arXiv subjects

Han Dong

Publications and source records attributed to Han Dong.

14 recordsLinked to original sources

arXiv2026

Sinkhorn Linearization and the Spectral Proxy: Unifying the Statistical and Algorithmic Theory of Feature-Parameterized Inverse Optimal Transport via a Single Spectral Sandwich

We develop the statistical and algorithmic theory of inverse optimal transport (IOT) under the feature-parameterized cost C_theta(i,j) = -theta^T phi(i,j). The core technical contribution is the Sinkhorn linearization -- the implicit-function sensitivity of the entropic OT plan to the cost -- together with its spectral proxy, a formula that is spectrally exact yet geometrically transparent. The restricted Hessian on the tangent space satisfies the spectral sandwich (pi_min/epsilon) I <= H_T^{-1} <= (pi_max/epsilon) I, yielding the single core bound sigma_min >= (pi_min/(a_max epsilon)) sqrt(lambda_min(Sigma)) that drives the entire theory. On this core we establish four theorems and one observation. T1 (identifiability): theta is globally injective on the quotient of the gauge kernel, with dimension bound F <= (K-1)^2. T2 (sparsistency): the l1-penalized estimator recovers the true support under irrepresentability and score concentration, with exponential failure probability. T3 (well-posedness): the feature-moment map M(theta) = Phi^T x_theta is strongly monotone, and the inverse is Lipschitz with constant L <= epsilon ||Phi^T S_a||_op / (pi_min lambda_min(Sigma)). T4 (convergence): local strong convexity with mu >= pi_min^2 lambda_min(Sigma) / epsilon^2 guarantees monotone gradient descent convergence. O5 (misspecification): the estimator converges to the OT-model projection of the truth; the Holder continuity of the projection map is assessed numerically, yielding setting-dependent empirical exponents alpha_eff in (0,1).

stat.ML↗

arXiv2026

VLA-Trace: Diagnosing Vision-Language-Action Models through Representation and Behavior Tracing

Understanding how Vision-Language-Action (VLA) models transform multimodal knowledge into embodied control remains an open challenge. We present VLA-Trace, a progressive diagnostic framework that analyzes VLA models through a unified evidence chain from representation dynamics to causal control attribution and behavioral manifestation. It specifically combines cross-modal and checkpoint-drift centered kernel alignment (CKA) to trace representation evolution, attention knockout interventions to identify modality-specific control pathways, and rollout-level behavioral probes to examine grounding, shortcut dependence, and semantic following. Experiments on $\pi_{0.5}$ and OpenVLA reveal three key findings. First, the two models exhibit distinct modality-specific adaptation dynamics during VLA finetuning. Second, they rely on different multimodal routing strategies and layer-wise dependencies during action decoding. Third, although VLA policies excel at visually grounded trajectory generation, they remain limited in fine-grained semantic following. These findings highlight future directions for representation-preserving adaptation, causal VLA circuits, and compositional semantic control.

cs.AI↗

arXiv2026

EPIC-Bench: A Perception-Centric Benchmark for Fine-Grained Embodied Visual Grounding in Vision-Language Models

While large vision-language models (VLMs) are increasingly adopted as the perceptual backbone for embodied agents, existing benchmarks often rely on question-answering or multiple-choice formats. These protocols allow models to exploit linguistic priors rather than demonstrating genuine visual grounding. To address this, we present EPIC-Bench, Embodied PerceptIon BenChmark, a fine-grained grounding benchmark designed to systematically evaluate the visual perceptual capabilities of VLMs in real-world embodied environments. Comprising 6.6k meticulously annotated tuples (Image, Text, Mask), EPIC-Bench spans 23 fine-grained tasks across three core stages of the embodied interaction pipeline: Target Localization, Navigation, and Manipulation. Extensive evaluations of over 89 leading VLMs reveal that while advanced reasoning models show promise, current VLMs universally struggle with complex visual-text alignment for physical interactions. Specifically, models exhibit critical bottlenecks in multi-target counting, part-whole relationship understanding, and affordance region detection. EPIC-Bench provides a robust foundation and actionable insights for advancing the next generation of vision-driven embodied models.

cs.CV↗

arXiv2025

High-Discretization Method of Moments for Capacitance Calculation: A Cube and a Hollow Cylinder

This paper employs the method of moments (MOM) to calculate the capacitances of a cube and a hollow cylinder. For the cube, each face was divided into a maximum of 600 x 600 sub-areas. By fully exploiting the geometric symmetry between sub-areas and incorporating parallel computing, computational resources were significantly conserved. Our results show that the calculated capacitance of the cube first increases and then decreases as the number of sub-areas increases. When each face was divided into 90 x 90 sub-areas, the capacitance of the unit cube (with an edge length of 1 m) reached a maximum reference value of 73.519014 pF. This indicates that higher accuracy cannot be achieved merely by indefinitely increasing the number of discretized sub-areas. Subsequently, the method was applied to compute the capacitance of a hollow cylinder. The results were compared with numerical solutions based on Lekner's theoretical formula and Cavendish's experimental values, showing good agreement among the three.

physics.class-ph↗

arXiv2025

Taming and Controlling Performance and Energy Trade-offs Automatically in Network Applications

In this paper, we demonstrate that a server running a single latency-sensitive application can be treated as a black box to reduce energy consumption while meeting an SLA target. We find that when the mean offered load is stable, one can find the "sweet spot" settings in packet batching (via interrupt coalescing) and controlling the processing rate (DVFS) that represents optimal trade-offs in the interactions of the software stack and hardware with the arrival rate and composition of requests currently being served. Trying a few combinations of settings on the live system, an example Bayesian optimizer can find settings that reduce the energy consumption to meet a desired tail latency for the current load. This research demonstrates that: 1) without software changes, dramatic energy savings (up to 60%) can be achieved across diverse hardware systems if one controls batching and processing rate, 2) specialized research OSes that have been developed for performance can achieve more than 2x better energy efficiency than general-purpose OSes, and 3) a controller, agnostic to the application and system, can easily find energy-efficient settings for the offered load that meets SLA objectives.

cs.OS↗

arXiv2024

A First-Principle Approach to X-ray Active Optics: Design and Verification

This paper presents the first-principle design approach for X-ray active optics, using the simulation-modulation cycle in place of the measurement-modulation feedback loops used in traditional active optics. Hence, the new active optics have the potential to outperform the accuracy of surface-shape metrology instruments. We apply an X-ray mirror with localized thermal elastic deformation to validate the idea. Both the finite element simulations and surface shape measurements have demonstrated that the active optics modulation accuracy limit can be achieved at the atomic layer level. It is believed that the implementation of the first-principle design strategy has the capacity to revolutionize both the manufacturing processes of X-ray mirrors and the beamline engineering of synchrotron radiation.

physics.optics↗

arXiv2021

Slowing Down for Performance and Energy: An OS-Centric Study in Network Driven Workloads

This paper studies three fundamental aspects of an OS that impact the performance and energy efficiency of network processing: 1) batching, 2) processor energy settings, and 3) the logic and instructions of the OS networking paths. A network device's interrupt delay feature is used to induce batching and processor frequency is manipulated to control the speed of instruction execution. A baremetal library OS is used to explore OS path specialization. This study shows how careful use of batching and interrupt delay results in 2X energy and performance improvements across different workloads. Surprisingly, we find polling can be made energy efficient and can result in gains up to 11X over baseline Linux. We developed a methodology and a set of tools to collect system data in order to understand how energy is impacted at a fine-grained granularity. This paper identifies a number of other novel findings that have implications in OS design for networked applications and suggests a path forward to consider energy as a focal point of systems research.

cs.OS↗

arXiv2021

Hypothesis Test of a Truncated Sample Mean for the Extremely Heavy-Tailed Distributions

This article deals with the hypothesis test for the extremely heavy-tailed distributions with infinite mean or variance by using a truncated sample mean. We obtain three necessary and sufficient conditions under which the asymptotic distribution of the truncated test statistics converges to normal, neither normal nor stable or converges to $-\infty$ or the combination of stable distributions, respectively. The numerical simulation illustrates an application of the theoretical results above in the hypothesis testing.

math.ST↗

arXiv2019

SEUSS: Rapid serverless deployment using environment snapshots

Modern FaaS systems perform well in the case of repeat executions when function working sets stay small. However, these platforms are less effective when applied to more complex, large-scale and dynamic workloads. In this paper, we introduce SEUSS (serverless execution via unikernel snapshot stacks), a new system-level approach for rapidly deploying serverless functions. Through our approach, we demonstrate orders of magnitude improvements in function start times and cacheability, which improves common re-execution paths while also unlocking previously-unsupported large-scale bursty workloads.

cs.OS↗

arXiv2019

Generalized Noether Theorem for Gauss-Bonnet Cosmology

Generalized Noether's theory is a useful method for researching the modified gravity theories about the conserved quantities and symmetries. A generally Gauss-Bonnet gravity $f(R,\mathcal{G})$ theory was proposed as an alternative gravity model. Through the generalized Noether symmetry, polynomial and product forms of the $f(R,\mathcal{G})$ theory with corresponding conserved quantities and symmetries are researched. Then suitable general forms of the polynomial form $f(R,\mathcal{G}) \!=\! k_1 R^n + (6)^{\frac{n}{2}}(-1)^{n+1} k_1 \mathcal{G}^{\frac{n}{2}}$ and the product form $f(R,\mathcal{G}) \!=\! k ( R / \sqrt{\mathcal{G}} )^n \mathcal{G}$ are found out, to contain the solution of accelerated expansion cosmology. Both forms of $f(R,\mathcal{G})$ concerned in this paper only possess time translational symmetry. And energy condition of these solutions are also checked. To some extent, the consistency of conservation of symmetry and energy condition is demonstrated. For the specific form of different $n$, it needs further detailed study. Noting that, the corresponding conserved quantities are both zero, and the only conservation relation is conservation of energy.

gr-qc↗

arXiv2013

Waveform Digitizing for LaBr$_3$/NaI Phoswich Detector

The detection efficiency of phoswich detector starts to decrease when Compton scattering becomes significant. Events with energy deposit in both scintillators, if not rejected, are not useful for spectral analysis as the full energy of the incident photon cannot be reconstructed with conventional readout. We show that once the system response is carefully calibrated, the full energy of those double deposit events can be reconstructed using a waveform digitizer as the readout. Our experiment suggests that the efficiency of photopeak at 662 keV can be increased by a factor of 2 given our LaBr$_3$/NaI phoswich detector.

physics.ins-det↗

arXiv2013

The distinctions between $Λ$CDM and $f(T)$ gravity according Noether symmetry

Noether's theory offers us a useful tool to research the conserved quantities and symmetries of the modified gravity theories, among which the $f(T)$ theory, a generally modified teleparallel gravity, has been proposed to account for the dark energy phenomena. By the Noether symmetry approach, we investigate the power-law, exponential and polynomial forms of $f(T)$ theories. All forms of $f(T)$ concerned in this work possess the time translational symmetry, which is related with energy condition or Hamilton constraint. In addition, we find out that the performances of the power-law and exponential forms are not pleasing. It is rational adding a linear term $T$ to $T^n$ as the most efficient amendment to resemble the teleparallel gravity or General Relativity on small scales, ie., the scale of the solar system. The corresponding Noether symmetry indicates that only time translational symmetry remains. Through numerically calculations and observational data-sets constraining, the optimal form $αT + βT^{-1}$ is obtained, whose cosmological solution resembles the standard $Λ$CDM best with lightly reduced cosmic age which can be alleviated by introducing another $T^m$ term. More important is that we find the significant differences between $Λ$CDM and $f(T)$ gravity. The $Λ$CDM model has also two additional symmetries and corresponding positive conserved quantities, except the two negative conserved quantities.

gr-qc↗

arXiv2012

Birkhoff's Theorem in f(T) Gravity up to the Perturbative Order

f(T) gravity, a generally modified teleparallel gravity, has become very popular in recent times as it is able to reproduce the unification of inflation and late-time acceleration without the need of a dark energy component or an inflation field. In this present work, we investigate specifically the range of validity of Birkhoff's theorem with the general tetrad field via perturbative approach. At zero order, Birkhoff's theorem is valid and the solution is the well known Schwarzschild-(A)dS metric. Then considering the special case of the diagonal tetrad field, we present a new spherically symmetric solution in the frame of f(T) gravity up to the perturbative order. The results with the diagonal tetrad field satisfy the physical equivalence between the Jordan and the so-called Einstein frames, which are realized via conformal transformation, at least up to the first perturbative order.

physics.gen-ph↗

arXiv2012

Extended Birkhoff's Theorem in the f(T) Gravity

The f(T) theory, a generally modified teleparallel gravity, has been proposed as an alternative gravity model to account for the dark energy phenomena. Following our previous work [Xin-he Meng and Ying-bin Wang, EPJC(2011), arXiv:1107.0629v1], we prove that the Birkhoff's theorem holds in a more general context, specifically with the off diagonal tetrad case, in this communication letter. Then, we discuss respectively the results of the external vacuum and internal gravitational field in the f(T) gravity framework, as well as the extended meaning of this theorem. We also investigate the validity of the Birkhoff's theorem in the frame of f(T) gravity via conformal transformation by regarding the Brans-Dicke-like scalar as effective matter, and study the equivalence between both Einstein frame and Jordan frame.

gr-qc↗