Group Theory
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 14 September
We establish three results on decision problems for finitely generated matrix groups, based on fiber products: $\bullet$ The torsion-freeness problem is undecidable for subgroups of $\mathrm{SL}_{4}(\mathbb{Z})$, $\bullet$ There exists a finitely generated subgroup $G$ of $\mathrm{SL}_{4}(\mathbb{Z})$ with undecidable square root problem, $\bullet$ No algorithm can, on input a tuple of invertible matrices, output the list of marked finite quotients of the group they generate.
High Energy Physics (Experiment)
Calibration of electromagnetic shower features in the CMS calorimeter with machine-learning techniques
Monte Carlo simulations are used extensively in high-energy particle physics analyses. However, an incomplete description of the underlying physics can lead to significant discrepancies between observables measured in simulated and collision data. To mitigate potential biases arising from such mismodelling, it is essential to calibrate simulations to data. This paper presents two novel calibration methods based on machine-learning techniques: a reweighting approach, which employs a classifier to learn the ratio of probability density functions between simulation and data; and a normalising-flow approach, which learns a high-dimensional transformation to map simulation to data. Compared to traditional calibration methods, both approaches offer continuous, unbinned corrections across high-dimensional feature spaces, enabling improved global agreement with data. The techniques are demonstrated in the context of correcting the features of simulated electromagnetic showers in the CMS calorimeters, using proton-proton collision data collected during 2022 at $\sqrt{s}$ = 13.6 TeV, corresponding to an integrated luminosity of 26.7 fb$^{-1}$. The strengths and limitations of the two methods are compared.
Read the full abstract
Logic
AKE principles in roughly deeply ramified henselian valued fields
We show that for any henselian valued field of mixed characteristic $(0,p)$, the (existential) $\mathcal{L}_{\mathrm{val}}$-theory of the valued field is determined by the (existential) theory of the value group in $\mathcal{L}_{\mathrm{oag}}$ with a constant for $v(p)$ and the (existential) theory of the residue ring $\mathcal{O}_v/(p)$ in an expansion $\mathcal{L}_{\mathrm{Witt}}$ of the language of rings, provided $\mathcal{O}_v/(p)$ is semi perfect. We moreover show that the $\mathcal{L}_{\mathrm{Witt}}$-structure on $\mathcal{O}_v/(p)$ is $\mathcal{L}_{\mathrm{ring}}$-definable using constants, and that this is exactly the structure induced on $\mathcal{O}_v/(p)$ by the ambient valued field. As a consequence, we obtain a relative quantifier elimination result (eliminating $K$-quantifiers) in a suitable language for the theory of roughly deeply ramified henselian valued fields of mixed characteristic $(0,p)$.
Read the full abstract
Cosmology and Nongalactic Astrophysics
Assembly Bias in eRASS1 X-ray Galaxy Clusters: Accounting for Non-uniform Survey Exposure
Halo assembly bias describes the dependence of halo clustering on properties beyond halo mass and provides a means of connecting halo assembly history to the large-scale matter distribution. We develop and test a relative clustering methodology for measuring differences in the large-scale bias of two halo populations using the ratio of their cluster–galaxy pair counts. In the linear regime, this ratio can be related to the difference in cluster bias, while the galaxy autocorrelation independently constrains the galaxy bias. We apply the method to a volume-limited X-ray selected galaxy clusters from the eROSITA eRASS1 catalogue, dividing them into high- and low-temperature subsamples with matched distributions in X-ray luminosity, redshift, and exposure. We first verify the validity of the linear approximation, finding $\xi(r)\ll1.0$ over $4$–$10\,h^{-1}\mathrm{Mpc}$. The joint clustering analysis yields $b_{\rm g}=1.610^{+0.018}_{-0.019}$ and a bias difference $\Delta b=0.8\pm1.1$. DES Y3 weak gravitational lensing measurements provide an independent test of the halo masses, yielding $\log(M_{200\mathrm{m}}/h^{-1}M_\odot)=14.33^{+0.07}_{-0.07}$ and $14.33^{+0.10}_{-0.09}$ for the high- and low-temperature samples, respectively, with a mass difference of $\Delta\log M_{200\mathrm{m}}=-0.01\pm0.12$. We therefore find no statistically significant evidence for assembly bias at the $1\sigma$ level, but our measurement is consistent with the bias difference measured for analogous halo populations in the FLAMINGO simulation. The result therefore does not rule out assembly bias at the level expected from the simulation, while demonstrating that the relative clustering methodology provides a framework for future assembly-bias measurements with larger X-ray cluster and galaxy surveys.
Read the full abstract
Analysis of PDEs
Parabolic uniform rectifiability is characterized by Carleson measure estimates for bounded caloric functions
We show that the parabolic uniform rectifiability of a set can be characterized by an interior PDE property, namely Carleson measure estimates for bounded solutions to the heat equation. In particular, under very mild background hypotheses on an open set $\Omega$ and its boundary $\partial \Omega$, we show that the following are equivalent: (i) The quantity $|\nabla u(\cdot)|^2 \text{dist}(\cdot, \partial\Omega)$ is the density of a Carleson measure on $\Omega$ for all bounded solutions to the heat equation in $\Omega$. (ii) The caloric measure for $\Omega$ admits a corona decomposition. (iii) $\partial \Omega$ is parabolic uniformly rectifiable. The implication (iii) implies (i) is the main result of the cited paper [Bortz, Hoffman, Hofmann, Luna-García, Nyström, Anal. PDE 2023]. The remaining implications are new, and are direct parabolic analogues of results of Garnett, Mourgoglou and Tolsa [Duke Math. J. 2018] in the elliptic setting for the Laplacian; however, our proofs require several new ideas. In particular, we must adapt arguments to account both for time-lag, and for the fact that some characterizations of (elliptic) uniform rectifiability have either not been developed or have been shown to be untrue in the parabolic setting. Our background assumptions, as noted above, are very mild: we assume that the domain satisfies the time-symmetric capacity density condition and has interior corkscrews, and that the boundary is (non time-directed) Ahlfors-David regular. These are weaker than the background assumptions of Bortz, Hofmann, Martell and Nyström in arXiv:2510.22047, and are essentially optimal for the boundary continuity properties of caloric functions on which our arguments rely.
Read the full abstract
Trading and Market Microstructure
Resolution Is Not Settlement, Part II: Protocol Finality and Observed Redemption on Polymarket
An Oracle result is not yet a protocol payout, a redeemable position is not yet collateral in a holder's account, and a redemption event is not a complete measure of economic entitlement. This companion paper develops an event-sourced framework for Polymarket conditions from preparation through protocol finality and observed holder realization. The empirical design uses three Conditional Tokens Framework event families derived from a pinned contract application binary interface (ABI): ConditionPreparation, ConditionResolution, and PayoutRedemption. It separates the contract-wide acquisition universe from the frozen Polymarket adapter-question cohort. The exact bridge contains 108,638 linked conditions. Of these, 99,283 have an observed protocol-resolution event by the fixed snapshot at Polygon block 90,114,204 (2026-07-12T17:11:41Z). Among resolved exact-linked conditions, 92,158 have an observed redemption of any amount and 91,817 have an observed positive-payout redemption; condition-specific Kaplan-Meier medians from first protocol resolution are 182 and 200 seconds respectively. The exact-linked payout taxonomy contains 53,847 canonical (0,1) vectors, 45,024 canonical (1,0) vectors, 410 fifty-fifty vectors, two other valid vectors, and 9,355 conditions with no observed resolution. Cross-contract Oracle-adapter-protocol ordering is reported conservatively: 823 conditions have an interval-qualified terminal generation, 48 have multiple candidate generations, 91,638 have no compatible terminal generation in the frozen evidence, and 16,129 are right-censored or otherwise unevaluable. The formal results show that Oracle finality does not identify protocol finality, protocol finality does not identify holder realization, and redemption events alone do not identify the fraction of entitlement redeemed without an independent balance-consistent entitlement denominator.
Read the full abstract
Game Theory
Differentially Private Multicolor Discrepancy and Fair Division of Indivisible Goods
We study the fair division of indivisible goods under pure differential privacy, continuing the line of work initiated by Manurangsi and Suksompong. For $n$ agents with nonnegative additive utilities over $m$ goods and a fixed privacy parameter, we give an entry-private algorithm that, with high probability, achieves consensus envy-freeness up to $O(\sqrt n+\log^3 m)$ goods. This substantially improves the dependence on $n$ over the previous $O(n\log m)$ guarantee for ordinary envy-freeness, while providing the stronger consensus guarantee. A key ingredient is a private algorithm for multicolor discrepancy, which may be of independent interest. Our algorithm may require exponential time. We also obtain substantially stronger guarantees under additional structure: when all item values belong to a public alphabet of size $D$, we give a polynomial-time entry-private algorithm achieving ordinary envy-freeness up to $O(\operatorname{polylog}(mD))$ goods with high probability. Finally, we prove an $\Omega(\log n)$ lower bound on the number of goods that must be removed to achieve ordinary envy-freeness under entry privacy, for sufficiently many goods, even with binary utilities.
Read the full abstract
Mesoscale and Nanoscale Physics
Unified contact-free formulation of linear-response transport theory
Thermal Hall transport is conventionally formulated as a current–current Kubo response supplemented by an energy-magnetization correction obtained through an auxiliary pseudogravitational magnetic field. Motivated by this structure, we develop a unified contact-free moment formulation of linear response to scalar sources. For channels whose source contributions vanish in the dc limit, the transport coefficient is the low-frequency residue of a retarded moment–moment correlator, with equal-time endpoint contributions incorporated before the dc limit is taken. The formulation uses only scalar sources, requires no separate magnetization subtraction, is independent of the local-current gauge, and is invariant under admissible redistributions of bond or interaction energy. It organizes particle, grand-energy, and specified spin-density channels within a common source–moment response matrix. Quadratic Landau–Lifshitz spin waves illustrate the method: the secular growth of polarization-moment correlations isolates the finite Hall residue and yields the magnon thermal Hall and spin Nernst coefficients from the same paraunitary band geometry. The formulation provides a unified operator framework that simplifies the organization of particle, thermal, and spin transport and can streamline calculations by eliminating separate contact and magnetization corrections.
Read the full abstract
Systems and Control
Volt-VAr-Watt Optimization in Four-Wire Low-Voltage Networks: Exact Nonlinear Models and Smooth Approximations - Extended Version
The proliferation of distributed energy resources is increasing the prevalence of both overvoltages and undervoltages in low-voltage (LV) distribution networks. Smart inverter functionalities, such as Volt-VAr and Volt-Watt control, can regulate voltage at the consumer level but are challenging to capture in optimization models due to their nondifferentiability. This paper proposes three nonlinear models for four-wire unbalanced optimal power flow that incorporate these nonsmooth functions without binary or integer variables. The first model encodes these nonsmooth functions directly as user-defined functions using control flow, a feat enabled by many state-of-the-art algebraic modeling languages. The second and third introduce bespoke smooth approximations with tunable approximation errors to address potential numerical issues arising from nondifferentiability, improving reliability while maintaining accuracy. All three methods are evaluated on real four-wire unbalanced LV networks with varying rooftop solar adoption levels, including a 539-bus system with 302 smart inverters. This paper is not only the first to demonstrate accurate, reliable, and tractable Volt-Var-Watt optimization (VVWO) on real four-wire unbalanced LV network models, but it also establishes that optimization-based, non-incremental methods offer superior reliability compared to commonly used incremental (quasi-steady-state) approaches.
Read the full abstract
Computation and Language
SlopShape: Identifying AI-Generated Commercial Web Content
Word-level detectors identify unedited AI-generated text almost perfectly, but the literature documents their brittleness under rewording, and a word-level score neither characterizes a text nor identifies which AI model wrote it. We ask whether AI-generated text can be identified one level deeper, from structural signatures: how information is presented, in what order, with what evidence, and in what voice. We replicate StoryScope (Russell et al., 2026), which showed such patterns for AI-generated fiction, on commercial content: 2,250 pre-ChatGPT human blog posts from 268 company domains against 11,250 AI mirrors from five frontier models. A 203-feature instrument, applied by an LLM and validated in a human gold-annotation session (human-human kappa 0.939, human-model 0.951), detects AI posts from its 176 structural features alone at 97.0 macro-F1 on held-out companies, nearly unchanged (96.1) when every AI post is reworded by its own model. The signal characterizes and attributes: AI posts share a tidy, self-announcing shape, 68.6% are attributed to the correct source against a 16.7% chance rate, and human posts occupy rare structural configurations. All effects replicate StoryScope's, consistent in direction and at least as large in magnitude. We release pipeline, instrument, prompts, code, and aggregate artifacts.
Read the full abstract
Trading and Market Microstructure
Resolution Is Not Settlement, Part I: Oracle Adjudication and Semantic Governance on Polymarket
Prediction-market resolution is often reduced to a terminal outcome and one timestamp. That representation is inadequate for leveraged event claims because rule versioning, request creation, proposal, dispute, reset, Oracle finality, and adapter terminality are distinct states with different observation precision and balance-sheet consequences. We reconstruct those states for Polymarket using Oracle request generations as the unit of adjudication. The population is frozen at Polygon block 79,721,080 and contains 185,550 initialized adapter-question instances and 350,703 decoded adapter logs. Exact requester-filtered extraction yields 504,332 decoded Oracle lifecycle events: 184,148 request creations, 159,447 proposals, 1,604 disputes, and 159,133 settlements. The accounting closes as 182,671 questions with at least one request plus 1,477 successor generations. Immutable chain identity and deployed request semantics provide exact linkage; unfinished histories remain right-censored. Request age is not semantic resolution age. Median request-to-first-proposal time is 182 seconds on the legacy route but 176,388-744,151 seconds on modern routes; post-reset successor proposals arrive within 300-2,909 seconds at the median. This descriptive contrast does not establish causal efficiency. Exact stable-ID metadata linkage recovers 104,032 of 185,550 questions (56.07%), leaves 81,518 unmatched, and produces no ambiguous exact match. External-source publication and contractual-decidability clocks remain unmeasured population-wide and are not replaced by mechanism timestamps. The results establish an event-sourced account of Oracle adjudication, semantic governance, and adapter terminality. Companion Part II reconstructs Conditional Tokens payout recording and observed redemption, preserving the boundary between adjudication, protocol settlement, and holder realization.
Read the full abstract
Computer Vision
CapsuleMotion: A Lightweight Real-Time Visual Motion Predictor for Capsule Endoscopy
Video Capsule Endoscopy (VCE) is a non-invasive medical examination that allows for the observation of the small intestine, which is otherwise difficult to access. A fundamental challenge persists in the form of their limited size in order to still be swallowable. The resulting restricted battery capacity, however, contradicts with the power-intensive nature of image capture and transmission. Therefore, we propose CapsuleMotion, a patient-specific dynamic capsule behavior that utilizes the available energy in a goal-oriented manner to increase the likelihood of a complete screening of the gastrointestinal tract. By investigating and combining metrics from the on-device image compression, CapsuleMotion predicts the motion between two successive frames. The camera's frame rate will be modified in accordance with the predicted magnitude of motion. Furthermore, prior to entering the small intestine, the capsule operates in a low power mode with a significantly reduced frame rate. In this mode, the LocalizationNet is employed to determine the current organ, provided that motion was predicted. The proposed framework is evaluated on the Rhode Island VCE dataset and deployed on an ultra-low power single-core RISC-V demonstrator with an integrated hardware accelerator. CapsuleMotion demonstrated the capability to reduce electric energy consumption by up to 20.66% in comparison with conventional capsules that lack a dynamic frame rate. Additionally, the accuracy of detecting the entry point of the small intestine has been improved.
Read the full abstract
High Energy Physics (Phenomenology)
Self-field of a moving string
We discuss the effectiveness of the recently proposed quantity $J_{s}$ in suppressing the contribution of a string's self-field to its spectrum, which is used in field theory simulations to track the emission of axions by decaying cosmic strings. We compute the contribution of the self-field to the spectrum of this quantity and to the usually computed spectrum of $\phi\partial_{t}\alpha$ for an infinitely long straight string. Although we demonstrate that the $J_s$ approach is a substantial improvement, we also point out that this highly symmetric model doesn't capture the full contribution of the self-field to the spectrum emitted by strings of a more complex shape, which are typical in network simulations. We then illustrate this point numerically using a simulation of a sinusoidally perturbed straight string. For this simple configuration, we managed to separate the contribution from the self-field and the radiation to the spectrum of $J_{s}$, using the self-field subtraction method. This allows us to show that the spectrum of $J_{s}$ is still dominated by the $n=1$ mode of the string's oscillation, which can be attributed to the self-field and is largely suppressed when the self-field is removed. We also demonstrate that this mode is predominantly sourced by variations along the direction parallel to the string, which is missing in the unconnected segment model, used by arXiv:2512.13653 to claim the effectiveness of $J_{s}$ in suppressing the self-field's contribution to the spectrum.
Read the full abstract
Instrumentation and Methods for Astrophysics
PYRA and MYTHRA: new statistical analysis for image reconstruction
Image reconstruction in optical interferometry remains a challenging inverse problem due to sparse Fourier sampling and the strong dependence of the reconstructed images on user-defined parameters. We present PYRA (Python for MiRA) and MYTHRA (Mean Astrophysical Images with PYRA), two tools designed to improve the robustness and reproducibility of interferometric imaging. PYRA automatically generates large ensembles of MiRA reconstructions by exploring a wide range of reconstruction parameters, while MYTHRA statistically analyzes the resulting images to identify the most consistent solutions and derive a final averaged image with associated uncertainty estimates. The methodology is validated using the blind datasets of the 2024 Interferometric Imaging Contest, where our reconstruction strategy achieved the best overall score. We further apply the method to VLTI/MATISSE observations of the triple AGB system {\pi}1 Gru and the A[e] supergiant 3 Pup. These results demonstrate that statistical analysis of large reconstruction ensembles provides a powerful framework for assessing image reliability and reducing reconstruction artifacts in optical interferometry.
Read the full abstract
Artificial Intelligence
RSIAgent: Autonomous Exploration for Recursive Self-improvement in New Environments
Digital agents must often adapt to new environments whose interfaces, tools, and failure modes are not fully captured by pretrained models. We introduce RSIAgent, a training-free multi-agent framework for recursive self-improvement through autonomous memory construction. RSIAgent coordinates curriculum, actor, and verifier agents to continually explore the environment, validate outcomes, and retain environment-specific knowledge, including reusable causal relationships between actions, conditions, and consequences. It further adopts a broad-then-deep exploration strategy, combining parallel broad recursive self-exploration for discovering diverse environment structures with focused deep self-exploration for uncovering hard cases, hidden constraints, boundary conditions, and previously unknown causal dependencies. The resulting memory is frozen and can be directly reused for downstream tasks without updating model parameters. Experiments on OSWorld-v2 and Agent's Last Exam show that RSIAgent substantially improves strong open-source models, enabling Kimi-K3 and GLM-5.3 to outperform frontier closed-source models including GPT-6.
Read the full abstract
Analysis of PDEs
Pointwise convergence for single-particle Schrödinger equation and Schrödinger equation with infinitely many particles on the torus and waveguide manifolds
In 1980, Carleson posed a question about the least regularity required for initial data in a Sobolev space $H^s$ to ensure pointwise convergence of the solution to the single particle linear Schrodinger equation. In this paper, we study Carleson problem on the $d$-dimensional waveguide manifold $\mathbb R^n\times \mathbb T^m$ and initiate its analogy for the system of infinitely many orthonormal particles (arising from the transition of many-body quantum mechanics to the thermodynamic limit) on torus and waveguide manifold. For a single particle on the waveguide manifold, we establish a maximal-in-time Strichartz estimate that yields almost everywhere convergence to the initial data in $H^s$ for $s > \frac{d}{d+2}$. On the other hand, we adopt Bourgain counterexample approach to show that this convergence fails for $s< \frac{d}{2(d+1)}.$ For infinitely many particles, we generalize the classical maximal-in-time Strichartz estimate and pointwise convergence result of Compaan, Luca and Staffilani (2021) in the setting of fermionic systems on the torus. Similar result is also established for the waveguide manifold. We also establish necessary condition (in the spirit of Bourgain's counterexample) for the pointwise convergence problem for fermionic systems. These are the first results in the setting of torus and waveguide manifold, and complement the works of Bez, Lee and Nakamura (2020) and Bez, Kinoshita, Shinya and Shiraki (2024) on Euclidean space.
Read the full abstract
Robotics
Understanding User Preferences of a Slope-Aware Variable-Admittance Filter for a Robot Guide Dog
This letter investigates how the parameters of a slope-aware variable-admittance filter influence user preferences in force-based interaction with a robotic guide dog for visually impaired individuals. The proposed system consists of a quadruped robot equipped with a sensor-free rigid handle for physical guidance. The framework combines path following, momentum-based interaction-wrench estimation, and a variable-admittance filter whose stiffness and damping are adapted online from slope information extracted by the robot's depth camera. The adaptation policies are evaluated through high-fidelity simulations and a human-subject study involving blindfolded sighted participants. Multiple strategies are compared using a Taguchi L9 design of experiments. Preliminary main-effect results suggest that increasing stiffness uphill and decreasing it downhill improves both objective and subjective metrics, whereas damping shows no significant main effect.
Read the full abstract
Machine Learning
Robust and Efficient Communication for Multi-Agent Learning
Effective communication is a cornerstone of distributed intelligence in Multi-Agent Reinforcement Learning (MARL), yet ensuring that generated messages are both informative and robust to physical constraints remains a significant challenge. This paper introduces Multi-Agent Regularized Communication (MARC), a novel framework inspired by information-theoretic principles of conditional mutual information. MARC employs an attention-based architecture coupled with a unique message regularization mechanism designed to minimize uncertainty regarding future system states, thereby inducing the learning of highly representative communication protocols. Crucially, we evaluate MARC under stringent communication bottlenecks and lossy channels, simulating the real-world constraints of autonomous robotic networks and decentralized systems. Our results demonstrate that MARC significantly outperforms state-of-the-art methods in complex cooperative domains. Furthermore, we provide a deep analysis of message characteristics, proving that MARC maintains high operational performance even under significant data compression, offering a scalable path for deploying intelligent agents in resource-constrained environments.
Read the full abstract
Analysis of PDEs
Boundedness and contractive estimates for orthotropic, widely degenerate, doubly nonlinear diffusion equations
We study the regularity of weak solutions to doubly nonlinear orthotropic evolution equations of the form \[ \partial_{t}(\vert u\vert^{\alpha-1}u)-\sum_{i=1}^{N}\partial_{i}\left[a_{i}(x,t)\,(|\partial_{i}u|-\delta_{i})_{+}^{p-1}\frac{\partial_{i}u}{\vert\partial_{i}u\vert}\right]=f\,\,\,\,\,\,\,\,\,\,\mathrm{in}\,\,\,\Omega_{T}=\Omega\times(0,T), \] where $\Omega$ is a bounded open subset of $\mathbb{R}^{N}$ for $N\geq2$, the coefficients $a_{i}$ are measurable and bounded, $\alpha>0$, $p\in(1,\infty)$ and $\delta_{1},\ldots,\delta_{N}$ are non-negative numbers. We show that weak solutions are locally bounded due to their membership in a suitable De Giorgi-type energy class. We also obtain contractive estimates and global boundedness in space for solutions to a Cauchy problem associated with the above PDE. Our analysis extends analogous results available in the literature for diffusion equations that either do not exhibit double nonlinearity or are less degenerate than those considered here. Another main novelty of this paper is the presence of a source term $f$ on the right-hand side of the equation, for which we impose suitable integrability assumptions in the space-time variables.
Read the full abstract
Artificial Intelligence
MAPS: Memory-Aware Predictive Scheduling Framework for Large Language Model Serving
The surge of large language model (LLM) applications on personal devices imposes massive, bursty workloads on cloud serving infrastructure. While prefill-decode disaggregation improves throughput and scalability, memory-bound decode instances often suffer from persistent load imbalance, as output lengths are unknown when requests arrive at the cloud. To address this, we propose MAPS, a Memory-Aware Predictive Scheduling framework tailored for disaggregated LLM serving. MAPS performs device-assisted speculative output length prediction overlapped with cloud-side prefilling, incurring negligible latency overhead. To handle generation uncertainty, MAPS applies uncertainty-aware calibration to derive output-length upper bounds with target coverage, enabling safe scheduling decisions. Building on these bounds, MAPS employs a hierarchical global-local scheduling strategy to mitigate inter-decoder queue buildup and intra-decoder head-of-line blocking. Extensive experiments on two real-world workloads and two LLMs show that MAPS significantly outperforms three state-of-the-art systems, reducing average end-to-end latency by 42.6 and tail latency by up to 84.8.
Read the full abstract
Game Theory
Reconfiguration in Fair Division Revisited
We revisit reconfiguration in the fair allocation of indivisible goods, where the goal is to transform one fair allocation into another through a sequence of exchanges while preserving fairness at every step. Our focus is on the hierarchy of envy-freeness up to $k$ goods (EF$k$). We show that for any fixed $k$, two EF1 allocations with the same size vector need not admit a reconfiguration path whose intermediate allocations satisfy EF$k$. This impossibility persists even when the two allocations arise from standard EF1 approaches: the envy cycle elimination algorithm or the maximum Nash welfare solution. In contrast, we prove that allocations with the same size vector produced by recursively balanced picking sequences, including round-robin, are always connected via a path that maintains EF2. We also show that deciding whether an EF1 reconfiguration path exists is NP-hard for any fixed number of agents. Furthermore, we complement these exchange-based results by studying a more permissive model that also allows transfers, establishing additional connectivity guarantees.
Read the full abstract
Computer Vision
Diffusion Trajectory Modeling for Semantic Correspondence
Diffusion models generate images through an iterative diffusion process, and recent studies have demonstrated that the intermediate feature maps produced during this process contain rich visual representations, leading to their adoption across a variety of downstream tasks. However, most existing approaches are limited to either using a single feature map at a specific timestep or aggregating feature maps across multiple timesteps. We observe that intermediate representations in the diffusion process form meaningful trajectories along the time axis. In particular, the representation of each spatial patch evolves progressively throughout the generative process, encoding semantics that are difficult to capture from static snapshots alone. This observation motivates the need to treat diffusion representations as temporally structured trajectories rather than static snapshots. To this end, we propose Diffusion Trajectory Modeling (DTM), a framework that interprets the temporal evolution of each spatial patch as a trajectory and leverages it for semantic correspondence. By effectively modeling patch-wise trajectories generated across multiple timesteps, DTM captures correspondence cues that prior methods are not designed to capture. We further demonstrate empirically that spatially corresponding patches form similar trajectory patterns throughout the diffusion process, suggesting that the temporal axis of diffusion carries semantic information. Experiments on SPair-71k, SPair-U and AP-10K show that DTM achieves strong performance, presenting a new perspective for exploiting diffusion representations from a trajectory-centric viewpoint.
Read the full abstract
Signal Processing
Sparse Channel Estimation and Signal Recovery for Reduced-PAPR Visible Light Optical OFDM Systems Relying on Bayesian Learning
A multipath CIR estimator is proposed, followed by sparse frequency-domain (FD) signal detection, which harnesses the simultaneous-sparsity innate in the multipath CIR and FD signals across measurement vectors, conceived for optical OFDM (O-OFDM) utilized for visible light communication (VLC) systems having reduced peak-to-average power ratio. At the outset, we derive the input-output relationships for the asymmetrically clipped O-OFDM (ACO-OFDM) as well as for the direct current-biased O-OFDM (DCO-OFDM) systems. Next, traditional benchmarking methods are introduced for channel estimation (CE), including both the LMMSE and LS methods, followed by an orthogonal matching pursuit (OMP)-based technique capable of exploiting sparsity in the multipath CIR of the VLC system. Subsequently, an CE technique based on a novel group sparse OMP (GOMP) concept is proposed, which capitalizes on the group-sparsity in the delay domain of the CIR across measurement vectors, attributed to the multipath characteristics inherent in the NLoS components of the VLC channel model. Additionally, an advanced group-sparse CE technique is put forth based on the group-sparse Bayesian learning (GBL) approach, which considerably reduces the pilot overhead. A low complexity version of GBL, termed LCGBL, is also developed that reduces the computational cost of GBL significantly. Consequently, the GOMP and GBL frameworks are also extended to the data detection of the sparse FD O-OFDM symbols, which utilizes the group-sparsity of the FD O-OFDM symbols across measurement vectors. The Bayesian Cramer-Rao lower bound (BCRLB) is computed to assess the estimation performance of the proposed CE techniques. Our simulations demonstrate that, despite its reduced pilot overhead, the proposed GBL technique outperforms the other CE schemes in terms of its BER, outage probability, and NMSE, quantitatively reinforcing its superiority.
Read the full abstract
Machine Learning (Statistics)
Approximating Smooth Functionals with ReLU Networks
We study the uniform approximation of smooth scalar-valued functionals on an infinite-dimensional separable Hilbert space by ReLU neural networks. A key feature in deep learning for functional data is the varying importance of different coordinates/dimensions. Representing the functional input in a basis expansion, we quantify the importance of each coordinate through both the magnitude of its corresponding basis score and the directional sensitivity of the target functional. Our analysis combines coordinate truncation, anisotropic partitioning, local Taylor approximation, and ReLU network realization, while allowing unrestricted interactions among the retained coordinates. We establish a general nonasymptotic upper bound for the uniform approximation error and a complementary pseudo-dimension-based lower bound for the worst-case approximation error. Under generalized exponential coordinate decay $w_ds_d\asymp\exp(-cd^\rho)$, with $\rho>0$, the upper and lower bounds match at the leading order, which is stretched-exponential in the logarithm of the network size budget, and thus yield the nearly optimal approximation rate. This is the first work to characterize neural network approximation error for infinite-dimensional functional inputs explicitly through the joint dimensional decay of coordinate magnitudes and directional sensitivities.
Read the full abstract
Computer Vision
End-to-End Cell Detection via Instance-aware Graph Modeling
Accurate cell detection and classification are crucial for pathological analysis, directly affecting diagnostic accuracy and treatment planning. To capture complex cellular interactions beyond visual appearance within the tumor microenvironment, several approaches have employed graph neural networks to model spatial and relational patterns among cell nuclei, yielding promising results. However, these methods typically adopt a two-stage paradigm of visual extraction followed by relational modeling, which necessitates separate tuning for each stage, thereby increasing pipeline complexity and hindering end-to-end joint optimization. In this paper, we propose an end-to-end framework for cell detection and classification that jointly models patch-level visual representations and instance-level interactions, which incorporates a dynamic graph construction module and an instance-aware graph network. Specifically, the graph construction module dynamically builds the graph structure using learnable queries derived from patch-level features as cell instance representations, with adjacency defined by integrating feature similarity and spatial distances. The instance-aware graph network performs adaptive instance filtering and feature reorganization, aggregating them over the cell graph into a topological latent state for a selective state-space transition driven by visual cues, fusing appearance and relational evidence. When evaluated on multiple datasets with different staining protocols for cell and nucleus detection, our method significantly outperforms existing approaches in both detection and classification performance. The code will be released at https://github.com/RuochenLiu23/IGM.
Read the full abstract
Statistical Methodology
Modelling multiple disease risk factors for microsimulation studies: a review of methods
Longitudinal microsimulation models to evaluate policies and scenarios for chronic disease risk reduction typically involve simulating multiple risk factors for a synthetic population over time. Various statistical methods have been used to accomplish this, but the principles behind them have never been comprehensively reviewed or critically compared. We describe and review the range of methods that have been used to simulate multiple risk factors over time, given either longitudinal or repeated cross-sectional data, highlight those which are most practically applicable and flexible, and introduce some useful variants of them. We describe procedures that can be used to check and compare these methods in practice based on fit to data, and present a case study where the methods are used and compared, based on a microsimulation model to evaluate mid-life cardiovascular health checks. We conclude that a useful range of procedures are available, and this paper improves the knowledge base to support their use in models to inform health policy.
Read the full abstract
Robotics
Assistance Torque Estimation via Dynamics-Aware Optimization for Lower-Limb Exoskeleton in Complex Environments
Ground-truth human joint torque estimation relies on motion capture systems, which suffer from limited outdoor usability and significant deployment expenses. Furthermore, direct scaling of ground-truth joint torques to obtain motor torque commands is not necessarily the optimal strategy. To address the aforementioned limitations, inspired by the human motion generation process, this paper proposes a novel assistance torque estimation method based on the dynamic model. From an optimization perspective, the proposed method directly generates motor-assist torque and lowers the cost of data acquisition. Then, a data-driven assistance torque prediction network is trained to enable accurate real-time prediction under complex outdoor environments. Experimental results demonstrate that optimized (estimated) assistance torque exhibits better phase consistency with gait trajectories and better alignment with task characteristics. Relative to the Zero torque condition, the predicted torque can decrease metabolic rate by 11.8%-17.7%, heart rate by 8.9%-14.3%, and peak muscle activation levels by 28.2%-54.0%, respectively. This provides a new perspective for low-cost adaptive exoskeleton assistance.
Read the full abstract
Atomic Physics
State-Dependent Diffusion and Spectra of Strongly Driven Thermal Atoms
We propose a state-dependent diffusion model for strongly driven thermal-atom spectra. Starting from the trajectory-dependent internal-state evolution of individual atoms, we derive a closed spatial equation for the local density-matrix field using a velocity-moment expansion. Measurements of an $^{85}$Rb atomic-filter transmission spectrum agree well with the model up to a maximum Gaussian peak intensity of $1.27\times10^{3}$ W/cm$^2$, approaching six orders of magnitude above the $^{85}$Rb D2-line saturation intensity. Counterintuitively, the model reveals an anomalous optical-pumping pathway in which intense light transfers atoms from nominally dark states into bright states. Hyperfine Paschen–Back splitting selectively enhances this anomalous pathway while suppressing conventional optical pumping, allowing the filter to maintain approximately 97$\%$ transmission at the highest intensity studied. This work provides a framework for controlling strongly driven atomic ensembles and designing saturation-resistant atomic optical devices.
Read the full abstract
Software Engineering
When Digitalization Transforms Itself: AI, Software, and the Next Technical Order
Agentic artificial intelligence (AI) marks a new phase of digitalization: digitalization is beginning to act back upon its own technical production base. Whereas earlier phases aimed at digitizing analog information, automating processes, and building digital value networks, AI is increasingly taking over tasks in the development of digital systems themselves. In software engineering (SE) in particular, agentic AI systems can already plan, execute, check, and iteratively refine development tasks across multiple steps. Digitalization thus becomes recursive: digital systems no longer merely support the digitalization of other sectors, but are increasingly developed and evolved by digital systems themselves. The technical production base of digitalization thereby becomes the object of its own transformation. This changes not only the division of labor between humans and machines in SE, but also the role of SE as a whole. This chapter analyzes this socio-technical change and argues that SE is evolving from a primarily supporting engineering discipline into a central infrastructure of societal value creation as well as of technological and digital sovereignty.
Read the full abstract
Physics and Society
Forced Coincidence and Its Boundary in Finite Condition Structures
Accounts of organizational dissolution explain survivor homogenization, temporally concentrated exit, and fragmentation into cohesive subgroups in different vocabularies. Their recurrence is the fact to be explained. A finite condition structure with a minimal dynamic extension generates all three, each in its own regime, with no actor responding to another; which conclusions survive when the restrictions defining those regimes are relaxed is then determined. With two requirement masks the union class is the only mediator between the pure types and the signature with the largest closed neighbourhood, and under positive endpoint loads it is also the unique maximizer of deficient load, so the independent derivations of fragmentation and homogenization both concern that class. With more than two masks the three roles need not coincide. Forced coincidence admits an exact characterization for every number of masks: it holds if and only if the requirement-derived join lies in the realized family and separates the endpoints. The proportion of family-pair instances meeting this condition vanishes under a uniform counting measure as the number of masks increases, and, at one actor per signature, an edge density above a stated bound excludes it. A common substrate does not make the mechanisms identical: selective attrition can hold while correlated collapse fails, and a change in exit dependence can change the population outcome while every actor's marginal lifetime law is unchanged.