arXiv subjects

Zhiyi Li

Publications and source records attributed to Zhiyi Li.

At least 19 recordsLinked to original sources

arXiv2026

The $6-\epsilon$ Expansion for Long-Range Lee--Yang and Percolation Criticality

The crossover from long-range (LR) to short-range (SR) criticality in percolation has remained unsettled because previous renormalization-group (RG) analysis within the $\epsilon'=3\sigma-d$ expansion fixes the anomalous dimension at $\eta=2-\sigma$, whereas SR percolation has $\eta_{\rm SR}<0$ near $d=6$. Sak's matching condition then places the crossover above $\sigma=2$, outside the regime in which the LR interaction dominates. In spatial dimension $d=6-\epsilon$, we formulate a perturbative expansion for the LR $\phi^3$ field theory and perform a one-loop RG analysis throughout the perturbatively accessible nonclassical regime $0<\delta<\epsilon/3$, where $\delta = 2-\sigma$. We derive the one-loop corrections to the critical exponents $\eta$ and $\nu$, which acquire nontrivial dependence on $\epsilon$ and $\delta$. They reduce to their mean-field values at the LR upper critical line and continuously recover the SR $6-\epsilon$ results as $\sigma\to2$. These results support a crossover threshold $\sigma_*=2$ and remove the apparent discontinuity of $\eta$ between the LR and SR values within this framework. The same approach also yields the anomalous and edge exponents of the LR Lee--Yang universality class and $q$-state Potts universality classes with $q<2$.

cond-mat.stat-mech↗

arXiv2026

Statistical Properties of Robust Learning under Distributional Shifts

Distributional shifts arise when the target deployment environment differs from the source environment that generated the training data. Robust learning frameworks such as Distributionally Robust Optimization (DRO) and Robust Satisficing (RS) aim to address this challenge, yet their finite-sample guarantees under such shifts, and their systematic comparison, remain underexplored: existing analyses typically establish guarantees either in the source environment or for adversarial worst-case performance over an ambiguity set. This paper instead studies generalization error in the target environment---the excess loss under the shifted target distribution. Our contributions are threefold. First, we derive finite-sample generalization error bounds in the shifted target environment for both DRO and RS. These bounds explicitly characterize the trade-off between reduced sensitivity to shift and the regularization penalty induced by each method's robustness hyperparameter, and they avoid the curse of dimensionality associated with Wasserstein empirical concentration. Second, when partial shift information such as shift magnitude or direction is available, we propose information-directed hyperparameter calibrations and compare the two methods given the same information. Under these calibrations, and in the partial-information regimes we study, DRO and RS exhibit complementary theoretical and empirical behavior. Finally, we apply the framework to a network lot-sizing problem, using it to interpret how robust policies respond to positive shifts in the demand distribution. Together, these results fill a gap in understanding the statistical properties of robust learning methods under distributional shifts and provide a principled basis for comparing DRO and RS.

stat.ML↗

arXiv2026

Joint Discrete-Continuous Flow Matching for Open-Vocabulary Inverse Design of Multilayer Optical Coatings

Amortized neural inverse design typically remains closed-world: component choices are fixed vocabulary tokens, coordinate grids are frozen at training time, and continuous variables are discretized into sequence tokens. Multilayer optical coatings are an industrially important instance, coupling material sequence, layer thickness and wavelength-dependent response. We present IrisFlow, a query-based, open-vocabulary flow-matching framework instantiated in coatings: the target reflectance/transmittance spectrum, wavelength grid, candidate-material optical constants and layer count are supplied at query time. Candidate materials enter as wavelength-aware optical tokens rather than learned identities; material sequences are sampled by discrete flow matching over the query's candidate bank, thicknesses by continuous flow matching without discretization. A single 136M-parameter model designs 2-100-layer stacks. Across a 224-task benchmark it reconstructs in-distribution targets faithfully and retains same-order accuracy on a 15-material held-out bank without retraining; it reconstructs bands up to 1100 nm beyond its training envelope, designs against analytic application specifications and outperforms an autoregressive baseline on that baseline's material library. With optical constants calibrated to our deposition process, IrisFlow designs four color-displaying coolers, fabricated by ion-assisted evaporation: the three chromatic devices reach a CIEDE2000 color error of 3.1-5.2 while retaining 93-95% solar near-infrared reflectance, demonstrating open-vocabulary design carried through to fabricated coatings.

physics.optics↗

arXiv2026

Structured 4D Latent Predictive Model for Robot Planning

Video predictive models are emerging as a powerful paradigm in robotics, offering a promising path toward task generalization, long-horizon planning, and flexible decision-making. However, prevailing approaches often operate on 2D video sequences, inherently lacking the 3D geometric understanding necessary for precise spatial reasoning and physical consistency. We introduce a Structured 4D Latent Predictive Model, which predicts the evolution of a scene's 3D structure in a structured latent space conditioned on observations and textual instructions. Our representation encodes the scene holistically and can be decoded into diverse 3D formats, enabling a more complete and 3D consistent scene understanding. This structured 4D latent predictive model serves as a planner, generating future scenes that are translated into executable actions by a goal-conditioned inverse dynamics module. Experiments demonstrate that our model generates futures with strong visual quality, substantially better 3D consistency and multi-view coherence compared to state-of-the-art video-based planners. Consequently, our full planning pipeline achieves superior performance on complex manipulation tasks, exhibits robust generalization to novel visual conditions, and proves effective on real-world robotic platforms. Our website is available at https://structured-4d-model.github.io/.

cs.RO↗

arXiv2026

Perturbative Renormalization and Universality Diagram for Long-Range Quantum Criticality

Experimental progress in quantum simulators highlights the role of long-range (LR) interactions in reshaping quantum criticality and stabilizing exotic phases beyond the short-range (SR) paradigm. We study ferromagnetic long-range quantum $O(n)$ models with interactions decaying as $1/r^{d+\sigma}$ and develop a perturbative renormalization-group expansion around the LR--SR boundary by setting $d=3-\epsilon$ and $\sigma=2-\delta$. In this parametrization, the full interacting LR window $2d/3<\sigma<2$ becomes $0<\delta<2\epsilon/3$, and is therefore perturbatively controlled. A two-loop calculation yields explicit expressions, in terms of $\epsilon$, $\delta$, and $n$, for the correlation-length exponent $\nu$ and for the frequency and momentum anomalous dimensions $\eta_\omega$ and $\eta_k$. The resulting exponents reduce to long-range Gaussian scaling at $\sigma=2d/3$ and to SR quantum Wilson-Fisher scaling in the $\sigma \to 2$ limit, thereby identifying $\sigma_*=2$ as the LR--SR boundary within the controlled $3-\epsilon$ expansion. Combining the RG results with scaling boundaries and classical LR analogies, we propose a $(d,\sigma)$ universality diagram for ferromagnetic long-range quantum $O(n)$ criticality and use it as an organizing framework for the phase diagram of long-range quantum spin chains.

cond-mat.stat-mech↗

arXiv2026

Finite-Temperature Spin Exchange-Correlation Kernel of the Uniform Electron Gas

The finite-temperature spin response of the uniform electron gas (UEG) is a fundamental reference for spin-polarized and magnetized electron liquids, including warm dense matter (WDM), yet it remains far less constrained than charge response. Using variational diagrammatic Monte Carlo, we compute the static spin exchange--correlation (XC) kernel $K_{xc}(q;T)$ of the unpolarized UEG at metallic densities across the quantum-degenerate, warm-dense, and classical regimes. The kernel connects smoothly to zero-temperature spin-response parametrizations at low temperature, while heating suppresses the Fermi-surface-scale spin-correlation structure and weakens the XC-driven Stoner enhancement. Its long-wavelength limit provides a direct response test of the spin stiffness implied by thermal local-spin-density-approximation (LSDA) parametrizations, showing low-temperature consistency while exposing a resolved warm-dense residual in current LSDA parametrizations. In the classical regime, the spin XC kernel becomes nearly local on the Fermi-momentum scale, in sharp contrast to the corresponding charge XC kernel. These results provide a first-principles basis for finite-temperature spin-response theory and magnetized WDM modeling.

cond-mat.str-el↗

arXiv2026

Effect of Solar Flares on Decayless Kink Oscillations in Nearby Coronal Loops

We present a statistical study of 130 solar flares (B to X class) that lack soft X-ray quasi-periodic pulsations and show no kink oscillations of nearby coronal loops visible in SDO/AIA 171~\AA~images. The aim is to investigate whether decayless kink oscillations of coronal loops respond to nearby flaring activity. Using the Fractional Anisotropy-based Video Motion Magnification technique, we detected low-amplitude decayless oscillations in all 130 loops before, during, and after each flare, confirming their ubiquitous nature. Oscillation periods are found to range from 122~s to 268~s, and the projected displacement amplitudes are 0.023--0.111~Mm. No amplitude--period correlation is found. For each event, we estimated the amplitude before, during, and after the flare. Across all flare classes, the average amplitude remains unchanged. However, in some specific cases, the oscillation amplitude may exhibit minor changes. For B-, C-, and M-class flares, the fraction of events with an amplitude change exceeding 10% is approximately 23%, 41%, and 36%, respectively. In M-class flares, such minor amplitude increases occur four times more often than decreases; in X-class flares (only six events), decreases dominate by a factor of three. The fraction of events that exhibit an increase in the amplitude of more than 20% appears to be highest when the loop centre is located at a distance of 100--120~Mm from the flare site, reaching 33% (6 out of 18 events). Overall, the amplitude of decayless kink oscillations does not undergo a major change in response to nearby flares, especially for less powerful classes, suggesting that flare-related processes such as blast waves and reconnection inflows have little effect on the energy supply to oscillating loops.

astro-ph.SR↗

arXiv2026

First-Principles Effective Mass in the Three-Dimensional Uniform Electron Gas

The quasiparticle effective mass $m^*$ of the three-dimensional uniform electron gas (UEG) is a fundamental Fermi-liquid parameter whose value and density dependence have remained controversial for decades. Using renormalized perturbation theory with explicit counterterms, we determine $m^*$ in the metallic regime ($r_s \le 6$) from first principles by two complementary routes -- the self-energy and the forward-scattering four-point vertex via the $p$-wave spin-symmetric Landau parameter $F_1^s$ -- that agree within uncertainties at each density through sixth renormalized order. The resulting $m^*/m$ remains close to unity throughout the metallic regime, with a shallow non-monotonic density dependence -- a minimum near $r_s\approx 1$ followed by a gentle upturn -- reflecting the interplay of exchange and dynamical screening in the self-energy, and disfavoring strong monotonic suppression. This finding supports a physical picture for the metallic UEG in which dominant charge correlations are concentrated in nearly forward scattering and generate only a weak $F_1^s$ component.

cond-mat.str-el↗

arXiv2026

Self Knowledge Re-expression: A Fully Local Method for Adapting LLMs to Tasks Using Intrinsic Knowledge

While the next-token prediction (NTP) paradigm enables large language models (LLMs) to express their intrinsic knowledge, its sequential nature constrains performance on specialized, non-generative tasks. We attribute this performance bottleneck to the LLMs' knowledge expression mechanism, rather than to deficiencies in knowledge acquisition. To address this, we propose Self-Knowledge Re-expression (SKR), a novel, task-agnostic adaptation method. SKR transforms the LLM's output from generic token generation to highly efficient, task-specific expression. SKR is a fully local method that uses only unannotated data, requiring neither human supervision nor model distillation. Experiments on a large financial document dataset demonstrate substantial improvements: over 40% in Recall@1 for information retrieval tasks, over 76% reduction in object detection latency, and over 33% increase in anomaly detection AUPRC. Our results on the MMDocRAG dataset surpass those of leading retrieval models by at least 12.6%.

cs.CL↗

arXiv2026

Percolation in the three-dimensional Ising model

Geometric representations provide a useful perspective on critical phenomena in the Ising model. In a recent study [Phys. Rev. E 112, 034118 (2025)], we found that the two-dimensional critical Ising model exhibits two consecutive percolation transitions for geometric spin clusters as the bond-occupation probability $p$ between parallel spins increases. Here, through extensive Monte Carlo simulations, we show that this phenomenon does not persist in three dimensions, where we observe only a single percolation transition on critical Ising configurations. Further theoretical analysis of the Ising model on the complete graph also yields the same scenario. In addition, we study percolation on a two-dimensional layer embedded in the three-dimensional critical Ising model. For this layer system, we estimate the red-bond exponent $y_p = 0.426(6)$ and the fractal dimensions of the largest cluster, hull, and shortest path as $d_f = 1.8926(20)$, $d_{\rm hull} = 1.663(4)$, and $d_{\rm min} = 1.080(10)$, respectively. These values indicate a distinct universality class induced by coupling to out-of-plane critical correlations.

cond-mat.stat-mech↗

arXiv2026

Unraveling Intrinsic Thermal Conductivity in Layered Conductive MOF Single Crystals

Layered conductive metal-organic frameworks (LCMOFs) show great promise in energy and electronics due to their high electrical conductivity and tunable pore structures. They are considered ideal "phonon-glass, electron-crystal" materials. However, their intrinsic thermal transport properties, particularly the thermal conductivity in the single-crystalline state, have never been explored before. The applicability of the Wiedemann-Franz law to such complex porous materials is a key scientific question to describe their thermoelectric relationship. We investigated single crystals of three LCMOFs (Cu3HHTP2, Co9HHTP4, Nd3HHTP2) using the microfabricated suspended device. Results showed ultralow thermal conductivities (0.075-0.194 W m-1 K-1) along the {\pi}-{\pi} stacking direction. Crucially, Nd3HHTP2 exhibited a high electrical conductivity of 398 S cm-1, yet its thermal conductivity (0.148 W m-1 K-1) was comparable to the other two LCMOFs with significantly lower electrical conductivities. Structural characterization revealed that the incommensurate modulation, and in-plane correlated disorder within the Nd3HHTP2 structure are the potential causes of strong phonon scattering and the observed ultralow thermal conductivity.

cond-mat.mtrl-sci↗

arXiv2026

From $\alpha$ decay to cluster decay: an extreme case of transfer learning

When training data are limited, data-driven models are especially vulnerable to optimization-related fluctuations from random initialization and to sampling-induced bias from insufficient training data. We address both challenges with transfer learning (TL): deep neural networks (DNNs) are first pretrained on $\alpha$ decay half-lives and then fine-tuned on a small cluster decay dataset. The pretraining stage provides a physically informed initialization that stabilizes optimization, while transferred global decay systematics regularize the fit and reduce sensitivity to training set composition. Despite extreme data sparsity, the resulting models accurately predict cluster decay half-lives for parent nuclei from $^{221}$Fr to $^{242}$Cm. We further quantify how initialization and sample selection affect predictive accuracy and robustness, demonstrating that TL enables stable and reliable learning in the small-sample regime.

nucl-th↗

arXiv2026

The 4-$\epsilon$ Expansion for Long-range Interacting Systems

The establishment of the Wilson-Fisher fixed point (WFP) for $O(n)$ spin models in $d=4-\epsilon$ dimensions stands as a cornerstone of the renormalization group (RG) theory for critical phenomena. However, when long-range (LR) interactions, algebraically decaying as $\propto 1/r^{d+\sigma}$, are introduced, the fate of the short-range WFP (SR-WFP) has remained a subject of intense debate since the 1970s. We employ two complementary techniques -- the standard field-theoretic RG and a perturbative bootstrap scheme, and perform the $\epsilon$-expansion calculations up to the two-loop level. We show that, as long as $\sigma<2$, the SR-WFP becomes unstable and a stable LR-WFP emerges, and, in the non-classical regime with $d/2 < \sigma < 2$, the critical exponents, including the anomalous dimension, are functions of $\epsilon$, $\delta=2-\sigma$ and $n$, which reduce to the exact results in the limiting cases $\epsilon \to 0$, $\delta \to 0$ or $n \to \infty$. Our $(4-\epsilon)$-expansion calculations support the scenario that the threshold between the LR- and SR-WFP occurs strictly at $\sigma_*=2$, well consistent with the recent high-precision numerical study while different from the widely accepted Sak's criterion.

cond-mat.stat-mech↗

arXiv2025

Analysis of Frequency and Voltage Strength in Power Electronics-Dominated Power Systems Based on Eigen-subsystems

The large-scale integration of inverter-based resources (IBRs) has deteriorated the frequency/voltage (F/V) responses of power systems, leading to a higher risk of instability. Consequently, evaluating the F/V strength has become an important task in power electronics (PE)-dominated power systems. Existing methods typically examine F/V strength separately, employing fundamentally different metrics, such as inertia (focusing on device dynamics) and short-circuit ratio (SCR, addressing network characteristics). These fragmented approaches have resulted in a lack of comprehensive understanding of the overall system strength, potentially overlooking critical aspects. To address this problem, this paper proposes a unified framework for analyzing F/V strength. First, a unified modeling of F/V regulations is introduced. Then, based on modal decoupling, the power systems are decomposed into several eigen-subsystems, where the F/V responses are both decomposed into common-mode (CM) and differential-mode (DM) components, namely, CM-F, DM-F, CM-V, and DM-V. The CM-F and CM-V represent the collective response of all devices to external active or reactive power disturbances, independent of the power network characteristics. In contrast, the DM-F and DM-V capture the redistribution of disturbance power within the system, which is strongly influenced by the network topology and the locations of devices. Notably, traditional strength analysis generally ignores the CM-V (global voltage response), which, as discovered in this paper, may also become unstable in PE-dominated power systems. Based on the proposed framework, new metrics are proposed to evaluate the strength of each modal component. Finally, the effectiveness of the proposed approach is validated through simulations.

eess.SY↗

arXiv2025

Two-Electron Correlations in the Metallic Electron Gas

We present high-precision \emph{ab initio} calculations of the four-point vertex function for the three-dimensional uniform electron gas using variational diagrammatic Monte Carlo. From these results, we extract Landau parameters that reveal a density-driven crossover from underscreening to overscreening, and obtain the full two-electron scattering amplitude on the Fermi surface with controlled accuracy. A residual analysis of the scattering amplitude against the charge-channel Kukkonen--Overhauser (KO$^+$) interaction shows that only a minimal s-wave correction in the antiparallel-spin channel is needed, defining the sKO$^+$ ansatz: KO$^+$ within the local-density approximation plus this short-range correction. Using both our direct VDMC amplitudes and the sKO$^+$ ansatz, we compute the electron-electron contribution to the thermal resistivity, obtaining quantitative agreement with experiments on simple metals (Al, Na, K, Rb). sKO$^+$ thus provides a controlled UEG-based effective interaction for simple-metal transport and future first-principles extensions.

cond-mat.str-el↗

arXiv2025

Generalized Dynamics Generation towards Scannable Physical World Model

Digital twin worlds with realistic interactive dynamics presents a new opportunity to develop generalist embodied agents in scannable environments with complex physical behaviors. To this end, we present GDGen (Generalized Representation for Generalized Dynamics Generation), a framework that takes a potential energy perspective to seamlessly integrate rigid body, articulated body, and soft body dynamics into a unified, geometry-agnostic system. GDGen operates from the governing principle that the potential energy for any stable physical system should be low. This fresh perspective allows us to treat the world as one holistic entity and infer underlying physical properties from simple motion observations. We extend classic elastodynamics by introducing directional stiffness to capture a broad spectrum of physical behaviors, covering soft elastic, articulated, and rigid body systems. We propose a specialized network to model the extended material property and employ a neural field to represent deformation in a geometry-agnostic manner. Extensive experiments demonstrate that GDGen robustly unifies diverse simulation paradigms, offering a versatile foundation for creating interactive virtual environments and training robotic agents in complex, dynamically rich scenarios.

cs.CV↗

arXiv2024

Orb: A Fast, Scalable Neural Network Potential

We introduce Orb, a family of universal interatomic potentials for atomistic modelling of materials. Orb models are 3-6 times faster than existing universal potentials, stable under simulation for a range of out of distribution materials and, upon release, represented a 31% reduction in error over other methods on the Matbench Discovery benchmark. We explore several aspects of foundation model development for materials, with a focus on diffusion pretraining. We evaluate Orb as a model for geometry optimization, Monte Carlo and molecular dynamics simulations.

cond-mat.mtrl-sci↗

arXiv2024

Comparative Analysis of Extrinsic Factors for NER in French

Named entity recognition (NER) is a crucial task that aims to identify structured information, which is often replete with complex, technical terms and a high degree of variability. Accurate and reliable NER can facilitate the extraction and analysis of important information. However, NER for other than English is challenging due to limited data availability, as the high expertise, time, and expenses are required to annotate its data. In this paper, by using the limited data, we explore various factors including model structure, corpus annotation scheme and data augmentation techniques to improve the performance of a NER model for French. Our experiments demonstrate that these approaches can significantly improve the model's F1 score from original CRF score of 62.41 to 79.39. Our findings suggest that considering different extrinsic factors and combining these techniques is a promising approach for improving NER performance where the size of data is limited.

cs.CL↗