Papers

Recent 4D LiDAR language models aim to reason about objects and their evolving spatial relationships. Yet, in our evaluation, always selecting the same option nearly matches the multiple-choice accuracy of two B4DL-derived configurations. We introduce LiDAR-Hallu, a geometry-referenced benchmark and diagnostic protocol with 10,000 questions across 150 nuScenes scenes. It covers object existence, ego-relative position, distance ordering, relative motion, and temporal localization, with explicit rules for selecting objects, comparing times, and determining reference answers. Our protocol combines fixed-answer and candidate-content controls, cross-scene pairs with identical prompts but opposite reference answers, and relation-specific recall. Analysis of 100,000 recorded responses reveals failures hidden by aggregate accuracy. Candidate duration alone makes temporal answers predictable without observing LiDAR. On paired questions, the models frequently give the same answer to scenes requiring opposite answers. Relation-specific analysis further shows that both configurations miss every positive lateral-motion case across all tested conditions. Temporal-shuffle contrastive decoding provides little net improvement, as repairs are largely offset by new errors and the main failures persist. These results show that evaluating spatio-temporal reasoning requires testing whether models distinguish the queried physical relationships, rather than relying on individual-answer accuracy alone. The source code, checkpoints, and data are released at https://github.com/Awesome4D/4DMLLM_Hallucination_Bench.

The branched transport problem is a nonconvex and nonsmooth variational optimization problem on normal $1$-currents in $\mathbb{R}^n$ with prescribed boundary. The optimality is with respect to some non-decreasing, lower semicontinuous, and subadditive function $\tau:\mathbb{R}_+\to\mathbb{R}_+$ with $\tau(0)=0$ describing the cost $\tau(m)$ to move an amount of mass $m$ per unit distance. The subadditivity leads to complicated, hierarchically ramified patterns in the support of (suboptimal) solutions. These network-like sets appear to have regularity properties similar to those of the singular surfaces arising in the so-called Brakke flow, a weak generalization of the mean curvature flow. We construct a geometric flow of transportation networks, which correspond to normal real $1$-rectifiable currents, by modifying Brakke's variational approximation scheme. We prove the existence of a limit that is Hölder continuous with respect to the flat norm and whose branched transport cost decreases along the geometric evolution. We further analyze a closely related geometric flow approximated by $1$-varifolds, whose weight measures model the branched transport cost, and establish its $1$-rectifiability together with a motion law analogous to Brakke's inequality.

Memory is an inherent feature of the dynamics of physical systems. It typically emerges when the complex structure of the system is simplified during the modeling process, obscuring a part of its underlying behavior. Consequently, the correct description of such a physical model requires knowledge of not only its present, but also its past states, leading to non-Markovian dynamics. The resulting interactions depending on the system history, however, introduce a significant layer of mathematical complexity. For this reason, when the memory time is significantly shorter than the time scales characterizing the dynamics, it is typically neglected at the cost of losing information regarding its role in the system behavior. In this dissertation, I introduce a novel methodology, namely the effective mass approach, which bridges the Markovian and non-Markovian regimes and provides an accessible framework for analyzing systems with short memory. In this method, the non-Markovian system is approximated with its Markovian counterpart, in which the short-memory effects are encapsulated within an effective mass. I then utilize this approach to demonstrate that the impact of short memory on physical dynamics can be remarkably pronounced. In particular, I show that the directed transport of a Brownian particle in an environment exhibiting temporal correlations can be reversed relative to its behavior in a memoryless medium. Moreover, I propose a setup wherein memory serves as a~control parameter for the generation of random information bits. Collectively, the results of this dissertation underscore the critical importance of short-time correlations in microscopic systems.

This paper investigates some properties of the compound marker endomorphisms appearing in the study of the monodromy action of Hénon maps. The main result is to present a computable criterion for checking the bijectivity of a single marker endomorphism. Some direct corollaries are also developed, keeping in mind the Hénon dynamical aspect. Another point of the paper is the reformulation of compound marker endomorphisms in terms of permutive functions, which have been introduced by Hedlund. This reformulation allows to directly apply some of his results.

In this paper, we propose and analyze a novel three-field symmetric formulation for optimal control problems governed by linear thermo-poroelasticity. The state variables are the solid displacement $\boldsymbol{u}$, the fluid pressure $p$, and the temperature $\theta$, and both the distributed fluid source $m_p$ and the distributed heat source $m_\theta$ serve as control variables. We prove well-posedness of the symmetric differentiated formulation, establish the existence and uniqueness of an optimal control pair $(\bar{m}_p,\bar{m}_\theta)$, and derive the first-order necessary optimality conditions via a coupled three-field adjoint system. For the fully discrete scheme based on dG(0) time discretization, we employ a variational discretization approach for both controls and derive a priori error estimates of order $\mathcal{O}(h^s + \Delta t)$ for the state, adjoint, and control errors. We also present manufactured-solution experiments that validate the fully discrete optimality system and illustrate the storage-degenerate regime covered by our analysis.

The Whitham equation has given its name to a wider family of nonlocal, nonlinear equations with very weak dispersion, which all feature highest waves. Recent work by Ehrnström, Maehlen and Varholm establishes leading-order asymptotics at the crest of solutions to such Whitham-type equations, and conjectures similar expansions for all derivatives of the solution. By extending their techniques we confirm the conjecture for a range of equations with highest waves of Hölder regularity $C^{s}$ for $s\in[0.35,1)$. We do this by strong induction, redistributing difference operators to deal with singularities in higher-order derivatives arising from integral convolution kernels. The restriction $s\geq0.35$ arises solely from the separate argument establishing the zeroth-order asymptotics, which requires a uniform sign estimate that we verify using rigorous interval arithmetic. The higher-order induction itself applies for every $s\in(0,1)$, so any extension of the zeroth-order result immediately yields the corresponding higher-order expansions.

Large language model (LLM) agents increasingly execute long-horizon workflows through external tools, allowing untrusted outputs to influence subsequent actions and exceed user authorization. Existing defenses isolate injected content or constrain execution with predefined plans and static policies, but these approaches are brittle under dynamic workflows and scale poorly across extensible tool ecosystems. In this work, we present ActGov, a runtime enforcement framework that validates each LLM-proposed tool action before it causes external effects. Built on a unified semantic model of authorization, actions, runtime context, and security constraints, the ActGov-Policy component iteratively constructs a policy set from tool specifications, benign tasks, and observed failure traces, with each update verified through SMT-based counterexample checking. At runtime, ActGov-Runtime abstracts each tool call into finite policy records and permits it only if it remains within the task-scoped authorization boundary and satisfies all applicable policies. This per-action enforcement preserves authorization throughout long-horizon, dynamically branching workflows. We evaluate ActGov on the AgentDojo and AgentDyn benchmarks across multiple models and attack configurations. It shows that ActGov consistently reduces the success rate of indirect prompt-injection attacks while preserving task utility, significantly outperforming existing defenses. These results demonstrate that ActGov can enforce fine-grained authorization over dynamic agent executions without relying on the underlying LLM to correctly identify malicious instructions.

A direct charge-\(4e\) superconductor exhibits coherent four-electron order without condensation in any charge-\(2e\) pairing channel. We establish three complementary results. First, a particle-number-conserving Hamiltonian with electron hopping and two-body interactions moves bound quartets efficiently: for electron- and pair-hopping amplitudes \(t\) and \(K\) and local gap \(U_0\), virtual pair transfer produces quartet hopping at order \(K^2/U_0\), whereas electron hopping alone contributes only at order \(t^4/U_0^3\). Second, the microscopic theory identifies an exactly solvable parent Hamiltonian at the phase-separation boundary. Its ground state combines quartet off-diagonal long-range order and finite phase stiffness without pair condensation, while rigorous bounds establish quartet order beyond the parent in a family with explicit quartet transfer. Third, a single convergent strong-coupling expansion simultaneously controls quartet kinetics and thermodynamic stability and rules out condensation in every pair channel of the microscopic model. On bipartite hypercubic lattices at half quartet filling, it yields a parametrically controlled homogeneous phase for \(d\geq2\), with quartet order, a linear collective mode, and gapped charge-\(2e\) excitations. Virtual pair transfer thus enables a direct charge-\(4e\) superconducting ground state without pair condensation.

Accurate, reliable, and deployable wind power forecasting is critical for power system dispatch, renewable energy integration, and electricity market operations. Progress in this field hinges on the ability to empirically and comprehensively benchmark forecasting methods. Yet existing benchmarks fall short of supporting systematic evaluation in four key aspects: 1) limited coverage of wind power scenarios across turbine scale, variable composition, and spatial structure; 2) incomplete coverage of forecasting model families; 3) evaluation metrics misaligned with wind power requirements; and 4) limited structure-aware diagnostics beyond individual temporal patterns. To address these limitations, we propose WPBench, a comprehensive, fair, and extensible benchmark for wind power forecasting. WPBench integrates 26 public datasets organized by turbine scale and variable composition, spanning single-turbine, multi-turbine, univariate, and multivariate settings. Under unified processing, training, and evaluation protocols, it benchmarks 19 representative models covering traditional methods, deep temporal models, spatio-temporal models, and foundation models. Beyond point-wise errors, WPBench assesses forecast-curve fidelity and computational efficiency, and delivers structure-aware diagnostics across temporal, variable-dependency, and spatial-dependency perspectives. Together, these capabilities enable systematic model comparison across diverse wind scenarios and provide a reusable platform for future research.

Disruption predictors must warn the plasma control system early enough for the intended response, and the required warning time differs between mitigation and avoidance. We view a warning system as a predictor that outputs a disruption risk and an alarm policy that acts on it, and ask which risk the predictor should learn. Binary classifiers do not specify when a disruption occurs, and survival analysis models the full distribution of the time to disruption, whereas an alarm depends only on the risk within a finite horizon. We introduce two horizon-focused targets into tokamak disruption prediction: a horizon-truncated survival objective and fixed-horizon early event prediction (EEP) with temporal label smoothing (TLS). We compare them with binary classification and survival analysis on the DIII-D, Alcator C-Mod, and EAST databases of DisruptionBench, evaluating the risk by its threshold-free ranking at a 40 ms deadline and the alarm policy by thresholds calibrated on validation data. On every device, TLS outperforms the deep survival models in both ranking and alarm performance; TLS performs best on DIII-D and EAST, and a random forest trained on a hard horizon target performs best on C-Mod. TLS also improves the calibrated alarm performance over hard targets placed at either end of its smoothing range on all devices, isolating the benefit of label smoothing. Horizon truncation alone did not consistently improve survival analysis. Alarm performance remains limited on C-Mod and DIII-D and varies strongly across training seeds, and ablations show that the target should extend to the required warning time.

Context. Helmet streamers are large ray-like structures extending from the solar corona that thin further away from the solar surface, forming an extended current sheet. These structures are quasi-stable, and are observed to support kink waves travelling outward from the Sun, called helmet streamer waves. Aims. Limited analytical models identify streamer waves as fast body kink eigenmodes of the streamer slab. To bridge the gap between analytical models and numerical simulations, we investigate the eigenmode spectrum of more realistic streamer slab configurations, obtained from simulations, to retrieve the kink profiles and firmly establish that streamer waves are eigenmodes of the streamer. Methods. Using a streamer slab model extracted from numerical simulations, the Legolas code is applied to compute the eigenmode spectrum, for wavelengths observed in the simulation. In the spectrum we identify the mode matching the behaviour in the non-linear simulations by comparing to analytical and numerical results for the established Epstein profile. The identified modes's phase speeds are then compared to analytical, simulation, and observational results. Results. For each case, the spectrum is found to contain a mode matching the properties of a streamer wave. The phase speeds of the identified modes are compatible with those measured in observations, strongly suggesting that the observed streamer waves are fast body kink modes. Comparison to the analytical slab model is difficult due to the sensitivity of the analytical dispersion relation to the internal values.

Multivariate time-series forecasting is essential to many real-world applications. Recent large vision models (LVMs) offer a promising paradigm by transferring cross-domain visual priors to time-series forecasting. However, existing LVM-based methods face two key challenges: balancing independent visual representation spaces with cross-variable dependency modeling, and adapting vision backbones pretrained on natural images to the distinct temporal semantics of time-series images. To address these challenges, we propose MUSE, a dependency-aware adaptation framework built on a fully frozen pretrained MAE. First, the Variable Context Refinement Module (VCR) aggregates shared temporal information within each variable and models cross-variable contextual dependencies while preserving independent visual spaces. Second, the Temporal-Periodic Refinement Module (TPR) performs lightweight refinement at different encoder depths and explicitly models across-period temporal dependencies and within-period periodic dependencies. The two modules independently produce forecasts, which are fused through a learnable prediction-level gate. Experiments on 10 real-world datasets demonstrate that MUSE achieves state-of-the-art performance.

The quadratic scaling of Transformer self-attention has driven the adoption of sub-quadratic Selective State Space Models (SSMs) like Mamba, which compress past context into a fixed-size recurrent hidden state. This strict informational bottleneck raises a foundational question for mechanistic interpretability: do SSMs and Transformers learn fundamentally distinct latent representations? In this work, we employ Sparse Autoencoders (SAEs) to conduct a large-scale, feature-level correspondence analysis between Mamba-130m and Pythia-70m over a 10-million token corpus. Contrary to hypotheses predicting widespread architectural divergence, we find no evidence of systematic representational divergence between architectures: across the observed Jaccard distribution, 99.98% of Mamba features cluster toward the upper alignment boundary, providing preliminary feature-level support for the Universality Hypothesis. We further identify and qualitatively characterize this microscopic fraction (0.02%) of diverging features, finding patterns consistent with the hypothesis that the recurrent bottleneck selectively limits the parsing of rigid syntax rather than broad semantic ontology. We demonstrate that while Pythia's unconstrained attention permits the monosemantic decomposition of distinct formatting edge-cases, Mamba is forced to compress unrelated syntactical anomalies into polysemantic "junk drawer" neurons to preserve state capacity. Collectively, these results suggest that architectural routing mechanisms may have negligible impact on core semantic understanding, with representational divergence confined to extreme structural margins.

In the path algebra $KE$, where $K$ is a field and $E$ is a directed graph (or quiver), every non-loop edge can be expressed in the form $arb$, where $a,b,r\in KE$ and $ab=0$. This fact leads us to the characterization that $KE$ is semicommutative if and only if $E$ contains no non-loop edges. For a ring $R$, let $Z_i(R)=\{x\in R: x=arb, a,b,r\in R, ab=0 \}$ and call elements of $Z_i(R)$ zero-insertive. It follows that $R$ is semicommutative if and only if $Z_i(R)\subseteq E(R)$ and weakly semicommutative if and only if $Z_i(R)\subseteq N(R)$. We establish that every non-unit element of the Leavitt path algebra $L_K(A_2)$ is zero-insertive. If $K$ is a field and $n\geq 2$ is an integer, then each zero-insertive element of $L_K(A_n)$ is nil-clean if and only if $K\cong \mathbb{F}_2$. We call a ring $R$ zero-insertive nil clean (ZINC) if every zero-insertive element is nil clean. We show that the Leavitt path algebra $L_{\mathbb{F}_2}(A_n)$ is a ZINC ring. Additionally, we investigate the behavior of zero-insertive elements and ZINC rings under various ring extensions.

We study the emergence of non-equilibrium steady states (NESS) in stochastic processes under threshold resetting, an event-driven protocol in which the system resets to its initial configuration upon crossing a prescribed spatial boundary (threshold). In contrast to externally driven resetting, whose steady-state properties are well understood, the behavior under threshold resetting remains largely unexplored. We derive general conditions for the existence of a NESS and show that, whenever it exists, the steady state at a given position $x$ can be expressed as the ratio of two fundamental quantities: the mean local time (MLT) at $x$ and the mean first-passage time (MFPT) to hit the threshold. In particular, for noisy systems, a finite MFPT guarantees the existence of a NESS. As an illustrative example, we analyze a drift-diffusion process in one dimension and uncover rich intermediate-time dynamics, including anomalous relaxation in the spatial distribution, damped oscillations in the moments and in the mean-squared displacement (MSD), governed by system parameters. Our findings provide a general understanding of the emergence of non-equilibrium steady states and relaxation dynamics under threshold resetting, revealing how induced events shape the spatial and temporal properties of a broad class of stochastic processes.

We establish topological transitivity for a class of analytic partially hyperbolic diffeomorphisms on thickened nilmanifolds. These systems take the form of skew products, constructed as interval extensions of Anosov diffeomorphisms on nilmanifolds. Specifically, we develop a mechanism for topological transitivity allowing arbitrary finite-order parabolic degeneracies in the central direction.

We introduce a novel functional-style concurrent programming language called Categorical Message Passing Language (CaMPL) which is designed using the mathematics of linear actegories. This mathematical underpinning gives CaMPL programs useful properties such as deadlock freedom, and additionally, livelock freedom for programs without general recursive processes. We explore CaMPL's type system through a series of code examples. The current proto-alpha version of the compiler and the abstract machine is implemented in Haskell. A reader is encouraged to experiment with writing CaMPL programs either using the online compiler https://campl-app.vercel.app/ or by installing CaMPL from https://campl-ucalgary.github.io – our website has detailed instructions on how to run CaMPL code.

We consider initial and initial boundary value problems for the one-dimensional compressible Navier–Stokes equations of an ideal polytropic gas in unbounded domains. The viscosity and heat conductivity are proportional to $\theta^\alpha$ and $\theta^\beta$, respectively. For every fixed $\beta\geq0$ and sufficiently small $\alpha\geq0$, depending on the initial data, we prove the global existence and asymptotic stability of a strong solution for arbitrarily large $H^1$ initial data. Dividing the momentum equation by the viscosity allows us to estimate the specific-volume derivative without an initial $H^2$ assumption.

As computational resources continue to grow, the strong-scaling limitations of spatial parallelism motivate the pursuit of additional concurrency in the temporal dimension, particularly for applications with hard time constraints, such as weather and climate simulations. The Parallel Full Approximation Scheme in Space and Time (PFASST) is a parallel-in-time method based on Spectral Deferred Corrections (SDC). It computes multiple timesteps concurrently by coupling fine- and coarse-grid SDC sweeps using multigrid Full Approximation Scheme (FAS) corrections. However, PFASST's convergence is often problem-dependent, demanding a variable number of parallel timesteps and, hence, computing resources at different times throughout the simulation. Dynamic Resource Management (DRM) provides a remedy for this challenge by enabling the adaptive adjustment of computational resources and algorithmic parameters at runtime. In this work, we present our novel approach to extending PFASST with DRM, which enables (a) dynamic adaptation of computing resources, (b) adaptive selection of the number of PFASST iterations based on local convergence behavior, and (c) coupling of these two adaptations into a single resizing strategy. With this approach, we demonstrate for the first time that optimal configurations can be identified in real time for each application, rather than relying on static allocation. Furthermore, we show that convergence-informed tuning of PFASST improves resource utilization and convergence efficiency.

Low-bit vision-language-action inference must reduce observation-to-action latency while preserving robot behavior. We present FoldQuantVLA, a post-training quantization framework that carries a consistent activation representation through calibration, weight rounding, and native integer execution. It combines channel scaling and block Hadamard transforms with dynamic per-token quantization, without policy retraining. Custom TensorRT plugins execute projections in both the language backbone and iterative action expert with four-bit weights and activations (W4A4) on Ada GPUs and Jetson AGX Orin. Evaluation spans LIBERO, SimplerEnv, and two robot platforms. Across three GR00T checkpoints and $\pi_{0.5}$, W4A4 achieves $1.20$ to $1.33\times$ speedups over floating-point TensorRT on Orin and $1.25$ to $1.52\times$ on desktop. Retaining language attention-output and feed-forward down projections at eight bits (W8A8) improves held-out action fidelity on all four checkpoints. Across four real-robot tasks, this configuration raises observed GR00T N1.7 success from $80.0\%$ with uniform W4A4 to $92.5\%$ over 80 trials per configuration, with a measured additional Orin latency of 1 ms.

Sparse on-policy distillation (OPD) allocates teacher supervision to a small subset of tokens in student-generated trajectories. However, useful teacher guidance can yield a noisy update when its gradient is estimated from a sampled next token. We study this estimation problem at a fixed prefix in information geometry and propose an information-efficiency ratio (IER) based on a signal-to-noise decomposition. IER characterizes relative gradient estimation error under an optimal scalar baseline. A candidate-set approximation enables token selection based on IER and its combination with existing usefulness scores, while retaining the sampled reverse-KL training objective. On mathematical and medical reasoning tasks, adding IER improves existing selectors in multiple settings, with sparse configurations matching or exceeding full OPD without token selection at small token budgets of 0.1%–1%. These results support accounting for both usefulness and gradient-estimation reliability when allocating sparse supervision. Our code is available at https://github.com/BruceSheng1202/IER-OPD.

In this paper, we study the anisotropic-length-preserving weighted anisotropic curvature flow for smooth convex closed plane curves. For any smooth, embedded, closed, convex initial curve, the flow exists smoothly for all time, and the anisotropic curvature converges smoothly to that of the boundary of the corresponding homothetic Wulff shape as time tends to infinity.

Generative recommendation has emerged as an active research direction, where items are commonly represented by semantic IDs (SIDs): discrete codes generated token by token. Despite strong empirical results, SID designs vary widely in construction strategy, codebook organization, and code length, making their true impact on recommendation performance unclear. We conduct a large-scale reproducibility study to systematically investigate the impact of semantic ID design on generative recommendation under a unified experimental framework. We focus on a fundamental question: What makes a good semantic ID for generative recommendation? To answer this question, we examine four aspects: the relative effectiveness of different semantic ID designs, the connection between codebook utilization and recommendation quality, the effect of semantic code length, and the influence of semantic ID design on local item semantic preservation. Through a unified evaluation and additional cross-dataset controlled analyses, we find that the effects of SID design are largely non-monotonic: no single SID design is universally best, and commonly used RQ-VAE- and OPQ-based designs can behave inconsistently across datasets. The method with the most balanced first-level codebook is not consistently the best recommender, showing that utilization is diagnostic but insufficient. Scaling either the generative backbone or the SID length is also not always beneficial. Finally, semantic-neighborhood analysis reveals that no single SID design dominates all notions of local semantic preservation; instead, different designs exhibit complementary strengths that remain stable across datasets and neighborhood sizes. Our study provides a controlled and reproducible understanding of semantic ID design and offers practical insights for future generative recommender systems.

Repulsively bound doublons are two-particle composites formed by strong interactions and are usually separated from the scattering continuum. Introducing spin-orbit coupling fundamentally alters the underlying band structure, providing a powerful tuning knob to shift these isolated pairs toward this continuum. However, because entering such a regime typically dictates immediate dissociation, whether this coupling can drive these pairs inside while preserving their bound nature constitutes a fundamental unresolved challenge. Here we show that spin-orbit coupling in the one-dimensional Fermi-Hubbard model can drive doublons into the two-particle scattering continuum. Most of these states hybridize with extended channels and decay, whereas a subset remains decoupled and spatially bound, forming many-body bound states in the continuum (BICs). We map the interacting two-particle problem onto a two-dimensional lattice of coupled acoustic cavities, and experimentally observe both the radiating doublon continuum and the confined BIC states. These results demonstrate that spin-orbit coupling can turn selected doublons into interaction-induced BICs, deepening the understanding of continuum physics for interaction-bound pairs.

We investigate the harvesting of genuine tripartite entanglement by three linearly arranged Unruh-DeWitt detectors locally coupled to a massless scalar field in a gravitational wave background. We show that gravitational waves play a dual role in entanglement harvesting, either enhancing or suppressing the harvested entanglement depending on the system parameters. In particular, genuine tripartite entanglement exhibits a pronounced nonmonotonic response to the gravitational wave frequency, characterized by successive suppression, enhancement resembling resonance, and renewed suppression, while approaching its corresponding value in Minkowski spacetime in the regime of high frequencies. Its dependence on the detector separation is also nonmonotonic, displaying local enhancement peaks that are absent in the corresponding bipartite entanglement behavior. Compared with bipartite entanglement, genuine tripartite entanglement therefore exhibits a higher sensitivity to gravitational wave perturbations in both the frequency and spatial domains. These results suggest that genuine tripartite entanglement may provide a sensitive quantum probe of spacetime perturbations induced by gravitational waves.

We analyse two-dimensional isotropic heterogeneous diffusion with a radially decaying diffusion coefficient $D(\rho)=D_0/\rho^2$, subject to Poissonian stochastic resetting to the origin. Since multiplicative noise makes the process interpretation-dependent, we treat the Itô, Stratonovich, and Hänggi–Klimontovich prescriptions within a single approach. For each, we obtain the exact probability density function, the non-equilibrium stationary state induced by resetting, the mean squared displacement, and the first-passage properties. The non-equilibrium stationary state is governed by modified Bessel functions and departs from the Laplace form found for homogeneous diffusion under resetting, while the mean squared displacement saturates at long times as $\langle\rho^2\rangle\sim\bar{r}^{-1/2}$, $\bar{r}$ being the resetting rate. The mean first-passage time to an absorbing boundary exhibits an optimal resetting rate at which it is minimised. In addition, we incorporate memory effects through a subordination (continuous-time random walk) approach with a power-law kernel, deriving the fractional Fokker–Planck equation together with the non-equilibrium stationary state, the mean squared displacement expressed via the three-parameter Mittag-Leffler function, and the first-passage statistics; the optimal resetting rate increases as the memory exponent decreases.

We consider $N \times N$ random band matrices $H = (H_{xy})$ with centered complex Gaussian entries, indexed by points $x,y$ on the two-dimensional discrete torus $(\mathbb{Z} / \sqrt{N} \mathbb{Z})^2$. The matrix entries $H_{xy}$ vanish whenever the distance between $x$ and $y$ exceeds a constant multiple of the bandwidth parameter $W$. We prove that, for any fixed $\varepsilon>0$, if $W\geq(\log N)^{1+\varepsilon}$, then all bulk eigenvectors are delocalized with high probability. Equivalently, this yields a lower bound of order $\exp(W^{1-\varepsilon})$ on the localization length of two-dimensional random band matrices for arbitrarily small constant $\varepsilon>0$, improving the superpolynomial lower bound $W^C$ for every fixed constant $C>0$ established in arXiv:2503.07606. A key ingredient of our proof is a new strategy that closes the resolvent flow analysis using only loop observables of order at most two, without relying on the higher-order loop hierarchy.

We propose a protocol for extracting work from a coherent quantum battery state by exploiting measurement-assisted feedback mediated by a continuous-variable pointer. The scheme relies on the unitary operator $U = \exp(-i k t \, \hat{H} \otimes \hat{P}/\hbar)$, which generates entanglement between the battery's energy eigenstates and the position of an auxiliary pointer. A subsequent projective measurement of the pointer's position conditionally prepares the battery in a pure state from which work can be extracted via a feedback unitary. We analyze the protocol for a two-level quantum battery and a Gaussian pointer, computing the conditional states and the corresponding daemonic ergotropy. For the pure initial state considered, we find that the daemonic ergotropy equals the standard ergotropy for all interaction strengths, demonstrating that the measurement-assisted feedback recovers the full extractable work that would otherwise become inaccessible due to entanglement with the pointer when its degrees of freedom are traced out. The protocol thus provides a physically transparent realization of a quantum Maxwell demon, where the pointer acts as a quantum measurement ancilla whose position becomes correlated with the battery's energy. The scheme is amenable to experimental implementation in trapped-ion systems, and it contributes to the ongoing efforts to understand the role of quantum coherence and measurement in thermodynamics.

Monocular RGB-based hand pose estimation has emerged as a critical research frontier in computer vision. The local hand pose estimation methods predict hand poses relative to the wrist, while global hand pose estimation also requires estimating the wrist's position in the camera coordinate system. However, this camera-space estimation confronts two fundamental challenges: (1) depth ambiguity in monocular settings, and (2) the coupling effect of hand local poses and global wrist positions in the perspective projections. In particular, this coupling reflects that the projections are jointly determined by local hand poses, wrist positions, and camera intrinsics. To overcome these challenges, our framework proposes two key innovations: Transformation-Isomorphism Supervision for hand-depth information extraction and Perspective Information Embedding for resolving above coupling effect of local pose and wrist position, both integrated within the mainstream encoder-decoder architecture. Besides, we propose a novel framerate-aware multi-dataset training strategy for sequential pose refinement. Our fully integrated approach achieves at most 37.1% superiority in CS-MJE over SOTA on HO3D. Project page: https://github.com/Mine268/CS-ViT.

We consider a standard convex composite optimization problem with either smooth or nonsmooth objective function, and under quadratic growth. In recent years, several works gave algorithms based on a weak proximal oracle (WPO) that essentially match in oracle complexities proximal (sub)gradient methods relying on exact prox operations. Importantly, such WPOs, which relax the strong optimality condition of the standard prox operator, may admit much more efficient implementation in terms of runtime when optimal solutions have some sparse structure. A question remained if such WPO-based methods can be accelerated (in the sense of Nesterov's accelerated gradient). In this work we provide a negative answer by establishing lower bounds against both deterministic and randomized methods. Thus, while WPOs can substantially reduce the cost of individual oracle calls, this comes with an inherent loss in oracle complexity. We also provide a new upper-bound for WPO-based nonsmooth convex composite optimization, nearly matching the proximal subgradient method.