Peer Revenue Returns subject to peer review

Papers

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

All of arXiv

Tuesday 1 September

Performance

DART: Aiming for Tail-Delay Control in Reconfigurable Networks

Many systems serve different job classes by switching among configurations. Often, reconfiguration takes a stochastic amount of time that depends on direction and can differ in distribution between configurations. Reaching a target configuration may require crossing intermediate ones. Under a high-percentile delay objective, rarely served or pass-through configurations accumulate the long-waiting jobs that dominate the sojourn-time tail. Controlling that tail couples three decisions: which target to choose, which path to take, and whether and how many jobs to serve at the configurations crossed on the way. We introduce DART, an opportunistic commitment policy that commits to a target configuration to avoid myopic detours while serving a traversed configuration only when its weighted delay or backlog justifies the pause. We show how these decisions shape the weighted sojourn-time tail. Across six stress topologies, each isolating a different tail-control challenge, DART achieves the lowest P99 sojourn time in every case, up to 23% below the strongest baseline. The advantage holds under heterogeneous, heavy-tailed, and asymmetric reconfiguration times.
Read the full abstract

Information Theory

List Size of Random Linear Codes: Optimality and the Limits of Average-Radius List Decoding

Guruswami, Li, Mosheiff, Resch, Silas and Wootters conjectured that, for every fixed prime power $q$ and $p\in(0,1-1/q)$, random linear codes of rate $1-h_q(p)-\varepsilon$ have list size $\frac{h_q(p)}{\varepsilon}\cdot(1+o(1))$ as $\varepsilon\to0$, both for list decoding and average-radius list decoding. The conjecture was known for $q=2$ and remained open for larger alphabets. We prove the conjecture for list decoding for every prime power $q$, with the sharper bound $L=\frac{h_q(p)}{\varepsilon}+O_{p,q}(1).$ In contrast, we disprove the conjecture for average-radius list decoding. We show that the required list-size leading constant is at least $p\left(1+\log_q\frac{(q-1)(1-p)}{p}\right)$, which is strictly larger than $h_q(p)$ whenever $q^p(1-p)>1$. Moreover, for $p=1-\lambda/q$ with fixed $\lambda>1$, the required leading constant can be as large as $\left(\frac{\log\lambda}{e\lambda}+o_q(1)\right)\frac{q}{\log q}.$
Read the full abstract

Materials Science

Exchange striction determines how fast antiferromagnetic insulators demagnetize

Antiferromagnets combine terahertz spin dynamics with insensitivity to stray fields, and how quickly their order can be manipulated sets the speed limit on device operation. Femtosecond optical pulses demagnetize antiferromagnetic insulators on timescales that span picoseconds to nanoseconds across compounds, and no material parameter is known that accounts for the spread or predicts where a new compound will fall. In a compensated antiferromagnet, no angular momentum needs to leave the spin system, so the rate is set by energy flow from the lattice into the spins. Time-resolved second-harmonic generation experiments show that Cr2O3 demagnetizes within 2 ps once the lattice is driven above the Neel temperature, two orders of magnitude faster than the structurally similar FeBO3. First-principles calculations trace the disparity to exchange striction: short Cr-Cr contacts make the exchange coupling tenfold more sensitive to atomic displacements and widen the phase space for phonon decay into magnon pairs. Spin-lattice simulations with ab initio parameters reproduce the order of magnitude of the measured ratio. The derivative of the exchange coupling with respect to the ionic displacement thus emerges as a computable parameter that predicts how fast an insulating antiferromagnet can be demagnetized. The results advance our understanding of ultrafast control in insulating antiferromagnets, and suggest a practical pathway to screen candidate materials for thermally assisted antiferromagnetic memory before synthesis.
Read the full abstract

All authors Aleksandr Buzdakov, Ravi Kaushik, Nikolai Khokhlov, Sergey Artyukhin, Alexey Kimel

Computation and Language

OUTLETS: Output-Length Prediction from Speculative Decoding Backbones

The heavy-tailed distribution of output lengths in Large Language Model (LLM) serving poses major challenges for resource provisioning and cluster scheduling. Although output-length prediction can mitigate these issues, existing approaches have key drawbacks: external proxy models add substantial latency and often have limited fidelity, whereas internal state-based methods are efficient but rely on shallow probes of current model states. We identify a structural connection between speculative decoding (SD) and length prediction: latent representations produced by the draft decoder in advanced frameworks (e.g., EAGLE-3) encode signals that are predictive of generation length. Building on this insight, we introduce OUTLETS (Output-Length Prediction from Speculative Decoding Backbones), which repurposes the speculative backbone as a trajectory-aware length predictor. When its draft representations are already computed for speculative decoding, OUTLETS adds only a lightweight regression head and achieves lower MAE than the evaluated methods. Under saturated disaggregated serving, OUTLETS predictions enable standard scheduling policies to prioritize shorter requests and distribute requests more evenly across decoding instances, reducing short-request P99 latency by 34.8%.
Read the full abstract

All authors Weihuang Wen, Yingying Liu, Yichuan Liu, Wenqi Zeng, Li Zhou, Chumin Sun, Jie Sun, Tianshu Yu

Information Retrieval

World Model-Guided Reinforcement Learning via Counterfactual User Engagement Simulation

Reinforcement learning for user-centric agents is limited by the cost, latency, and risk of collecting online feedback, as well as by the lack of counterfactual comparisons under the same user state. In this paper, we propose World Model-Guided Reinforcement Learning via counterfactual user engagement simulation (WMG-RL), a framework in which a frozen user simulator provides reward supervision before real user exposure. Motivated by language world models, we instantiate the simulator as a User Engagement World Model (UEWM), which treats a recommended item as the agent action and the user's heterogeneous feedback as the environment observation. Rather than learning one fixed environment transition, UEWM learns to infer user-specific dynamics from engagement history and apply them to candidate items. In WMG-RL, a downstream policy proposes multiple candidate items for the same history; UEWM predicts the corresponding engagement feedback in parallel; and the simulated feedback is converted into dense rewards for policy optimization. Experiments show that UEWM provides reliable and transferable reward signals across domains, and that WMG-RL enables a compact 1.7B student policy to match or surpass much larger LLMs on downstream recommendation tasks.
Read the full abstract

All authors Ang Li, Xin Xu, Bin Liang, Yue Ma, Fubang Zhao, Yangyang Kang, Kam-Fai Wong

Quantum Physics

Scattering and Tunneling in Real Quantum Mechanics

We establish the general form of the evolution equation and study scattering and tunneling in real quantum mechanics. We prove an analogue of Stone's theorem for a real Kahler space and derive the corresponding general evolution equation, whose evolution operator is both orthogonal and symplectic. We formulate scattering theory in terms of real wave operators and the associated S-matrix. For the class of potentials considered, we show that the differential scattering cross section is equivalent to that obtained in standard complex quantum mechanics. We also show that the tunneling probability in real quantum mechanics is identical to that in complex quantum mechanics.
Read the full abstract

Astrophysics of Galaxies

Investigating the Young Stellar Populations and Hierarchies in Nearby Galaxies with the UVIT. III. Evidence for a Largest Scale of Correlated Stellar Structures and a Non-universal Fractal Dimension

Scale-free turbulent motions, gravitational collapse and galactic dynamics govern galactic-scale, hierarchical organization of star formation (SF) within galaxies. Past studies suggest that properties of SF hierarchies depend upon host galaxy properties and interstellar medium (ISM) conditions. To characterize SF hierarchies, we performed two-point correlation function analysis on ~25000 UV-selected star-forming clumps (SFCs) identified in a morphologically diverse sample of 8 classic spirals, 6 flocculent spirals and 3 dwarf irregulars. We found that SF hierarchies in galaxies exhibit a maximum spatial scale – the correlation length ($l_{\rm corr}$) – largest scale up to which SF is spatially correlated, presumably owing to ISM turbulence. The $l_{\rm corr}$ values range from ~100 pc to 3.4 kpc and exhibit strong dependence on the galaxy's stellar mass, morphology and nature of spiral arms. This suggests that a galaxy's gravitational potential and spiral structure place an upper limit on the sizes of the largest, hierarchically structured SF complexes. Connecting $l_{\rm corr}$ with turbulence injection sources suggests that stellar feedback in dwarf irregulars, whereas disk instabilities and spiral structure in classic/flocculent spirals dominate towards sustaining their SF hierarchies up to the $l_{\rm corr}$ scale. These hierarchies disperse to near-random distributions on timescales ($T_{\rm dis}$) ranging from 20-160 Myr. The broad range of derived $l_{\rm corr}$, projected fractal dimension ($D_2$ $\in$ 0.71$-$1.73), and $T_{\rm dis}$ indicates a non-universal, galaxy-specific nature of SF hierarchies. In this work, full coverage of each galaxy's star-forming extent with the AstroSat-UltraViolet Imaging Telescope uniquely enabled us to connect global parameters of SF hierarchies with large-scale galaxy properties.
Read the full abstract

All authors Gairola Shashank, Smitha Subramanian, Shyam H. Menon, Chayan Mondal, Kathryn Grasha, Richard de Grijs, Prasun Dutta, Annapurni Subramaniam

Sound

Artificial Rosetta Stone: Constrained Maximum A Posteriori (MAP) Reconstruction of Symbolic Raga Sequences via Order-k Markov Models

Reconstructing a damaged musical fragment is an inverse problem: the observed sequence contains partial information, while a raga encodes constraints limiting allowable completions. This paper formalizes a mathematical framework for this, proposing the Artificial Rosetta Stone (ARS). We separate three claims often conflated: a symbolic sequence can be reconstructed probabilistically; a sequence can be consistent with an explicit grammar; and a historical performance can be authenticated. We only support the first two. We model a raga via a finite alphabet and constraint system, using an order-k Markov model for melodic probabilities. A symmetric Dirichlet prior yields a tractable posterior. We pose missing-note reconstruction as a constrained MAP problem. For fixed-length sequences and finite-order constraints, optimization admits an exact dynamic-programming solution with worst-case time complexity $O(TN^{k+1})$. We derive the parameter count $N^k(N - 1)$, prove a concentration bound under explicit mixing assumptions, and analyze estimation error propagation. A reproducible synthetic experiment uses six raga-inspired alphabets, orders $k \in \{1, 2, 3\}$, and masking rates up to 50%. This is a proof of concept, not historical reconstruction. A real-audio feasibility pilot evaluates 30 usable sequences from 42 Yaman clips via automated pitch extraction, segmentation, and quantization. Lacking documented provenance and relying on automated transcription, this is not expert-validated archival reconstruction. Claims are tied to stated conditions, not universal properties of Hindustani music. Code: https://github.com/mathacker23/ArtificialRosettaStone.
Read the full abstract

General Relativity and Quantum Cosmology

Colored-Noise-Induced Horizon Fluctuations Near a Double Root in an Effective Black-Hole Geometry

Stress-tensor fluctuations carry information that is absent from the mean semiclassical Einstein equation, but their observable effect depends as much on the response of the geometry as on the noise itself. We formulate an effective, quasistationary spherical model organized by the response–noise structure of a quantum Langevin equation. A prescribed mean dressing and a horizon-localized colored source are varied independently. For every source realization we reconstruct the lapse, locate its outer trapping horizon, and evaluate the adiabatic Hayward–Kodama temperature from the slope at that same root. As the mean lapse approaches an inner–outer root merger, its inverse slope acts as a geometric susceptibility: a fixed source covariance produces enhanced horizon fluctuations, horizon–temperature covariance, and a positively skewed, greybody-filtered luminosity. The specific advance is a single-realization map from colored metric noise to correlated geometric, thermal, and radiative statistics, together with an explicit separation of reservoir strength from near-critical amplification. The model is a stochastic-semiclassical testbed, not a microscopic evaluation of Unruh-state response and noise kernels.
Read the full abstract

Artificial Intelligence

Space Generative AI with Solar Energy Harvesting

Satellites are emerging as promising platforms to extend generative artificial intelligence (AI) services to remote areas lacking terrestrial infrastructure. However, deploying space generative AI is fundamentally constrained by the limited, time-varying onboard energy supplied by solar energy harvesting (EH). This paper presents a framework for solar-powered space generative AI in which a satellite receives a user prompt, executes a diffusion-based image-generation model, and downlinks the compressed result within a strict time window. We identify the fundamental computation–communication (C$^2$) trade-offs governed by the shared harvested-energy budgets. Specifically, increasing the number of generation steps improves intrinsic image quality but depletes energy and time available for downlink transmission, whereas prioritizing communication guarantees reliable delivery but sacrifices semantic quality. To balance these trade-offs and maximize end-to-end (E2E) generative performance, we exploit the predictable solar-EH dynamics induced by deterministic orbital motion and develop a joint C$^2$ resource-optimization framework using a tractable two-step approach. First, we characterize the maximum downlink throughput for a fixed generation depth under continuous solar EH. This establishes a separation principle that decouples waiting-time selection from optimal transmit-power control. Next, we formulate a joint C$^2$ utility-maximization problem and derive a closed-form, low-complexity step-selection policy in the dominant constant-power regime. Extensive experiments under realistic orbital dynamics demonstrate that the proposed policy dynamically balances generation quality and transmission reliability. This yields significant E2E performance gains over static computation- and communication-centric baselines across diverse solar-EH states.
Read the full abstract

All authors Jierui Zhang, Jianhao Huang, Zhanwei Wang, Kaibin Huang

Robotics

Accelerating Reinforcement Learning via MPC Solver-Gradient Guidance for Weights-varying MPC

In Model Predictive Control (MPC), cost-function weights shape closed-loop behavior, yet changing conditions often make fixed parametrizations suboptimal and motivate context-dependent online adaptation. Learning such policies is difficult because behavior depends implicitly on numerical MPC solutions, producing nonlinear, potentially nonsmooth, long-horizon dependencies on policy parameters. This creates a bias-variance tradeoff: Reinforcement Learning (RL) optimizes realized closed-loop return from environment samples but is sample-inefficient, whereas Gradient-Based Policy Learning (GB-PL) uses low-variance solver gradients from differentiable MPC to optimize surrogate losses on predicted trajectories but can be biased under model mismatch. We propose Solver-Gradient Guided Reinforcement Learning (SG-RL), a solver-sensitivity augmentation for RL-based online MPC cost-weight adaptation. SG-RL keeps sampled closed-loop return as the objective and uses bounded solver-derived gradients as auxiliary guidance to improve stability and sample efficiency. We instantiate SG-RL in Proximal Policy Optimization (PPO) with four modular algorithms that inject solver-gradient guidance into actor-update scaling, policy loss, advantage estimation, and value-function learning. On two full-scale autonomous racing platforms with intentional model mismatch, SG-RL reaches PPO's best closed-loop return with up to 70.6% fewer samples, outperforms GB-PL baselines by at least 54% in closed-loop return, and generalizes zero-shot to unseen environments.
Read the full abstract

All authors Baha Zarrouki, Arslan Thobani, Jasper Hoffmann, Mattia Piccinini, Rudolf Reiter, Felix Jahncke, Sébastien Gros, Davide Scaramuzza, Johannes Betz

Image and Video Processing

Lightweight Interpretable RGB-Guided Hyperspectral Super-Resolution under Real Cross-resolution Misalignment

Compact snapshot hyperspectral cameras provide rich instantaneous spectral measurements for ground-level machine vision, but at lower spatial resolution than standard RGB cameras. RGB-guided hyperspectral super-resolution (HSR) addresses this limitation by transferring spatial detail from a high-resolution RGB guide to a low-resolution hyperspectral image (HSI). These dual-camera systems are typically in a horizontal rig geometry, requiring cross-camera image alignment due to different fields of view. However, residual misregistration can inject spurious high-frequency details. Existing learned unaligned-fusion methods are usually trained for a fixed spectral support and spatial scale factors and can be computationally demanding, limiting their flexibility across sensors. We propose a lightweight and interpretable RGB-guided HSR framework combining cross-modal flow alignment with model-based Gram-Schmidt orthogonalization fusion. The method first warps the RGB guide onto the HSI grid, then estimates an energy-based confidence weight map by measuring local alignment reliability. This map is then used both in a weighted least-squares spectral regression and in a gated fusion between the super-resolved estimate and an HSI-preserving estimate. Unlike existing learned methods, the proposed framework has a low computational footprint and supports VIS-NIR spectral supports and scale factors without retraining. Experiments on the Real benchmark show that the proposed method improves reconstruction accuracy over learned fusion baselines while remaining substantially faster. On a 34-frame sequence acquired with our real RGB-HSI dual-camera setup, a reduced-resolution quantitative evaluation validates the method under genuine cross-sensor radiometric, noise, and geometric differences, while native-resolution qualitative results demonstrate deployment on the full 51-band VIS-NIR acquisition.
Read the full abstract

Computer Vision

Dyn-3D: Unveiling and Resolving Ego-Motion Ambiguity in Vision-Language Models

As Vision-Language Models (VLMs) tackle dynamic 3D spatial reasoning, ego-motion perception becomes essential to resolve monocular scale ambiguity. However, current models often overfit to smooth trajectory priors rather than genuinely understanding physical motion. Consequently, their spatial reasoning degrades severely under large displacements, a phenomenon we term Kinematic Collapse. This failure stems from spurious visual-motion correlations in natural videos and a lack of explicit physical supervision. To evaluate this, we introduce Dyn-3D, a benchmark using counterfactual 3D rendering to rigorously decouple visual changes from true kinematic properties. Furthermore, we propose the TempoVista framework, featuring the Kinematic-GSPO algorithm. By embedding metric physical ground truth into policy optimization, TempoVista explicitly grounds visual representations in 3D space. Experiments demonstrate that our approach significantly improves both motion estimation and robust spatial reasoning by utilizing camera dynamics as an effective geometric calibration signal.
Read the full abstract

All authors Jiayu Ding, Zhuodong Liu, Lei Zhang, Manyu Xiong, Hongbo Jin, Haoran Tang, Hongbo Zhang, Changen Zhu, Wenbo Xing

Artificial Intelligence

ARISE-RL: Agentic Rubric-Grounded Iterative Self-Evolution with Reinforcement Learning

Training open-ended agents via reinforcement learning (RL) is hindered by the lack of verifiable gold answers and scalable rubrics. Moreover, even near the model's capability boundary, long-horizon open-ended agentic tasks often yield brittle and unstable rewards, resulting in weak or noisy rollout contrast that obscures fine-grained optimization signals for group-based policy learning. To address these challenges, we propose ARISE-RL, a novel full-cycle self-evolution framework that couples a task/rubric Generator and a reasoning Solver through rubric-mediated co-evolution. The Generator grounds tool-related rubric criteria in real tool observations and is rewarded for producing valid, intermediate-difficulty tasks aligned with the Solver's evolving capability boundary. The Solver, in turn, learns from fine-grained rubric satisfaction signals through multi-step reasoning and tool use. We further introduce Reward-Gated Self-Evolution Distillation (RG-SED), which selectively distills a memory-augmented variant of the same policy back into itself only when the memory yields empirical reward improvement, thereby reducing distribution mismatch and avoiding blind imitation of noisy guidance. Finally, to support rigorous evaluation, we present ECR-Bench, an expert-calibrated rubric benchmark suite covering single-tool deep research and multi-tool travel planning. Extensive experiments demonstrate that ARISE-RL consistently achieves robust and stable overall state-of-the-art performance across all evaluated benchmarks.
Read the full abstract

All authors Fanrui Zhang, Ruixue Ding, Qiang Zhang, Xi Chen, Boli Chen, Shihang Wang, Qiuchen Wang, Hongmin Zhan, Jinxin Bian, Li xingchao, Peijin Zheng, Hao cheng, Pengjun Xie, Kaipeng Zhang, Jiawei Liu, Zheng-Jun Zha

Artificial Intelligence

User Representation via Cross Multi-source Behavior Pre-training for Mobile Games

User representation pre-training has become a fundamental paradigm for alleviating data sparsity in downstream personalization tasks. However, existing studies predominantly focus on single-app or app-level behaviors, overlooking the inherently cross-source and multi-granular nature of user activities on mobile devices. At the device level, user intent emerges from complex interactions among heterogeneous behavior sources and hierarchical action structures, posing challenges that cannot be addressed by conventional app-centric modeling. To tackle this issue, we propose CM-PTM, a novel Cross Multi-source Behavior Pre-Training Model tailored for mobile game user representation learning on device-level behavioral logs. CM-PTM employs hierarchical cascaded mask-then-predict proxy tasks that first infer the source of the next behavior and then progressively refine predictions at the app-action level. This design enables unified modeling of cross-source dependencies and fine-grained behavioral dynamics within a single pre-training paradigm. Extensive experiments on large-scale real-world mobile datasets demonstrate that CM-PTM effectively captures users' endogenous interests and consistently delivers significant performance gains on downstream mobile game recommendation tasks.
Read the full abstract

All authors Chengqi Yang, Yiran Qiao, Feng Liu, Xingyu Lou, Zijun Zhou, Xiaoyun Mo, Changwang Zhang, Jiayuan Xu, Jun Wang, Xiang Ao

Artificial Intelligence

WorldBench: Culturally Grounded Benchmark for Multilingual Agents

Despite the growing use of LLM-powered agents to solve multi-step tasks in complex environments, existing benchmarks rarely test state preservation, performance across languages, and application to realistic, grounded scenarios. To address these concerns, we present WorldBench: a comprehensive, multilingual benchmark of genuine, persona-grounded everyday workflows, where agents can act in a sandbox via structured actions. WorldBench comprises 1,600 tasks across seven languages and eight cultures, filtered and refined through feedback from human annotators with language- and culture-specific expertise. For evaluation, we extend metrics from previous works and introduce Constrained Task Success (CTS), which combines natural language instructions and testbeds to score task completion, minimal modification, and other complementary metrics through deterministic and LLM-as-a-Judge evaluations. Our experiments show that frontier models reach only 49.2% CTS, with all models demonstrating large gaps between correctness and environment preservation. We thereby show that current agents remain brittle in multilingual, agentic scenarios, especially for long-horizon tasks and under state-preservation constraints
Read the full abstract

All authors Leonardo Ranaldi, Sherrie Shen, Jushi Kai, Alexandra Birch

Neurons and Cognition

Active Visual Semantics: A large-scale MEG and eye-tracking dataset for understanding visual intelligence in action

Here we present the Active Visual Semantics (AVS) dataset, a large-scale collection of magnetoencephalography (MEG) and eye-tracking data recorded while five participants freely explored 4,080 natural scenes over 10 sessions each, yielding more than 200,000 fixation epochs in total. Unlike existing neuroimaging datasets that rely on passive viewing with enforced central fixation, AVS captures brain activity during active scene exploration, including self-generated saccades and fixations. A semantic captioning task on 25% of the trials provides behavioural measures linking gaze to scene understanding and memory. In addition to neural and behavioural data, AVS includes per-fixation object category labels, human-rated annotations of the appearance of fixation targets in the scene captioning task and pupil dynamics. Individual head stabilisation casts were used during MEG data collection, which alongside with structural MRI scans, enabled precise cross-session source reconstruction. Using artificial neural network (ANN) encoding models we demonstrate that individual fixation-aligned MEG epochs hold visual content-specific signal, despite the challenges that active scene viewing poses for MEG signal quality. Further, we use fixation-aligned representation similarity analysis (RSA) and demonstrate that we can derive fixation object category averages that yield representational geometries which are highly reliable across participants. Both in MEG sensor and source space this structure is validated by its robust alignment with ANN object-level representational geometry. Taken together, AVS provides a rich resource for investigating a large variety of questions regarding the neural mechanisms of active vision, object recognition and scene captioning during natural viewing, and the relationship between gaze behaviour and memory encoding.
Read the full abstract

All authors Philip Sulewski, Carmen Amme, Peter König, Martin N. Hebart, Tim C. Kietzmann

General Relativity and Quantum Cosmology

High-Frequency Gravitational-Wave Transduction in a SQUID-Terminated Superconducting Cavity

High-frequency gravitational waves in the MHz–GHz range require detection strategies beyond conventional interferometers. We study the response of a SQUID-terminated superconducting microwave cavity as a narrowband parametric transducer. The cavity boundary conditions and the flux dependence of the Josephson inductance are used to derive the quarter-wave mode spectrum and the first-order eigenfrequency response to changes in the physical line length, phase velocity, and SQUID inductive length. By projecting the perturbed dynamics onto the static cavity modes, we show that the resonance-frequency response and the photon-pair-production response are generally distinct. We therefore introduce two independent kernels, $R_n^{\omega}$ and $R_n^{\rm pair}$, which coincide only when changes in mode normalization and spatial profiles can be neglected. Near $\Omega_{\rm GW}\simeq 2\omega_n$, the isolated-mode dynamics reduce to a Mathieu-type parametric amplifier, allowing the photon number, quadrature variances, gain, and instability threshold to be obtained in the presence of dissipation. Spontaneous photon production scales quadratically with the gravitational-wave strain and is extremely small for representative parameters, whereas phase-sensitive responses with a coherent probe can scale linearly with strain. The framework therefore provides a theoretical description of narrowband gravitational-wave transduction, while a quantitative sensitivity estimate requires device-specific calibration of the mechanical–electromagnetic response, verification of mode isolation, and a complete treatment of loss and noise.
Read the full abstract

Materials Science

First-Principles Electronic Structure Calculation of Crystals in Laboratory Magnetic Fields

External magnetic fields can qualitatively reshape the electronic structure of crystals, underpinning quantum Hall physics, Landau-level spectra and field-induced topological phases. Their first-principles treatment at laboratory-scale fields is, however, hindered by magnetic-flux quantization, which requires magnetic unit cells with areas inversely proportional to the applied field. Such cells contain a large number of chemical unit cells, rendering real-space and plane-wave calculations prohibitively expensive. Here we, for the first time, construct a magnetic Bloch basis built from linear combinations of gauge-including Gaussian-type atomic orbitals, which incorporate the magnetic-field phase factors required by magnetic translation symmetry. The framework requires far fewer basis functions than real-space or plane-wave representations of the same magnetic supercell and retains the sparsity of an atom-centred basis, together substantially reducing computational cost. We validate the framework by reproducing Landau-level spectrum of graphene from first principles. This approach provides a practical route to simulations of crystalline materials under experimentally accessible magnetic fields.
Read the full abstract

All authors Sichao Wang, Chengye Lü, Xingao Gong, Yingwei Chen, Hongjun Xiang

Solar and Stellar Astrophysics

Archival transients detected with the MeerLICHT telescope I: Supernovae

Wide-field optical transient surveys collectively discover $\sim$50 new transients per night, making rapid, multi-filter and/or spectroscopic follow-up of all but the most exceptional events unfeasible. Consequently, the (early) light-curve properties of most supernovae (SNe), particularly sub-luminous events, are often poorly characterised. Here, we showcase recovery of key parameters by combining publicly available data from multiple, contemporaneous surveys. We built a catalogue of $125$ SNe by cross-matching the MeerLICHT (ML) transient survey with the Transient Name Server list of spectroscopically classified SNe detected between August 2017 and October 2022. For a subset of $14$ SNe with good multi-survey photometric coverage, we combined ML-$q$, ZTF-$g,r$, and ATLAS-$c,o$ data to construct composite early-time light curves. We then applied an expanding fireball ($t^2$) model to constrain their explosion dates and rest-frame rise times. We find that this simple model provides a reliable description for the early emission of most SNe in our sample and successfully constrains their rise time, a key parameter for characterising SN categories and probing progenitor origins. This multi-survey approach, only using standard survey products, provides a powerful method to characterise large numbers of SNe that would otherwise be lost, a technique that will be essential for maximising the scientific return from current and future surveys such as GOTO, BlackGEM, the Vera C. Rubin Observatory's LSST, and Global Open Transient Telescope Array (GOTTA).
Read the full abstract

All authors O.Mogawana, P.J. Groot, N. Blagorodnova, P.M. Vreeswijk, D.L.A. Pieterse, S. Bloemen, G. Leloudas, R.A.D. Wijnands, D. A. H. Buckley, P. A. Woudt, R. Fender, B. W. Stappers

Machine Learning

SAGE: Subpopulation-Aware Generative Enhancement for Mitigating Spurious Correlations

Spurious correlations pose a significant challenge to the robustness of modern machine learning. The inherent imbalance in dataset distributions often leads traditional Empirical Risk Minimization (ERM) models to rely on majority spurious attributes for classification, resulting in poor performance on minority groups. This problem becomes particularly challenging when the spurious attributes are unavailable. Existing group-label-free methods often upsample minority groups or misclassified real training examples; repeating the same instances can reduce effective diversity and encourage overfitting. To mitigate these spurious correlations from a data-centric perspective in the absence of prior knowledge, we introduce Subpopulation-Aware Generative Enhancement (SAGE), a two-stage generative augmentation framework. Using cluster-derived sub-labels and class labels, we fine-tune a conditional generative model and text encoder, generating targeted synthetic data to fill underrepresented regions in the training set and construct a balanced validation set for last-layer reweighting. We experimentally show that SAGE achieves 89.5%, 85.7%, and 79.1% worst-group accuracy on Waterbirds, CelebA, and MetaShift, respectively, outperforming the best group-label-free baselines by up to 7.7 percentage points.
Read the full abstract

Cryptography and Security

What Limits Robustness in Deep Image Watermarking: An Analysis of Mechanisms and Their Scaling Across Capacities

Robustness remains the principal open problem in deep image watermarking, and what limits it becomes sharper as payload grows. This paper asks whether capacity is itself the limit or only makes other limits visible, and answers in two parts. The first organizes the distortions a watermark must survive and the strategies developed to resist them, ordering each by the axis that governs it: payload capacity for the distortions, differentiability for the strategies. The second identifies and measures three mechanisms that limit robustness in schemes mapping the payload onto a spatial block grid with extraction trained separately from a frozen embedder: desynchronization of the payload grid, the resistance of codec-induced distortion to training, and the narrowing of the usable embedding-strength window. Payloads from 64 to 16384 bits are measured, well beyond the range those strategies address. Training the extraction stage against a codec proves not merely ineffective but harmful, degrading the reading at the operating points used in training. The limits follow the class of distortion rather than capacity itself, and none is removed by further training on the extraction side, because all three arise before extraction. An evaluation protocol making claims of generalization verifiable is also contributed. The conclusions are properties of a class of designs rather than of one implementation.
Read the full abstract

Artificial Intelligence

QILP-0: Constructing Observational Declarative Twins of Quantum Circuits

This paper introduces QXymb, a general framework for constructing observational declarative twins of quantum circuits, and develops QILP-0, its first complete order-0 specialization. QILP-0 constructs a finite multi-valued propositional logic program from observed circuit behaviour within a declared observational scope. The pipeline traverses a declared family of quantum observables incrementally according to a reproducible structural grading and a declared observational reference horizon. Progress is quantified through reference-relative coverage against a fixed target-independent reference. Observable responses are organized through target-independent geometry, while retained latent structure is mapped deterministically back to original observable columns before symbolic processing, preserving observational semantics and provenance. Selected observable profiles are converted into a finite relation through admissible target-independent discretization. The target is used only afterwards to audit twin-admissibility and induce the declarative theory. A theory is certified as an exact observational declarative twin when it completely and correctly reconstructs the resulting finite task-conditioned discrete relation. Logical exactness is therefore separated from numerical, backend, provider, and discretization uncertainty, which is retained as audit metadata. Validation uses two complementary QML settings. Exhaustive Bars & Stripes experiments compare product and grid-CZ embeddings from 16 to 100 qubits and exercise the native-discrete branch. Low-Depth MNIST analyses all 14,708 digit-0/1 instances before and after a trained variational quantum transformation and exercises continuous discretization. In every reported relation, the induced QILP-0 theory achieves complete, conflict-free reconstruction with strict accuracy equal to one.
Read the full abstract

All authors Marina de la Cruz Echeandía, César Luis Alonso, Tony Ribeiro, Alfonso Ortega de la Puente

Computation and Language

Lagged Coupling: Internal Representations Become Readable Before They Become Causal

Across the full Pythia suite (160M-12B, eight checkpoints, four task families), a linear probe can read a target variable from the residual stream as early as step 1,000 at every scale – yet steering along that same reading direction remains null-equivalent in 43 of 48 model-checkpoint cells. Internal readability systematically outruns causal efficacy, and the lag does not shrink with scale. We call this structure lagged coupling and decompose it into three dissociable tracks: (i) internal readability, saturated (AUROC >= 0.990) from the first checkpoint everywhere; (ii) behavioral readability, which develops gradually and progressively later at larger scales (12B reaches 0.909 only at the final checkpoint); (iii) causal efficacy, almost always null-equivalent, occasionally counterproductive early, with one isolated positive pulse (12B, step 8,000, z = +2.49) our grid cannot resolve. The ordering is dominantly read-before-write (11/11 units, no inversion). Representation headroom along the probe direction grows up to 57x with training and scale while causal write-in stays below 0.11% of headroom – the variable is increasingly written into the representation and increasingly ignored by the readout. Under a fully pre-registered protocol, both single-onset hypotheses resolve INDETERMINATE (scale slope +0.24, 95% CI [-0.60, +0.87]; time vote 3:3) – a disciplined negative explained by the three-track decomposition. A pre-registered OLMo-2 replication preserves the direction at attenuated magnitude. Our results caution against inferring steerability from probe accuracy and establish a developmental bottleneck: representation formation reliably outpaces causal readout consolidation.
Read the full abstract

General Relativity and Quantum Cosmology

Ultraviolet Closure in Asymptotically Weyl-Invariant Gravity

We investigate asymptotically Weyl-invariant gravity (AWIG) in the Palatini formulation, defined by an exponent n that interpolates between an Einstein-like infrared regime, where n goes to one, and a Weyl-invariant ultraviolet regime, where n goes to two. In four dimensions, we show that exact Weyl invariance within the minimal scalar Palatini f(R) sector uniquely selects the R^2 action as the ultraviolet endpoint. Interpreting the exponent as an effective curvature-dependent quantity, we constrain a phenomenological class of smooth autonomous flows compatible with the required infrared and ultraviolet endpoints. We then analyse the strict ultraviolet theory. On the regular branch R not equal to zero, and within a restricted torsionless, parity-even sector constructed purely from curvature, all admissible divergent on-shell counterterms reduce at arbitrary loop order to the original \sqrt{-g}R^2 term plus a topological Euler density. Although this conditional result does not establish full off-shell perturbative renormalizability, it provides nontrivial evidence for improved ultraviolet behaviour relative to Einstein gravity and identifies a concrete target for a future first-principles renormalization-group calculation.
Read the full abstract

Cryptography and Security

HiveTraceGuard-Pro: A Compact Generative Guardrail for Prompt Injection, Jailbreaks, and Adversarial Obfuscation

Production LLMs must handle inputs that attempt to override system instructions, bypass safety policies or elicit harmful responses. A common mitigation is a separate guardrail model. Existing reports, however, provide little evidence on Russian prompt injection or Russian surface obfuscation. We present HiveTraceGuard-Pro, a 0.6B generative guardrail LoRA-tuned from Qwen3-0.6B. It is trained on Russian and English and uses one binary scoring rule (safe/unsafe) for the final target turn. Its training corpus pairs harmful examples, where a counterpart exists, with benign examples from the same domain and applies eight obfuscation transforms to both labels. In one harness, we compare HiveTraceGuard-Pro with thirty-four other guards on nineteen benchmark groups, sixteen of which are public. Its aggregate key is 0.7432, behind 0.7641 and 0.7552 for the two higher-scoring guards. Over the sixteen public groups alone, its key is 0.7153 and four of the thirty-four other suite guards score higher. In a fifteen-model comparison, HiveTraceGuard-Pro has the highest clean Russian robustness combined-F1 (0.88) and Russian prompt-injection recall (0.999). Both results use Russian sets assembled by our team, and at least 27.1% of the prompt-injection set overlaps the training corpus. Its 14.3 ms median latency is the lowest among those fifteen models in that run. Across the suite, FPR is 0.268 and FNR is 0.156. All reported response results use a legacy standalone-reply serialization rather than the natural assistant-role path of the shipped chat template. We release the merged weights on Hugging Face under Apache-2.0. The corpus, evaluation sets and evaluation code remain internal.
Read the full abstract

Artificial Intelligence

AgentFactory: Towards Automated Agentic System Design and Optimization

Large Language Models (LLMs) have demonstrated remarkable capabilities as powerful components in agentic systems, enabling sophisticated reasoning and complex task execution. However, current approaches to manually designing and optimizing agentic systems heavily rely on manual effort, limiting their adaptability and scalability. Recent work has explored the automated optimization of workflow designs. However, these approaches often overlook the crucial role of model capabilities and focus on single performance metrics, failing to address real-world deployment constraints. In this paper, we present AgentFactory, a framework that jointly optimizes both foundation models and workflow structures in agentic systems while considering multiple objectives including performance, cost, and efficiency. AgentFactory leverages advanced LLMs as optimizers to navigate the vast search space of possible configurations, employing a three-stage optimization pipeline to automatically discover effective combinations of fine-tuned models and optimized workflows. Through an iterative optimization process, our framework systematically explores and evaluates different agentic system designs, adapting to task-specific requirements while maintaining operational efficiency. We evaluate AgentFactory across eight benchmarks spanning five domains, including general reasoning, coding, mathematics, medicine, and finance. Our experiments demonstrate that AgentFactory consistently outperforms both manually designed methods and existing automated approaches, achieving an average improvement of 9.1% across all benchmarks, with particularly significant gains in domain-specific tasks (19.6% on MedQA and 18.7% on FinEval). These results establish AgentFactory as a promising approach for developing more capable and efficient agentic systems through automated optimization.
Read the full abstract

All authors Enci Zhang, Haofeng Wang, Yuesheng Zhu, Xiaole Cui, Guibo Luo

Soft Condensed Matter

Mass-conserving growth percolation in polymer gelation

Polymer gelation involves the emergence of a system-spanning network from growing polymer-rich domains, yet conventional percolation models typically prescribe particle size independently of material consumption. We formulate gelation as a locally mass-conserving growth-percolation process in which Voronoi capture zones define finite material reservoirs for individual nuclei. Local depletion determines the evolving supersaturation and limiting particle size, while particle contacts generate a dynamic network whose first spanning cluster defines the gel point. Three mechanisms, namely, interface-, diffusion-, and polymer-blob-controlled growth produce distinct gelation kinetics while sharing the same accessible final state. Spatial fluctuations in nucleation further generate a distribution of gelation times. The framework separates gelation from saturation and naturally captures post-gel aging as continued growth and topological maturation.
Read the full abstract

Machine Learning

From Truncation to Commitment: Persistent Context in Uniform Discrete Diffusion

Uniform-state discrete diffusion models update all tokens in parallel while keeping every position revisable. Even when the commonly used top-$p$ rule leaves only one candidate at a position, that choice affects only the current reverse step and can be revised at the next sampling step. We ask what changes when selected hypotheses instead become persistent context for later predictions. We therefore propose committed reveal sampling (CRS), a training-free sampler that stores selected argmax tokens and inserts them into subsequent model inputs. Our analysis separates persistent context from truncation. Within one reverse step, adaptive top-$p$, randomized or not, reproduces the update induced by a visible token only when a likelihood-ratio ordering holds, while in a simple latent-mode model, keeping the selected token visible helps later parallel predictions agree on the same sequence-level choice. Empirically, paired experiments on Duo-distilled then separate this persistent effect from single-step top-$p$ restriction and scalar temperature scaling. Under the same finalization rule, CRS without top-$p$ truncation reaches lower generative perplexity (GenPPL) than fixed $p=0.95$ and $p=0.9$ baselines across budgets of 8–64 function evaluations (NFE). At 64 NFE, the comparison at matched unigram entropy also gives lower GenPPL for CRS, shifting the GenPPL–entropy tradeoff under this metric. Base Duo shows the same direction in a descriptive comparison, while other diversity and continuation metrics can rank these operating points differently. A distributional distance to held-out text (SOL) under a fixed GPT-2-Large representation also decreases with persistent context at the same late top-$p$ threshold. These results identify support restriction and persistent context as distinct controls of that tradeoff.
Read the full abstract

Physics and Society

Behavioral calibration of mobile-phone GPS data for population-representative analyses

Mobile phone mobility data have transformed the study of human behavior, but demographic and behavioral biases can compromise their representativeness and distort population-level inference. Existing calibration approaches primarily address demographic and geographic representativeness, leaving behavioral discrepancies largely uncorrected. Here we introduce the Behavioral Population (BePop) framework, which jointly calibrates mobility data to representative demographic and behavioral distributions using census data and time-use surveys. BePop embeds mobility sequences into behavioral profiles and estimates person-level weights that align both population composition and daily activity patterns. Across three U.S. metropolitan areas, the framework consistently improves agreement between GPS-derived mobility and representative behavioral distributions, including time allocation, activity transitions, and mobility motifs. Calibration also substantially alters downstream mobility indicators, demonstrating that behavioral biases can propagate into commonly used mobility measures. Our results establish behavioral representativeness as a critical complement to demographic calibration and provide a general framework for population-representative mobility inference.
Read the full abstract

All authors Nicolò Alessandro Girardini, Unchitta Kan, Eduardo López, Bruno Lepri, Lorenzo Lucchini, Simone Centellegher