Searcharxiv⌕ Search

arXiv subjects

Gang Chen

Publications and source records attributed to Gang Chen.

At least 271 records · Page 15Linked to original sources
arXiv2023

ACQ: Improving Generative Data-free Quantization Via Attention Correction

Data-free quantization aims to achieve model quantization without accessing any authentic sample. It is significant in an application-oriented context involving data privacy. Converting noise vectors into synthetic samples through a generator is a popular data-free quantization method, which is called generative data-free quantization. However, there is a difference in attention between synthetic samples and authentic samples. This is always ignored and restricts the quantization performance. First, since synthetic samples of the same class are prone to have homogenous attention, the quantized network can only learn limited modes of attention. Second, synthetic samples in eval mode and training mode exhibit different attention. Hence, the batch-normalization statistics matching tends to be inaccurate. ACQ is proposed in this paper to fix the attention of synthetic samples. An attention center position-condition generator is established regarding the homogenization of intra-class attention. Restricted by the attention center matching loss, the attention center position is treated as the generator's condition input to guide synthetic samples in obtaining diverse attention. Moreover, we design adversarial loss of paired synthetic samples under the same condition to prevent the generator from paying overmuch attention to the condition, which may result in mode collapse. To improve the attention similarity of synthetic samples in different network modes, we introduce a consistency penalty to guarantee accurate BN statistics matching. The experimental results demonstrate that ACQ effectively improves the attention problems of synthetic samples. Under various training settings, ACQ achieves the best quantization performance. For the 4-bit quantization of Resnet18 and Resnet50, ACQ reaches 67.55% and 72.23% accuracy, respectively.

cs.CV↗
arXiv2023

Rethinking loss of available work and Gouy-Stodola theorem

Exergy represents the maximum useful work possible when a system at a specific state reaches equilibrium with the environmental dead state at temperature To. Correspondingly, the exergy difference between two states is the maximum work output when the system changes from one state to the other, assuming that during the processes, the system exchanges heat reversibly with the environment. If the process involves irreversibility, the Guoy-Stodola theorem states that the exergy destruction equals the entropy generated during the process multiplied by To. The exergy concept and the Gouy-Stodola theorem are widely used to optimize processes or systems, even when they are not directly connected to the environment. In the past, questions have been raised on if To is the proper temperature to use in calculating the exergy destruction. Here, we start from the first and the second laws of thermodynamics to unambiguously show that the useful energy loss (UEL) of a system or process should equal to the entropy generation multiplied by an equivalent temperature associated with the entropy rejected out of the entire system. For many engineering systems and processes, this entropy rejection temperature can be easily calculated as the ratio of the changes of the enthalpy and entropy of the fluid stream carrying the entropy out, which we call the state-change temperature. The UEL is unambiguous and independent of the environmental dead state, and it should be used for system optimization rather than the exergy destruction.

cond-mat.stat-mech↗
arXiv2023

Interfacial cooling and heating, temperature discontinuity and inversion in evaporation and condensation

Although ubiquitous in nature and industrial processes, transport processes at the interface during evaporation and condensation are still poorly understood. Experiments have shown temperature discontinuities at the interface during evaporation and condensation but the experimentally reported interface temperature jump varies by two orders of magnitude. Even the direction of such temperature jump is still being debated. Using kinetic-theory based expressions for the interfacial mass flux and heat flux, we solve coupled problem between the liquid and the vapor phase during evaporation and condensation. Our model shows that when evaporation or condensation happens, an intrinsic temperature difference develops across the interface, due to the mismatch of the enthalpy carried by vapor at the interface and the bulk region. The vapor temperature near the interface cools below the saturation temperature on the liquid surface during evaporation and heats up above the latter during condensation. However, many existing experiments have shown an opposite trend to this prediction. We explain this difference as arising from the reverse heat conduction in the vapor phase. Our model results compare favorably with experiments on both evaporation and condensation. We show that when the liquid layer is very thin, most of the applied temperature difference between the solid wall and the vapor phase happens at the liquid-vapor interface, leading to saturation of the evaporation and the condensation rates and the corresponding heat transfer rate. This result contradicts current belief that the evaporation and condensation rates are inversely proportional to the liquid film thickness.

cond-mat.soft↗
arXiv2023

Double exchange, itinerant ferromagnetism and topological Hall effect in moiré heterobilayer

Motivated by the recent experiments and the wide tunability on the MoTe$_2$/WSe$_2$ moiré heterobilayer, we consider a physical model to explore the underlying physics for the interplay between the itinerant carriers and the local magnetic moments. In the regime where the MoTe$_2$ is tuned to a triangular lattice Mott insulator and the WSe$_2$ layer is doped with the itinerant holes, we invoke the itinerant ferromagnetism from the double exchange mechanism for the itinerant holes on the WSe$_2$ layer and the local moments on the MoTe$_2$ layer. Together with the antiferromagnetic exchange on the MoTe$_2$ layer, the itinerant ferromagnetism generates the scalar spin chirality distribution in the system. We further point out the presence of spin-assisted hopping in addition to the Kondo coupling between the local spin and the itinerant holes, and demonstrate the topological Hall effect for the itinerant electrons in the presence of the non-collinear spin configurations. This work may improve our understanding of the correlated moiré systems and inspire further experimental efforts.

cond-mat.str-el↗
arXiv2023

A mixed policy to improve performance of language models on math problems

When to solve math problems, most language models take a sampling strategy to predict next word according conditional probabilities. In the math reasoning step, it may generate wrong answer. Considering math problems are deterministic, we propose a mixed policy exploration approach to solve math problems with reinforcement learning. In peculiar, we propose a two level token exploration policy: the abstract level explores next token with probability and the second level is deterministic. Specifically, the abstract level policy will decide whether the token is operator or operand with probability sampling, while the second level is deterministic to select next token with the highest score in a greedy way. We test our method on GSM8K dataset with GPT-2 model, and demonstrate more than $2\%$ performance gain. Our implementation is available at https://github.com/vividitytech/math_lm_rl.

cs.CL↗
arXiv2023

Deep learning-based reduced order model for three-dimensional unsteady flow using mesh transformation and stitching

Artificial intelligence-based three-dimensional(3D) fluid modeling has gained significant attention in recent years. However, the accuracy of such models is often limited by the processing of irregular flow data. In order to bolster the credibility of near-wall flow prediction, this paper presents a deep learning-based reduced order model for three-dimensional unsteady flow using the transformation and stitching of multi-block structured meshes. To begin with, full-order flow data is provided by numerical simulations that rely on multi-block structured meshes. A mesh transformation technique is applied to convert each structured grid with data into a corresponding uniform and orthogonal grid, which is subsequently stitched and filled. The resulting snapshots in the transformed domain contain accurate flow information for multiple meshes and can be directly fed into a structured neural network without requiring any interpolation operation. Subsequently, a network model based on a fully convolutional neural network is constructed to predict flow dynamics accurately. To validate the strategy's feasibility, the flow around a sphere with Re=300 was investigated, and the results obtained using traditional Cartesian interpolation were used as the baseline for comparison. All the results demonstrate the preservation and accurate prediction of flow details near the wall, with the pressure correlation coefficient on the wall achieving an impressive value of 0.9985. Moreover, the periodic behavior of flow fields can be faithfully predicted during long-term inference.

physics.flu-dyn↗
arXiv2023

Locking-free HDG methods for Reissner-Mindlin plates equations on polygonal meshes

We present and analyze a new hybridizable discontinuous Galerkin method (HDG) for the Reissner-Mindlin plate bending system. Our method is based on the formulation utilizing Helmholtz Decomposition. Then the system is decomposed into three problems: two trivial Poisson problems and a perturbed saddle-point problem. We apply HDG scheme for these three problems fully. This scheme yields the optimal convergence rate ($(k+1)$th order in the $\mathrm{L}^2$ norm) which is uniform with respect to plate thickness (locking-free) on general meshes. We further analyze the matrix properties and precondition the new finite element system. Numerical experiments are presented to confirm our theoretical analysis.

math.NA↗
arXiv2023

Anisotropic in-plane heat transport of Kitaev magnet Na$_2$Co$_2$TeO$_6$

We report a study on low-temperature heat transport of Kitaev magnet Na$_2$Co$_2$TeO$_6$, with the heat current and magnetic fields along the honeycomb spin layer (the $ab$ plane). The zero-field thermal conductivity of $κ^a_{xx}$ and $κ^{a*}_{xx}$ display similar temperature dependence and small difference in their magnitudes; whereas, their magnetic field (parallel to the heat current) dependence are quite different and are related to the field-induced magnetic transitions. The $κ^a_{xx}(B)$ data for $B \parallel a$ at very low temperatures have an anomaly at 10.25--10.5 T, which reveals an unexplored magnetic transition. The planar thermal Hall conductivity $κ^a_{xy}$ and $κ^{a*}_{xy}$ show very weak signals at low fields and rather large values with sign change at high fields. This may point to a possible magnetic structure transition or the change of the magnon band topology that induces a radical change of magnon Berry curvature distribution before entering the spin polarized state. These results put clear constraints on the high-field phase and the theoretical models for Na$_2$Co$_2$TeO$_6$.

cond-mat.str-el↗
arXiv2023

Classical Spin Gravitational Compton Scattering

We introduce a novel bootstrap method for classical Compton scattering amplitudes involving two massless gluon/graviton particles and two arbitrary-spin infinite-mass particles in a heavy-mass effective field theory context. Using a suitable ansatz, we deduce new and explicit classical spin results for gluon four and five-point infinite mass processes that exhibit exponentiated three-point factorizations to all orders in spin and feature no spurious poles. We discuss the generalization of our bootstrap to higher multiplicities and summarize future potential applications.

hep-th↗
arXiv2023

Exchange renormalized crystal field excitation in a quantum Ising magnet KTmSe$_2$

Rare-earth delafossite compounds, ARCh$_2$ (A = alkali or monovalent ion, R = rare earth, Ch = chalcogen), have been proposed for a range of novel quantum phenomena. Particularly, the Tm series, ATmCh$_2$, featuring Tm ions on a triangular lattice, serves as a representative group of compounds to illustrate the interplay and competition between spin-orbit coupling, crystal fields, and exchange couplings in the presence of geometric frustration. Here we report the thermodynamic and inelastic neutron scattering studies on the newly discovered triangular-lattice magnet KTmSe$_2$. Both heat capacity and neutron diffraction reveal the absence of long-range magnetic order. Magnetic susceptibility shows strong Ising-like interactions with antiferromagnetic correlations. Furthermore, inelastic neutron scattering measurements reveal a branch of dispersive crystal field excitations. To analyze these observations, we employ both the transverse field Ising model and the full crystal field scheme, along with exchange interactions. Our results suggest a strong competition between spin exchange interactions and crystal field effects. This work is expected to offer a valuable framework for understanding low-temperature magnetism in KTmSe$_2$ and similar materials.

cond-mat.str-el↗
arXiv2023

Byzantine-Robust Learning on Heterogeneous Data via Gradient Splitting

Federated learning has exhibited vulnerabilities to Byzantine attacks, where the Byzantine attackers can send arbitrary gradients to a central server to destroy the convergence and performance of the global model. A wealth of robust AGgregation Rules (AGRs) have been proposed to defend against Byzantine attacks. However, Byzantine clients can still circumvent robust AGRs when data is non-Identically and Independently Distributed (non-IID). In this paper, we first reveal the root causes of performance degradation of current robust AGRs in non-IID settings: the curse of dimensionality and gradient heterogeneity. In order to address this issue, we propose GAS, a \shorten approach that can successfully adapt existing robust AGRs to non-IID settings. We also provide a detailed convergence analysis when the existing robust AGRs are combined with GAS. Experiments on various real-world datasets verify the efficacy of our proposed GAS. The implementation code is provided in https://github.com/YuchenLiu-a/byzantine-gas.

cs.LG↗
arXiv2023

Systematic study of hadrons and their quark-component nuclear modification factors

We have systematically studied the connection between hadron and its quark component nuclear modification factors and the flavor (mass) ordering at both parton and hadron levels in the nucleus-nucleus collisions at the LHC energies by the PACIAE model. It turns out that these two physical phenomena explored in our last publication (J. Phys. G 49 065104, 2022) are generally held, irrespective of the rapidity, centrality, reaction energy, and the collision system size. The mass ordering at hadron level in nuclear modification factor seems to be really existed, which should be studied further both theoretically and experimentally.

hep-ph↗
arXiv2023

Deep Metric Tensor Regularized Policy Gradient

Policy gradient algorithms are an important family of deep reinforcement learning techniques. Many past research endeavors focused on using the first-order policy gradient information to train policy networks. Different from these works, we conduct research in this paper driven by the believe that properly utilizing and controlling Hessian information associated with the policy gradient can noticeably improve the performance of policy gradient algorithms. One key Hessian information that attracted our attention is the Hessian trace, which gives the divergence of the policy gradient vector field in the Euclidean policy parametric space. We set the goal to generalize this Euclidean policy parametric space into a general Riemmanian manifold by introducing a metric tensor field $g_ab$ in the parametric space. This is achieved through newly developed mathematical tools, deep learning algorithms, and metric tensor deep neural networks (DNNs). Armed with these technical developments, we propose a new policy gradient algorithm that learns to minimize the absolute divergence in the Riemannian manifold as an important regularization mechanism, allowing the Riemannian manifold to smoothen its policy gradient vector field. The newly developed algorithm is experimentally studied on several benchmark reinforcement learning problems. Our experiments clearly show that the new metric tensor regularized algorithm can significantly outperform its counterpart that does not use our regularization technique. Additional experimental analysis further suggests that the trained metric tensor DNN and the corresponding metric tensor $g_{ab}$ can effectively reduce the absolute divergence towards zero in the Riemannian manifold.

cs.LG↗
arXiv2023

Super-Resolution Imaging via Angular Magnification

The far-field resolution of optical imaging systems is restricted by the Abbe diffraction limit, a direct result of the wave nature of light. One successful technological approach to circumventing this limit is to reduce the effective size of a point-spread-function. In the past decades, great endeavors have been made to engineer an effective point-spread-function by exploiting different mechanisms, including optical nonlinearities and structured light illumination. However, these methods are hard to be applied to objects in a far distance. Here, we propose a new way to achieve super-resolution in a far field by utilizing angular magnification. We present the first proof-of-concept demonstration of such an idea and demonstrate a new class of lenses with angular magnification for far-field super-resolution imaging. Both theoretical and experimental results demonstrate a more than two-fold enhancement beyond the angular-resolution limit in the far-field imaging. The proposed approach can be applied to super-resolution imaging of objects in far distance. It has promising potential applications in super-resolution telescopes and remote sensing.

physics.optics↗
arXiv2023

Schur rings over infinite dihedral group

Schur rings over the infinite dihedral group $\mathcal{Z}\rtimes\mathcal{Z}_2$ are studied according to properties of Schur rings over infinite groups and the classification of Schur rings over infinite cyclic groups. Schur rings over $\mathcal{Z}\rtimes{\mathcal{Z}}_2$ are classified under the assumption that $\mathcal{Z}$ is an $\mathcal{A}$-subgroup. Those Schur rings are proved to be traditional.

math.CO↗
arXiv2023

Point convolutional neural network algorithm for Ising model ground state research based on spring vibration

The ground state search of the Ising model can be used to solve many combinatorial optimization problems. Under the current computer architecture, an Ising ground state search algorithm suitable for hardware computing is necessary for solving practical problems. Inspired by the potential energy conversion of springs, we propose a point convolutional neural network algorithm for ground state search based on spring vibration model, called Spring-Ising Algorithm. Spring-Ising Algorithm regards the spin as a moving mass point connected to a spring and establish the equation of motion for all spins. Spring-Ising Algorithm can be mapped on the GPU or AI chips through the basic structure of the neural network for fast and efficient parallel computing. The algorithm has very productive results for solving the Ising model and has been test in the recognized test benchmark K2000. The algorithm introduces the concept of dynamic equilibrium to achieve a more detailed local search by dynamically adjusting the weight of the Ising model in the spring oscillation model. Finally, there is the simple hardware test speed evaluation. Spring-Ising Algorithm can provide the possibility to calculate the Ising model on a chip which focuses on accelerating neural network calculations.

physics.comp-ph↗
arXiv2023

Deep Partial Multi-Label Learning with Graph Disambiguation

In partial multi-label learning (PML), each data example is equipped with a candidate label set, which consists of multiple ground-truth labels and other false-positive labels. Recently, graph-based methods, which demonstrate a good ability to estimate accurate confidence scores from candidate labels, have been prevalent to deal with PML problems. However, we observe that existing graph-based PML methods typically adopt linear multi-label classifiers and thus fail to achieve superior performance. In this work, we attempt to remove several obstacles for extending them to deep models and propose a novel deep Partial multi-Label model with grAph-disambIguatioN (PLAIN). Specifically, we introduce the instance-level and label-level similarities to recover label confidences as well as exploit label dependencies. At each training epoch, labels are propagated on the instance and label graphs to produce relatively accurate pseudo-labels; then, we train the deep model to fit the numerical labels. Moreover, we provide a careful analysis of the risk functions to guarantee the robustness of the proposed model. Extensive experiments on various synthetic datasets and three real-world PML datasets demonstrate that PLAIN achieves significantly superior results to state-of-the-art methods.

cs.LG↗