy, we apply several on-policy reinforcement learning methods, including RLOO, RLOO++, GRPO, and DAPO, using automatically verifiable final-answer rewards. To enable a more reliable evaluation, we construct a fresh 1,024-problem held-out Countdown benchmark with no exact overlap with the training data. Our results show that calculator tool integration consistently improves both SFT and RL baselines, yielding roughly 10 percentage-point gains across pass@k. Among the RL methods, Tool-DAPO achieves the strongest performance, improving pass@1 from 35.8% for Tool-SFT to 66.0%. Further analysis shows that RL encourages more effective tool use even when only final-answer rewards are provided. These findings suggest that tool integration reduces arithmetic and verification errors, while RL increases the probability of correct reasoning traces.
Read the full abstract

Statistics Theory

Generalized Splines and Gaussian Processes

For finite-dimensional linear inverse problems where the variables are Gaussian, it is well-known that the minimum-mean-square error estimator takes the form of a regularized least-squares data fit. In this chapter, we show that this equivalence extends to a much broader infinite-dimensional setting where generalized splines take the role of linear regressors and generalized Gaussian processes on a nuclear space $S$ are the counterpart of Gaussian random vectors. The scope of this extension is of the same nature as the switch from the classic notion of function to that of a distribution, also known as a "generalized function." Our formalism involves a whitening/regularization operator $L: S\to S'$ whose continuous extension induces a native Hilbert space $H\subset S'$ that plays a central role in our characterization. The presentation is self-contained for the most part and remarkably general and powerful. It allows for the recovery of all known instances of such equivalences; in particular, the methods involving innovations and reproducing-kernel Hilbert spaces developed by Kailath and his students, and the mathematical correspondence between fractional splines and Mandelbrot's fractional Brownian motion (fractals), with the former being the optimal estimators of the latter. It also covers general Bayesian methods for the resolution of infinite-dimensional inverse problems.
Read the full abstract

Operator Algebras

Associativity of operator convolutions for group actions on von Neumann algebras

Given a pair of commuting, ergodic, trace-preserving and trace-integrable actions of locally compact unimodular groups on a common semifinite von Neumann algebra we prove a form of associativity for convolutions of triples of operators. We view this as a basis for a general version of quantum harmonic analysis.
Read the full abstract

Computation and Language

Sliding-window beats linear attention

Due to the nature of quadratic attention, Large Language Models (LLMs) consume a lot of memory and energy: every new token costs more than the previous one, and its keys and values must be stored in memory indefinitely, which is unsustainable. Two main lines of work address this: compressing the KV cache, e.g., by evicting or quantizing keys and values, and retrofitting LLMs to use Linear Attention, which replaces the KV cache with a fixed-size state. Retrofitting has attracted a lot of attention, given its promise to solve the quadratic scaling problem with state-of-the-art performance at low cost. However, it has not been properly compared to the simplest form of KV-cache eviction: Sliding Window Attention (SWA) with attention sinks. In this work, we show that SWA with sinks performs as well or better than most retrofitted Linear Attention models across multiple LLMs and downstream tasks, with the largest gains on long-context and generative tasks. On long-context reasoning tasks (Needle-in-a-Haystack and BABILong), SWA achieves massively higher performance (2 to 10 times higher than linear attention). SWA requires no additional training, is extremely fast, and requires little memory, making it an extremely cheap and reliable solution. When the training budget is limited, switching to SWA is a much more effective way to reduce inference memory cost than retrofitting linear attention. Linear attention models have shown promise, but they require training from scratch or extensive retrofitting to reap their architectural benefits and come close to SWA.
Read the full abstract

All authors Alexia Jolicoeur-Martineau, Rhea Sanjay Sukthanker, Pashmina Cameron, Emy Gervais

Logic in Computer Science

Self-extensional logics of formal inconsistency: Decidability and limits for paraconsistency

RmbC is a self-extensional paraconsistent logic in the family of Logics of Formal Inconsistency (LFIs). This system is obtained from mbC (the basic LFI) by adding the replacement property via two global inference rules. RmbC is characterized by a non-explosive negation $\neg$ and a consistency operator $\circ$, which recovers the principle of explosion in a controlled way. Together with its principal axiomatic extensions, RmbC admits a standard Lindenbaum-Tarski algebraization, with Boolean algebras with LFI operators (BALFIs) as its algebraic semantics. In this paper, we study how far this self-extensional paraconsistent behavior can be extended axiomatically, starting from RmbC. We classify pairs of very natural consistency axioms according to whether they preserve paraconsistency or force classical collapse; identify six minimal explosive combinations that collapse to a single algebraic core; and isolate a separate structural obstruction for the combination of excluded middle for $\neg$ with an involutive negation. We also investigate, for the first time, the decidability of this family of self-extensional LFIs. As a first result, we prove the finite model property for RmbC with respect to BALFI semantics via an algebraic filtration, which yields decidability, and transfer this result to several paraconsistent axiomatic extensions of RmbC. Finally, we establish a coNEXPTIME upper bound for the validity problem of RmbC and a coNP-hardness lower bound, and prove coNP-completeness for the principal extensions containing one of the six minimal explosive pairs.
Read the full abstract

Machine Learning

Curvature-Conditioned Multiscale Momentum with Sphere Constraints for LLM Pretraining

Pretraining accounts for a large fraction of the total computational cost in LLM training. However, noise-dominant gradients and the highly ill-conditioned loss landscape bring severe challenges. Although modern adaptive optimizers such as AdamW and Muon have achieved great success in large-scale pretraining, their reliance on gradient normalization offers limited mitigation of the ill-conditioned curvature. The progress along flat directions (eigen-directions of small eigenvalues), which dominates the final loss reduction, remains relatively slow. To enhance training dynamics along flat directions, we propose a curvature-conditioned multiscale momentum method with sphere constraints, delivering steady acceleration in LLM pretraining. This multiscale momentum, applied only along flat directions, pairs a slow-decay component for noise reduction with a fast-decay component for rapid curvature adaptation, harnessing their complementary strengths. Crucially, we employ a sphere constraint technique to prevent parameter inflation and excessively rapid effective learning rate decay that would otherwise arise from a naive combination. Extensive experiments show that the proposed method significantly accelerates Muon across diverse architectures (dense, MoE) and model sizes (0.12B–2.3B parameters). Theoretically, we verify the acceleration effect and provide insight into the design principles underlying the flat-direction multiscale momentum.
Read the full abstract

All authors Shuchen Zhu, Yuxin Fang, Mingze Wang, Kun Yuan

Information Theory

Significance-Driven Semantic Communication

In this paper, we study a significance-driven cross- layer semantic communication design problem. Based on sta- tistical decision theory, we introduce an information-theoretic measure of per-sample data significance that quantifies the task-specific value of each individual observation. Using this metric, we formulate a cross-layer optimization problem that simultaneously optimizes (i) physical-layer semantic encoding and inference and (ii) MAC-layer resource allocation, with the objective of maximizing semantic spectrum efficiency, defined as the semantic value delivered per unit bandwidth per unit time. At the physical layer, we develop Meta-Learning Variational Information Bottleneck (Meta-VIB), a new semantic transceiver that employs a meta-learned hypernetwork to compress high- dimensional observations into semantically significant latents, enabling instantaneous adaptation to dynamic channel conditions and varying symbol budgets without online retraining. At the MAC layer, we model channel allocation as a Multi-Action Restless Multi-Armed Bandit (MA-RMAB) and adopt the Q- Maximization algorithm, which dynamically allocates channel resources to sensors based on their semantic value of information. Experimental results on a real-world pedestrian safety dataset demonstrate that our joint design achieves substantial gains in semantic spectrum efficiency over baselines, reaching up to 1000 times gain at an average SNR of 0 dB and 40 times gain at an average SNR of 5 dB.
Read the full abstract

All authors Christian McDowell, Andrea Panebianco, Sirin Chakraborty, Yin Sun

Quantum Physics

From quantum reservoirs to quantum extreme learning machines through a nearest-neighbor spin chain with tunable quantum memory

Quantum Reservoir Computing (QRC) processes temporal data by retaining a memory of past inputs in the recurrent state of a quantum system, whereas a Quantum Extreme-Learning Machine (QELM) discards that memory, resetting the system at every step so that only the most recent input shapes the response. The two are usually treated as separate computational paradigms. We show that they are the two limits of a single architecture, connected by the input-encoding length, that is, the number of qubits overwritten with fresh data at each step. When a single qubit is re-encoded the system operates as a standard QRC, when the whole register is re-encoded it operates as a QELM, and intermediate lengths interpolate between them. The overwritten qubits hold the recent past in an explicit register, while the remaining qubits are never reset and carry older inputs forward in their evolving quantum state, so the encoding length redistributes memory between explicit and recurrent storage at fixed system size. Tuning the reservoir Hamiltonian and the evolution time with Bayesian optimization at each encoding length, we find that recurrent quantum memory is essential when a task must reach far into the past, and dispensable when the relevant history is short, where the memoryless reset limit already suffices. For every task the best reservoirs operate at the edge of chaos, where they perform as well as a densely connected reservoir with random all-to-all couplings of the same size, indicating that what temporal processing requires is the dynamical regime rather than the connectivity.
Read the full abstract

Computation and Language

Fidelity Is Not Enough: Dispatch-Level Instrumentation for Agentic Datasheet Extraction

One model passed our fidelity check without ever opening the datasheet. We found it while qualifying models for an internal extraction service: a structured-output constraint had silently disabled tool use, and the model answered anyway, with fabricated source text. Only the per-tool trace exposed it. Fidelity – whether an extracted value matches the source – is the standard measure for agentic document extraction, and it scores that run a success. We therefore log every tool call in an agentic benchmark of 25 hand-curated claims over three components, with 12 more on a fourth, 37 in all. From that dispatch record we build two instruments: a rule-based failure-attribution classifier, and a silent-failure detector whose two rules check only which tools were called, never the extracted value. The detector raises no flag on 207 clean fidelity-passing extractions across three model families, and recovers all 50 planted faults that withhold exactly the tools its rules check. The two results are not symmetric: the first bounds the false-positive rate, the second is recall by construction, and detection power against runs that call their tools and still answer wrongly is unmeasured. A second, independent oracle, a causal chamber that tests whether the datasheet's claims hold under physical measurement, is intentionally partial: it confirms only what the apparatus can exercise, a verifiable envelope of 2 of those 37 claims, and we give a taxonomy of why the rest are not physically gradable. Under a controlled perturbation, fidelity passes throughout while the chamber verdict flips exactly at the measurement uncertainty. Across three deployed model stacks (one destabilised by its serving stack, not by any capability gap) the tool layer buys portability and observability rather than accuracy, and earns its premium only once a document outgrows the context window.
Read the full abstract

Algebraic Geometry

Fujita freeness for projectivized toric vector bundles

Let $X$ be a smooth projective toric variety of dimension $n\geq1$ over an algebraically closed field of characteristic zero, let $\mathcal E$ be a toric vector bundle of rank $r\geq2$, and let $\pi\colon Y=\mathbb P_X(\mathcal E)\to X$ be the projective bundle of one-dimensional quotients. Write an ample line bundle on $Y$ as $A=\mathcal O_Y(a)\otimes\pi^*L$, with $a\geq1$. We record a blow-up argument proving that $K_Y+mA$ is globally generated whenever an integer $m$ satisfies $ma\geq r$ and $m\delta(A)>n$, where $\delta(A)$ is a positive integer obtained from the degrees of $A$ on the invariant quotient sections over the torus-invariant curves of $X$. In particular, $K_Y+mA$ is globally generated for $m\geq n+1$ and $ma\geq r$. Consequently every projectivized toric vector bundle satisfies Fujita's freeness conjecture. The uniform bound is sharp. We also formulate the result as a global-generation theorem for adjoint symmetric powers of $\mathcal E$ and explain its relation with the Seshadri-constant results of Hering–Mustaţă–Payne and Fulger–Murayama. ChatGPT (OpenAI) was used to assist with mathematical discussion, language, and bibliographic searches.
Read the full abstract

Systems and Control

LUCID: An Agentic AI Framework on Digital-Twin in the Loop for QoS-Guaranteeing Robotic Control

Cloud robotics relies on the timely uplink of high-volume sensing streams, yet dynamic environments continually shift the feasible combinations of trajectories, active-robot count, and per-robot QoS. Because existing approaches formulate trajectory planning (TP) and radio resource management (RRM) as a single fixed optimization problem, they cannot reconfigure these coupled decisions as conditions evolve, resulting in transient QoS violations. However, evolving operator intents change which quantities-such as the active-robot count and per-robot QoS-are fixed, optimized, or relaxed. Furthermore, the computational cost of evaluating trajectory-dependent wireless conflicts has made it difficult to build large-scale Digital-Twin-in-the-Loop (DITL) testbeds responsive enough for such dynamic orchestration. We present LUCID, an LLM-agent–orchestrated, uplink-aware cloud-robotics pipeline that moves TP–RRM from solving a fixed formulation to dynamically orchestrating optimization problem schemas within a DITL environment. Driven by the operator's high-level intent, LUCID treats the TP–RRM formulation as a bounded template whose variables, objectives, and constraints are dynamically configured, while SimBridge enables repeated ray-tracing evaluation by converting large-scale robotics scenes into wireless-ready DTs. By integrating collision-free path planning with a spectral-radius RRM validator, LUCID identifies wireless bottlenecks and restructures the problem schema on the fly to efficiently find the verified feasible state. Experiments confirm that LUCID robustly adapts to changing intents, active-robot counts, and scenes, while a multimodal surrogate model, FastConfigNet, reduces planning latency.
Read the full abstract

All authors Hyeonsu Lyu, Minwoo Kim, Sehyun Ryu, Hyun Jong Yang

Combinatorics

Singular difference graphs of vector spaces of square matrices

The singular difference graph, denoted by $\Gamma$, of the vector space of square matrices over a field is a graph whose vertex set is the set of all elements of the vector space, where two distinct vertices are adjacent if and only if the difference of the corresponding matrices is singular. In this paper, we investigate fundamental graph-theoretic properties of $\Gamma$, including connectivity, diameter, regularity, the Eulerian property, independence number, clique number, and domination number. We show that $\Gamma$ is a connected regular graph with diameter two. Over finite fields, we obtain an explicit formula for the degree of each vertex and characterize precisely when $\Gamma$ is Eulerian. We determine the independence number and clique number and provide explicit constructions attaining these values using companion matrices of irreducible polynomials. We also construct an explicit dominating set, yielding an upper bound for the domination number.
Read the full abstract

Robotics

Linear Temporal Logic Translation via Human-Inspired Self-Constrained Reasoning for Robot Task Specification

Many robotic tasks are temporally extended and demand precise specifications of subgoals, constraints, and their temporal ordering. Yet human operators typically communicate such tasks in natural language, which is inherently ambiguous, underspecified, and context dependent. Translating human instructions into formal task specifications, such as Linear Temporal Logic (LTL), is therefore essential for verifiable and safe robotic execution. Existing LLM-based translators attempt to bridge this gap through open-ended reasoning or post-hoc constraint enforcement, but the former may violate domain constraints, whereas the latter can disrupt the reasoning needed for novel instructions. This paper proposes Self-Constrained Reasoning (SCR), a framework that mitigates this trade-off by internalizing structural knowledge into the model's decision-making process rather than imposing it as an external filter. By combining a structural constraint representation with a hierarchical decision-making formulation, SCR guides reasoning within a formally grounded space while preserving adaptability to unseen instructions. Experiments show that SCR improves both domain-constraint satisfaction and generalization, providing an effective and interpretable approach for translating human intent into verifiable specifications for robotic execution.
Read the full abstract

All authors Haofei Hou, Fanxu Meng, Shunyi Zhao, Kairui Yang, Mengchen Cai, Lecheng Ruan, Qining Wang

Astrophysics of Galaxies

Characterising the Globular Cluster Systems of Three Local Group Dwarf Galaxies: NGC6822, NGC147 and NGC185

We present deep HST photometry for 26 globular clusters (GCs) residing in three Local Group dwarf galaxies: the isolated dwarf irregular (dIrr) galaxy, NGC6822, and the M31 dwarf elliptical (dE) satellites, NGC147 and NGC185. From their colour-magnitude diagrams (CMDs), we quantify their red giant branch (RGB) and horizontal branch (HB) morphologies, and employ new empirical relationships to derive measurements of metallicity, reddening and distance. Additionally, we measure sizes and $V$-band magnitudes from their integrated light profiles. We find that the clusters span a range of metallicities, from [Fe/H] $\sim-0.7$ to $\lesssim-2$; however, the three dwarfs have very similar mean GC metallicities ($\sim-1.7$ dex) despite their very different evolutionary histories. In contrast, we find that almost all of NGC6822's GCs exhibit red HB morphologies, whereas those in the two dEs are predominantly blue. We highlight three outlying GCs in NGC6822 that have very low metallicities yet very red HBs; they are also the most extended clusters in our sample, with half-light radii of $\sim 13-17$pc. If these clusters are young, their origin is difficult to explain given their very remote locations, and metallicities which are significantly lower than those of the old and intermediate-age stars in NGC6822. These three clusters are strikingly similar to several GCs linked to substructure in the outer halo of M31, suggesting either the recent accretion of an NGC6822-like dwarf by M31, or that both M31 and NGC6822 have recently accreted a similar low-mass system. Additionally, we find compelling evidence that another NGC6822 cluster, SC7, is tidally distorted.
Read the full abstract

All authors Gracie McGill, Annette M. N. Ferguson, Dougal Mackey, Chiara Crociati, Jess M. Howell, Avon P. Huxor, Mike J. Irwin, Alan W. McConnachie, Alessandro Savino, Nial R. Tanvir, Daniel R. Weisz

Artificial Intelligence

Prove2Me: An Open Collaborative Platform for Scaling Math Formalization

Proof assistants such as Lean 4 promise the paradigm of formally verified mathematics, but large-scale formalization projects have faced major barriers to entry, including the need for expertise in formal verification (as well as the underlying mathematics) and the significant time required for writing formal proofs. AI coding agents have dramatically reduced these barriers; human users can now use natural language to prompt agents to write complex proofs in Lean. This opens up the intriguing possibility of internet-scale mathematical collaboration involving both humans and AI agents, where correctness is machine-checked. To realize this possibility, we introduce Prove2Me (https://prove2.me), an open collaborative platform for formalizing mathematics. Users launch formalization "missions", to which AI agents contribute formal proofs toward completion. We designed mechanisms and a specialized harness in Prove2Me that enable large-scale collaboration so that agents can build on one another's work and freely reuse existing results. In doing so, Prove2Me aims to turn math formalization into a scalable, crowd-sourced effort open to anyone with an agent.
Read the full abstract

All authors Shuze Chen, Kunal Marwaha, Xiaoyang Lu, Henry Yuen, Tianyi Peng

Computation and Language

Are These Modules Worth Their Cost? A Paradigm-Level Accuracy-Cost Analysis of In-context Learning Text-to-SQL

Recent advances in in-context learning (ICL) text-to-SQL have substantially improved execution accuracy on public benchmarks by assembling increasingly elaborate pipelines around the base generator, yet existing studies typically report aggregate end-to-end accuracy, without quantifying the marginal accuracy-cost contribution of individual design choices. Consequently, providing a unified, paradigm-level cost-accuracy quantification remains a critical challenge for understanding and configuring modern text-to-SQL. To address this, we instantiate 17 paradigm-level configurations across five recurring modules of the ICL text-to-SQL pipeline under a single controlled implementation, and attribute each paradigm's marginal contribution and incurred cost across all four backbones spanning diverse capability levels and reasoning styles. Our analysis reveals that execution-feedback refinement is the only paradigm whose benefit holds universally at consistently low cost, while most other modules help only under backbone-dependent conditions. Token accounting shows that input demand is more closely tied to pipeline structure, whereas output demand is more sensitive to backbone generation behavior. Cross-module analysis further shows that stacking improves accuracy on most backbones, although how the gains compose varies with backbone capability. We also find that a fixed budget is often better spent engineering a more elaborate pipeline over a mid-tier backbone than upgrading to a frontier model with a lean pipeline. These findings distill into an actionable, cost-aware tiered guideline that transfers to five additional backbones without per-paradigm search.
Read the full abstract

All authors Jiayan Lin, Yujia Liu, Zijin Hong, Zheng Yuan, Yilin Xiao, Hao Chen, Qinggang Zhang, Xiao Huang, Feiran Huang

Neurons and Cognition

Adaptive self-organized criticality in deep neural networks

Deep neural networks are high-dimensional dynamical systems whose function depends on the stable propagation of activity and perturbations across many layers. Maintaining suitable dynamical regimes may therefore be essential for robust learning and for preventing dynamical instabilities during training. Here, we show that the global dynamical state of a deep neural network can be autonomously regulated by purely local homeostatic plasticity. Neuronal activity is inferred from responses across inputs, and individual synapses are strengthened or weakened using only the activity of their postsynaptic neuron. Without measuring any global network property, this rule drives networks from both subcritical and supercritical initial conditions toward a common critical state, characterized by conserved activity propagation and a vanishing largest finite-time Lyapunov exponent. When combined with gradient-based learning, homeostatic adaptation counteracts the training-induced drift toward supercritical dynamics, while revealing a competition between dynamical regulation and task optimization. Our results demonstrate how adaptive self-organization can be implemented in deep neural networks and how local plasticity can control their collective dynamical operating point.
Read the full abstract

Cosmology and Nongalactic Astrophysics

Follow-up of SN 2025wny V: Lens Modelling and Cosmography of a Strongly Lensed Superluminous Supernova at $z = 2.015$ using Space Data

We present a lensing and cosmographic analysis of the strongly lensed Type I superluminous supernova SN 2025wny at redshift $z=2.015$, multiply imaged by two foreground galaxies at $z=0.376$. Using imaging obtained with the Hubble Space Telescope and the James Webb Space Telescope, we model the lens system with two elliptical power-law mass distributions and an external shear component. In addition to the supernova image positions, the modelling incorporates the surface-brightness distribution of the lensed host galaxy. The inferred Einstein radii are $\theta_{\rm E,1}\simeq 1.6$" and $\theta_{\rm E,2}\simeq 0.7$ - $0.8$", with broadly consistent results across all filters. After accounting for microlensing by stars in the lens galaxies, the posterior distribution spans total magnifications of approximately $\mu_{\rm tot}\sim 5$–$50$, with flux ratios of the multiple images consistent with observations. Combining the lens models with spectroscopically and photomerically measured time delays yields a filter-marginalized constraint of \[ H_0 = 66.7^{+7.6}_{-6.3}\; {\rm km\,s^{-1}\,Mpc^{-1}}, \] for a fiducial model with isothermal mass profiles. Allowing the density slopes of the lens galaxies to vary over a broad range results in \[ H_0 = 70.8^{+8.2}_{-6.1}\; {\rm km\,s^{-1}\,Mpc^{-1}}. \] These values are conditional on the adopted parameterization of the lens mass distribution, the assumed priors on the density slopes, and possible additional lensing contributions from the surrounding large-scale environment. We find that incorporating the currently available stellar kinematic measurements has only a modest effect on the inferred value of $H_0$. Future measurements of the lens-galaxy kinematics and a detailed characterization of the lens environment will further strengthen the utility of SN 2025wny as a cosmological probe.
Read the full abstract

All authors Edvard Mörtsell, Joel Johansson, Ariel Goobar, Alice Townsend, Hannah C.Turner, Suhail Dhawan, Cameron Lemon, Peter Nugent, Thomas E. Collett, Stephen Thorp, Jacob Osman Hjortlund, Jakob Nordin, Lin Yan, Graham P. Smith, Christoffer Fremling

Computer Vision

Lossy Event Compression: From Event Stream Distortion to Task Performance

Event cameras generate asynchronous, sparse data streams with microsecond temporal resolution, but in moderate-to-high motion scenes they can produce as many as hundreds of millions of events per second, creating significant bandwidth and storage challenges. Lossy compression is therefore essential for practical deployment, yet existing event stream distortion metrics fail to reliably predict compression-induced degradation at the task level, forcing codec optimization to rely on expensive task-specific evaluations. To address this gap, this paper introduces two fundamentally different event compression pipelines: i) an aggregation-based pipeline that converts the event stream into polarity-based histogram frames for compression with the conventional image codec JPEG 2000, and ii) a frame-free point cloud-based pipeline that codes events natively as 3D points using the octree-based codec G-PCC. Both pipelines are then assessed within a unified task-driven evaluation framework that relates event stream distortion to downstream application performance across four representative tasks: i) video reconstruction, ii) object detection, iii) optical flow estimation, and a delay-sensitive task iv) asynchronous feature tracking under a reference-relative protocol. Building on this framework, five classification-based distortion metrics are applied to event compression for the first time, to the best of the authors' knowledge, and benchmarked against existing event stream metrics. Experimental results demonstrate that the proposed metrics reliably predict compression-induced task degradation across different coding frameworks. This demonstrates that event stream distortion assessment can be an efficient alternative to repeated task-specific evaluation, providing direct guidance for the development and optimization of future event data coding solutions.
Read the full abstract

Quantum Physics

Kerr nonlinearity and three-wave mixing in superconducting resonators hosting Al-InAs weak links

Nonlinear microwave resonators are a versatile tool in quantum information processing, enabling parametric amplification, continuous variable quantum computing, and engineered mode interactions. Many of these applications especially benefit from cubic nonlinearities enabling three-wave mixing; at the same time, they are limited by quartic nonlinearities giving rise to undesired Kerr effects. A recurrent challenge is therefore to engineer resonators with a finite cubic nonlinearity while suppressing quartic terms. Here, we investigate a superconducting resonator hosting two weak links fabricated from an aluminum-capped indium arsenide nanowire. We characterize the Kerr nonlinearity as a function of magnetic flux and gate bias, showing that it can be tuned to zero with either control parameter. Furthermore, we experimentally demonstrate three-wave mixing in a semiconductor-superconductor hybrid device, establishing nonzero cubic nonlinearity. An effective model based on Andreev bound states qualitatively captures the observed trends. Our results validate semiconductor-superconductor hybrid devices as a promising platform for tunable nonlinear superconducting circuits, with applications in parametric amplification, quantum control of bosonic modes, and engineering interactions between microwave modes.
Read the full abstract

All authors Vittorio Buccheri, Ivo P. C. Cools, Nermin Trnjanin, Ankit Khola, Oleg Shvetsov, Thomas Kanne, Jesper Nygård, Attila Geresdi, Simone Gasparinetti

Cosmology and Nongalactic Astrophysics

Follow-up of SN 2025wny IV: Photometric Time-delay Measurements of a Strongly Lensed Superluminous Supernova

We present photometric time-delay measurements of SN 2025wny, the first strongly lensed Type I superluminous supernova (SLSN-I), discovered at $z = 2.015$. Time-delay measurements from strongly lensed supernovae provide an independent probe of cosmology and the Hubble constant, $H_0$, without reliance on the local distance ladder. Using multi-facility imaging data, we performed scene-modelling photometry to deblend four of the lensed images (A-D) and construct $grizJ$-band light curves. We modelled the resolved light curves with Gaussian process regression using GausSN (Hayes et al. 2024) to infer relative time delays and magnifications between the lensed images. We found that a constant magnification model provides a suboptimal description of the data, motivating a time-dependent sigmoid magnification model to account for evolving relative magnification of image A. We measured time delays of $\Delta t_{AB} = -10.6^{+2.2}_{-2.5}$ days and $\Delta t_{AC} = 1.2^{+2.7}_{-2.6}$ days (68% credible intervals), consistent with independent spectroscopic measurements from Johansson et al. (2026). Combining the photometric time delays with the lens model of Mörtsell et al. (2026) gives $H_{0,\:\rm photo} = 80.5^{+26.4}_{-16.7}\;\rm km\,s^{-1}\,Mpc^{-1}$, while including the spectroscopic time delays as well yields $H_{0,\:\rm comb} = 70.8^{+8.2}_{-6.1}\;{\rm km\,s^{-1}\,Mpc^{-1}}$. Our results further demonstrate the potential of strongly lensed supernovae as independent probes of $H_0$.
Read the full abstract

All authors Alice Townsend, Suhail Dhawan, Erin E. Hayes, Maggie L. Li, Joel Johansson, Edvard Mörtsell, Ariel Goobar, Lin Yan, Charlotte Ward, Veena Krishnaraj, Steve Schulze, Jacob Osman Hjortlund, Yu-Jing Qin, Hannah C. Turner, Peter Massey, Jakob Nordin, Jule Augustin, Aleksandra Bochenek, Malte Busmann, Christoffer Fremling, Daniel Gruen, Xander J. Hall, K.-R. Hinds, Ezequiel J. Marchesini, Zoë McGrath, Conor M. B. Omand, Eliana Palazzi, Daniel A. Perley, Andrea Rossi, Killa Santer, Jesper Sollerman, Chiara Ventura, Jacob L. Wise, Tracy X. Chen, Steven L. Groom, Mansi M. Kasliwal, Josiah Purdum

Dynamical Systems

On the slow passage through a Hopf in a class of real-analytic systems: Exponential asymptotics and maximal delay

In this paper, we consider the slow passage through a Hopf in $\mathbb R^3$ in a certain class of real-analytic systems. Classically, the (simplified) Shishkova problem: $\epsilon \frac{dz}{d\mu} = \lambda(\mu) z+\mathcal O(\epsilon)$, $\lambda(\mu) = \mu-\i$, has been used to illustrate the delay associated with slow passage through a Hopf at $\mu=0$. In this paper, we consider a general $\lambda(\mu)$ with $\lambda(0)\in i\mathbb R\setminus \{0\}$, $\real[\lambda'(0)]\ne 0$, and add an equation for the complex conjugate. It is a basic fact that these systems are local normal forms for any real-analytic $(1,2)$ slow-fast system in $\mathbb R^3$ experiencing slow passage through a Hopf. In this paper, we study these systems under additional (global) assumptions on $\lambda$. In particular, we suppose that $\lambda$ has a simple zero away from the real axis. The remaining assumptions then relate to invariant manifolds of the so-called elliptic system $\dot \mu = -\i\overline{\lambda(\mu)}$ as well as properties of the level set $\operatorname{Re}[\i{\lambda(\mu)}{\lambda(\overline \mu)}]=0$. Under these assumptions, we then provide an asymptotic formula for the exponentially small splitting of attracting and repelling slow manifolds. As a corollary, we also obtain an asymptotic formula for the maximal delay. Our approach is geometric and is inspired by the work of Hayes et al (2016) (using blowup) and Neishtadt (1987,1988) (using so-called elliptic paths) but also by recent work of the second author on exponentially small splitting in unfoldings of the zero-Hopf bifurcation.
Read the full abstract

Machine Learning

Euclidean Fourier Neural Operators

Fourier neural operators (FNOs) provide an efficient framework for learning mappings between function spaces as they are, by construction, independent of the grid resolution at which they are trained and evaluated. However, FNOs are not independent of the periodic domain they are applied to: their discrete spectral weights are indexed by integer Fourier mode numbers, which correspond to physical wavevectors. When applied to a different domain, the same trained weights act at different wavevectors, and the FNO silently represents a different operator. This makes FNOs unsuitable for tasks where transfer across domains is crucial. We propose Euclidean Fourier neural operators~(EFNOs) as a domain-independent alternative to FNOs. By parameterizing the spectral kernel as a continuous function of the physical wavevector, the EFNO can learn operators that act consistently across periodic domains of varying shape and size. We evaluate the EFNO on a simple heat equation and on a practically relevant materials science task of learning exchange-correlation potentials across different crystal structures, and demonstrate that the EFNO is able to generalize to unseen grid sizes and domains.
Read the full abstract

Chemical Physics

Third-Order Algebraic Diagrammatic Construction Renormalized by Particle-Particle Ladders

The algebraic diagrammatic construction (ADC) of the electronic self-energy is based on Møller–Plesset perturbation theory. Its vertices contain bare energy denominators that do not capture renormalization of Hartree–Fock excitations through excited-state correlations. As we show here, replacing the Møller–Plesset amplitudes by amplitudes obtained from Epstein–Nesbet (EN) partitioning of the electronic Hamiltonian overcomes this deficiency at no extra cost by introducing dressed denominators that resum to infinite order the diagonal elements of ladder diagrams described by the pair propagator in the particle-particle $T$-matrix approximation. Third-order ADC with EN-dressed denominators leads to an almost fourfold increase in accuracy compared to standard ADC(3) on a challenging benchmark set for ionization potentials of closed-shell molecules, reaching an accuracy between that of equation-of-motion coupled cluster with singles and doubles (EOM-CCSD) and with singles, doubles, and triples. For open-shell systems, EN-ADC(3) halves the error of ADC(3) for weakly spin-contaminated molecules, outperforming EOM-CCSD, and still improves on ADC(3) for strongly spin-contaminated ones, for which EOM-CCSD remains more accurate.
Read the full abstract

Optics

High-performance silicon-metal laser welding resisting extreme conditions

Reliable material joining is essential for countless industrial applications. While femtosecond laser welding provides a route beyond conventional bonding methods, demonstrations of silicon-metal joints are rare due to nonlinear propagation effects and have so far been limited to shear joining strengths of a few MPa. Here, we demonstrate high-strength silicon-Kovar laser welding using sub-nanosecond pulses. By optimizing the focal position, the welding pattern, the laser polarization, and the metal roughness, remarkable shear joining strengths up to 15.8 MPa are achieved. The silicon-metal joints withstand temperatures of up to 500 {\deg}C and are hermetically sealed. Together with the remarkable strength values, the resistance to harsh environments underpins the applicability of silicon-metal welding in various fields including aerospace, nuclear science, and metallurgy.
Read the full abstract

All authors Jiawei Yin, Markus Blothe, Hagen P. Kohl, Christina Schütze, Stefan Nolte, Maxime Chambonneau

Audio and Speech Processing

Effects of HRTF Augmentation on Predicted Spatial Release from Masking in Music

Separating individual musical instruments within a complex mixture of sounds poses a persistent challenge for listeners with hearing loss. Although spatial separation of sources improves speech recognition in this population, the potential benefits of spatial cue enhancement for music perception remain largely unexplored. This paper introduces a method to increase spatial cue salience through the augmentation of individual head-related transfer functions (HRTFs). Auditory model analyses indicate that augmented HRTFs may enhance the separability of musical instruments relative to individual HRTFs. Predicted benefits persist when moderate sensorineural hearing loss is modelled, though they are substantially reduced. Simulated hearing aid processing does not restore these benefits to normal-hearing levels.
Read the full abstract

All authors Jack Webb, Christophe Lesimple, Volker Kuehnel, Lorenzo Picinali

Artificial Intelligence

Program Learning with Verifiable Rewards: Symbolic Backpropagation for Post-Training LLMs

Post training a language model to reason means updating its weights. Supervised finetuning and reinforcement learning both place the acquired capability inside the model where it cannot be inspected cannot be checked step by step and cannot be moved to another model. We argue that for tasks whose intermediate steps admit verification, reasoning is better placed outside the base models weights as an explicit program composed from deterministic and neural primitives. We introduce PLVR (Program Learning with Verifiable Rewards): a post training method that learns such programs directly from input-output examples. Its mechanism is symbolic backpropagation: each program layer carries a typed ontology a loss is computed at the output against ground truth and required input ontologies are propagated backward by type inference over primitive signatures: an analogue of the chain rule in which credit assignment is a derivation rather than an estimate. Where RLVR verifies a terminal outcome, PLVRs reward is a per step contract verdict dense over program structure. On LiveCodeBench v6 and Tau2Bench, 30B base models with PLVR outperform RL at matched budget by 27.8 points on average and frontier models an order of magnitude larger by 13.6 points. A single primitive library serves two benchmarks, so the marginal cost of a new task is 100 examples of program search and no new finetuning data. Replacing the loss guided search with uniform sampling over the same type admissible space at equal budget collapses the median program from 65.6 to 17.5, identifying the backward pass rather than the type system as the source of the advantage. We release the symbolic backpropagation library and a conformance checker so the method can be applied to primitive libraries other than our own.
Read the full abstract

Human-Computer Interaction

Between Algorithm (AI) and Intuition (Human): Preserving Designer Agency in AI-Assisted Sensemaking of Qualitative UX Data

The integration of AI into qualitative design research presents a fundamental tension: how do we leverage AI while preserving the subjective, intuitive judgments that define design expertise? This paper examines this question through a case study of analyzing 20 user responses about video conferencing platforms for educational contexts. We argue that AI sensemaking tools risk flattening the rich data patterns, amplifying contradictory textures of user feedback into sterile categories thereby transforming design research from an interpretive craft into a mechanical sorting exercise (rigid and formal). Through comparative analysis of AI-assisted sensemaking versus human-centered approaches to the same dataset, we identify when algorithmic efficiency enhances understanding and when it diminishes the designer's interpretive agency (uncovering hidden needs, critical enquiry, what if enquiries, making decisions, having trade-offs). We present a framework for augmented sensemaking that positions AI as an instrument for amplifying human judgment rather than replacing it. Our findings suggest that the most valuable role for AI in design research is not to eliminate subjectivity, but to make it more intentional, reflective, and accountable.
Read the full abstract