Searcharxiv⌕ Search

arXiv subjects

Dong Li

Publications and source records attributed to Dong Li.

At least 361 records · Page 20Linked to original sources
arXiv2022

A Statistical Study of Short-period Decayless Oscillations of Coronal Loops in an Active Region

Coronal loop oscillations are common phenomena in the solar corona, which are often classified as decaying and decayless oscillations. Using the high-resolution observation measured by the Extreme Ultraviolet Imager (EUI) onboard the Solar Orbiter, we statistical investigate small-scale transverse oscillations with short periods (<200 s) of coronal loops in an active region, i.e., NOAA 12965. A total of 111 coronal loops are identified in EUI 174 A images, and they all reveal transverse oscillations without any significant decaying, regarding as decayless oscillations. Oscillatory periods are measured from about 11 s to 185 s, with a median period of 40 s. Thus, they are also termed as short-period oscillations. The corresponding loop lengths are measured from about 10.5 Mm to 30.2 Mm, and a strong dependence of oscillatory periods on loop lengths is established, indicating that the short-period oscillations are standing kink-mode waves in nature. Based on the coronal seismology, kink speeds are measured to about 330-1910 km/s, and magnetic field strengths in coronal loops are estimated to about 4.1-25.2 G, while the energy flux carried by decayless kink oscillations lies in the range from roughly 7 W m^(-2) to 9220 W m^(-2). Our estimations suggest that the wave energy carried by short-period decayless kink oscillations can not support the coronal heating in the active region.

astro-ph.SR↗
arXiv2022

Quasi two-dimensional nature of high-Tc superconductivity in iron-based (Li,Fe)OHFeSe

The intercalated iron selenide (Li,Fe)OHFeSe has a strongly layered structure analogous to the quasi two-dimensional (2D) bismuth cuprate superconductors, and exhibits both high-temperature (Tc) and topological superconductivity. However, the issue of its superconductivity dimensionality has not yet been fully investigated so far. Here we report that the quasi-2D superconductivity features, including the high anisotropy γ = 151 and the associated quasi-2D vortices, are also revealed for (Li,Fe)OHFeSe, based on systematic experiments of the electrical transport and magnetization and model fittings. Thus, we establish a new vortex phase diagram for (Li,Fe)OHFeSe, which delineates an emergent quasi-2D vortex-liquid state, and a subsequent vortex-solid dimensional crossover from a pancake-like to a three-dimensional state with decreasing temperature and magnetic field. Furthermore, we find that all the quasi-2D characteristics revealed here for the high-Tc iron selenide superconductor are very similar to those reported for the high-Tc bismuth cuprate superconductors.

cond-mat.supr-con↗
arXiv2022

State-Aware Proximal Pessimistic Algorithms for Offline Reinforcement Learning

Pessimism is of great importance in offline reinforcement learning (RL). One broad category of offline RL algorithms fulfills pessimism by explicit or implicit behavior regularization. However, most of them only consider policy divergence as behavior regularization, ignoring the effect of how the offline state distribution differs with that of the learning policy, which may lead to under-pessimism for some states and over-pessimism for others. Taking account of this problem, we propose a principled algorithmic framework for offline RL, called \emph{State-Aware Proximal Pessimism} (SA-PP). The key idea of SA-PP is leveraging discounted stationary state distribution ratios between the learning policy and the offline dataset to modulate the degree of behavior regularization in a state-wise manner, so that pessimism can be implemented in a more appropriate way. We first provide theoretical justifications on the superiority of SA-PP over previous algorithms, demonstrating that SA-PP produces a lower suboptimality upper bound in a broad range of settings. Furthermore, we propose a new algorithm named \emph{State-Aware Conservative Q-Learning} (SA-CQL), by building SA-PP upon representative CQL algorithm with the help of DualDICE for estimating discounted stationary state distribution ratios. Extensive experiments on standard offline RL benchmark show that SA-CQL outperforms the popular baselines on a large portion of benchmarks and attains the highest average return.

cs.LG↗
arXiv2022

Prototypical context-aware dynamics generalization for high-dimensional model-based reinforcement learning

The latent world model provides a promising way to learn policies in a compact latent space for tasks with high-dimensional observations, however, its generalization across diverse environments with unseen dynamics remains challenging. Although the recurrent structure utilized in current advances helps to capture local dynamics, modeling only state transitions without an explicit understanding of environmental context limits the generalization ability of the dynamics model. To address this issue, we propose a Prototypical Context-Aware Dynamics (ProtoCAD) model, which captures the local dynamics by time consistent latent context and enables dynamics generalization in high-dimensional control tasks. ProtoCAD extracts useful contextual information with the help of the prototypes clustered over batch and benefits model-based RL in two folds: 1) It utilizes a temporally consistent prototypical regularizer that encourages the prototype assignments produced for different time parts of the same latent trajectory to be temporally consistent instead of comparing the features; 2) A context representation is designed which combines both the projection embedding of latent states and aggregated prototypes and can significantly improve the dynamics generalization ability. Extensive experiments show that ProtoCAD surpasses existing methods in terms of dynamics generalization. Compared with the recurrent-based model RSSM, ProtoCAD delivers 13.2% and 26.7% better mean and median performance across all dynamics generalization tasks.

cs.LG↗
arXiv2022

Breaking the Curse of Dimensionality in Multiagent State Space: A Unified Agent Permutation Framework

The state space in Multiagent Reinforcement Learning (MARL) grows exponentially with the agent number. Such a curse of dimensionality results in poor scalability and low sample efficiency, inhibiting MARL for decades. To break this curse, we propose a unified agent permutation framework that exploits the permutation invariance (PI) and permutation equivariance (PE) inductive biases to reduce the multiagent state space. Our insight is that permuting the order of entities in the factored multiagent state space does not change the information. Specifically, we propose two novel implementations: a Dynamic Permutation Network (DPN) and a Hyper Policy Network (HPN). The core idea is to build separate entity-wise PI input and PE output network modules to connect the entity-factored state space and action space in an end-to-end way. DPN achieves such connections by two separate module selection networks, which consistently assign the same input module to the same input entity (guarantee PI) and assign the same output module to the same entity-related output (guarantee PE). To enhance the representation capability, HPN replaces the module selection networks of DPN with hypernetworks to directly generate the corresponding module weights. Extensive experiments in SMAC, Google Research Football and MPE validate that the proposed methods significantly boost the performance and the learning efficiency of existing MARL algorithms. Remarkably, in SMAC, we achieve 100% win rates in almost all hard and super-hard scenarios (never achieved before).

cs.LG↗
arXiv2022

Linear Video Transformer with Feature Fixation

Vision Transformers have achieved impressive performance in video classification, while suffering from the quadratic complexity caused by the Softmax attention mechanism. Some studies alleviate the computational costs by reducing the number of tokens in attention calculation, but the complexity is still quadratic. Another promising way is to replace Softmax attention with linear attention, which owns linear complexity but presents a clear performance drop. We find that such a drop in linear attention results from the lack of attention concentration on critical features. Therefore, we propose a feature fixation module to reweight the feature importance of the query and key before computing linear attention. Specifically, we regard the query, key, and value as various latent representations of the input token, and learn the feature fixation ratio by aggregating Query-Key-Value information. This is beneficial for measuring the feature importance comprehensively. Furthermore, we enhance the feature fixation by neighborhood association, which leverages additional guidance from spatial and temporal neighbouring tokens. The proposed method significantly improves the linear attention baseline and achieves state-of-the-art performance among linear video Transformers on three popular video classification benchmarks. With fewer parameters and higher efficiency, our performance is even comparable to some Softmax-based quadratic Transformers.

cs.CV↗
arXiv2022

SA: Sliding attack for synthetic speech detection with resistance to clipping and self-splicing

Deep neural networks are vulnerable to adversarial examples that mislead models with imperceptible perturbations. In audio, although adversarial examples have achieved incredible attack success rates on white-box settings and black-box settings, most existing adversarial attacks are constrained by the input length. A More practical scenario is that the adversarial examples must be clipped or self-spliced and input into the black-box model. Therefore, it is necessary to explore how to improve transferability in different input length settings. In this paper, we take the synthetic speech detection task as an example and consider two representative SOTA models. We observe that the gradients of fragments with the same sample value are similar in different models via analyzing the gradients obtained by feeding samples into the model after cropping or self-splicing. Inspired by the above observation, we propose a new adversarial attack method termed sliding attack. Specifically, we make each sampling point aware of gradients at different locations, which can simulate the situation where adversarial examples are input to black-box models with varying input lengths. Therefore, instead of using the current gradient directly in each iteration of the gradient calculation, we go through the following three steps. First, we extract subsegments of different lengths using sliding windows. We then augment the subsegments with data from the adjacent domains. Finally, we feed the sub-segments into different models to obtain aggregate gradients to update adversarial examples. Empirical results demonstrate that our method could significantly improve the transferability of adversarial examples after clipping or self-splicing. Besides, our method could also enhance the transferability between models based on different features.

cs.SD↗
arXiv2022

Multiwavelength observations of a partial filament eruption on 13 June 2011

In this paper, we report the multiwavelength observations of the partial filament eruption associated with a C1.2 class flare in NOAA active region 11236 on 13 June 2011. The event occurred at the eastern limb in the field of view (FOV) of Atmospheric Imaging Assembly (AIA) on board the Solar Dynamics Observatory (SDO) spacecraft and was close to the disk center in the FOV of Extreme-UltraViolet Imager (EUVI) on board the behind Solar Terrestrial Relations Observatory (STEREO) spacecraft. During eruption, the filament splits into two parts: the major part and runaway part. The major part flows along closed loops and experiences bifurcation at the loop top. Some of the materials move forward and reach the remote footpoint, while others return back to the original footpoint. The runaway part flows along open field lines, which is evidenced by a flare-related type III radio burst. The runaway part also undergoes bifurcation. The upper branch of escapes the corona and evolves into a jet-like narrow coronal mass ejection (CME) at a speed of 324 km s-1, while the lower branch falls back to the solar surface. A schematic cartoon is proposed to explain the event and provides a new mechanism of partial filament eruptions

astro-ph.SR↗
arXiv2022

Gain without Pain: Recycling Reflected Energy from Wireless Powered RIS-aided Communications

In this paper, we investigate and analyze energy recycling for a reconfigurable intelligent surface (RIS)-aided wireless-powered communication network. As opposed to the existing works where the energy harvested by Internet of things (IoT) devices only come from the power station, IoT devices are also allowed to recycle energy from other IoT devices. In particular, we propose group switching- and user switching-based protocols with time-division multiple access to evaluate the impact of energy recycling on system performance. Two different optimization problems are respectively formulated for maximizing the sum throughput by jointly optimizing the energy beamforming vectors, the transmit power, the transmission time, the receive beamforming vectors, the grouping factors, and the phase-shift matrices, where the constraints of the minimum throughput, the harvested energy, the maximum transmit power, the phase shift, the grouping, and the time allocation are taken into account. In light of the intractability of the above problems, we respectively develop two alternating optimization-based iterative algorithms by combining the successive convex approximation method and the penalty-based method to obtain corresponding sub-optimal solutions. Simulation results verify that the energy recycling-based mechanism can assist in enhancing the performance of IoT devices in terms of energy harvesting and information transmission. Besides, we also verify that the group switching-based algorithm can improve more sum throughput of IoT devices, and the user switching-based algorithm can harvest more energy.

cs.IT↗
arXiv2022

Learning State Representations via Retracing in Reinforcement Learning

We propose learning via retracing, a novel self-supervised approach for learning the state representation (and the associated dynamics model) for reinforcement learning tasks. In addition to the predictive (reconstruction) supervision in the forward direction, we propose to include "retraced" transitions for representation / model learning, by enforcing the cycle-consistency constraint between the original and retraced states, hence improve upon the sample efficiency of learning. Moreover, learning via retracing explicitly propagates information about future transitions backward for inferring previous states, thus facilitates stronger representation learning for the downstream reinforcement learning tasks. We introduce Cycle-Consistency World Model (CCWM), a concrete model-based instantiation of learning via retracing. Additionally we propose a novel adaptive "truncation" mechanism for counteracting the negative impacts brought by "irreversible" transitions such that learning via retracing can be maximally effective. Through extensive empirical studies on visual-based continuous control benchmarks, we demonstrate that CCWM achieves state-of-the-art performance in terms of sample efficiency and asymptotic performance, whilst exhibiting behaviours that are indicative of stronger representation learning.

cs.LG↗
arXiv2022

Flare Quasi-Periodic Pulsation Associated with Recurrent Jets

Quasi-periodic pulsations (QPPs), which carry time features and plasma characteristics of flare emissions, are frequently observed in light curves of solar/stellar flares. In this paper, we investigated non-stationary QPPs associated with recurrent jets during an M1.2 flare on 2022 July 14. A quasi-period of about 45$\pm$10 s, determined by the wavelet transform technique, is simultaneously identified at wavelengths of soft/hard X-ray and microwave emissions, which are recorded by the Gravitational wave high-energy Electromagnetic Counterpart All-sky Monitor, Fermi, and the Nobeyama Radio Polarimeters, respectively. A group of recurrent jets with an intermittent cadence of about 45$\pm$10 s are found in Atmospheric Imaging Assembly (AIA) image series at 304 Å, but they are 180-s earlier than the flare QPP. All observational facts suggest that the flare QPP could be excited by recurrent jets, and they should be associated with nonthermal electrons that are periodically accelerated by a repeated energy release process, like repetitive magnetic reconnection. Moreover, the same quasi-period is discovered at double footpoints connected by a hot flare loop in AIA 94 Å, and the phase speed is measured to 1420 km/s. Based on the differential emission measure, the average temperatures, number densities, and magnetic field strengths at the loop top and footpoint are estimated to 7.7/6.7 MK, 7.5/3.6*10^{10} cm ^{-3}, and 143/99 G, respectively. Our measurements indicate that the 45-s QPP is probably modulated by the kink-mode wave of the flare loop.

astro-ph.SR↗
arXiv2022

On the Convergence Theory of Meta Reinforcement Learning with Personalized Policies

Modern meta-reinforcement learning (Meta-RL) methods are mainly developed based on model-agnostic meta-learning, which performs policy gradient steps across tasks to maximize policy performance. However, the gradient conflict problem is still poorly understood in Meta-RL, which may lead to performance degradation when encountering distinct tasks. To tackle this challenge, this paper proposes a novel personalized Meta-RL (pMeta-RL) algorithm, which aggregates task-specific personalized policies to update a meta-policy used for all tasks, while maintaining personalized policies to maximize the average return of each task under the constraint of the meta-policy. We also provide the theoretical analysis under the tabular setting, which demonstrates the convergence of our pMeta-RL algorithm. Moreover, we extend the proposed pMeta-RL algorithm to a deep network version based on soft actor-critic, making it suitable for continuous control tasks. Experiment results show that the proposed algorithms outperform other previous Meta-RL algorithms on Gym and MuJoCo suites.

cs.AI↗
arXiv2022

Multi-Document Scientific Summarization from a Knowledge Graph-Centric View

Multi-Document Scientific Summarization (MDSS) aims to produce coherent and concise summaries for clusters of topic-relevant scientific papers. This task requires precise understanding of paper content and accurate modeling of cross-paper relationships. Knowledge graphs convey compact and interpretable structured information for documents, which makes them ideal for content modeling and relationship modeling. In this paper, we present KGSum, an MDSS model centred on knowledge graphs during both the encoding and decoding process. Specifically, in the encoding process, two graph-based modules are proposed to incorporate knowledge graph information into paper encoding, while in the decoding process, we propose a two-stage decoder by first generating knowledge graph information of summary in the form of descriptive sentences, followed by generating the final summary. Empirical results show that the proposed architecture brings substantial improvements over baselines on the Multi-Xscience dataset.

cs.CL↗
arXiv2022

First detection of transverse vertical oscillation during the expansion of coronal loops

In this Letter, we perform a detailed analysis of the M5.5-class eruptive flare occurring in active region 12929 on 2022 January 20. The eruption of a hot channel generates a fast coronal mass ejection (CME) and a dome-shaped extreme-ultraviolet (EUV) wave at speeds of 740$-$860 km s$^{-1}$. The CME is associated with a type II radio burst, implying that the EUV wave is a fast-mode shock wave. During the impulsive phase, the flare shows quasi-periodic pulsations (QPPs) in EUV, hard X-ray, and radio wavelengths. The periods of QPPs range from 18 s to 113 s, indicating that flare energy is released and nonthermal electrons are accelerated intermittently with multiple time scales. The interaction between the EUV wave and low-lying adjacent coronal loops (ACLs) results in contraction, expansion, and transverse vertical oscillation of ACLs. The speed of contraction in 171, 193, and 211 Å is higher than that in 304 Å. The periods of oscillation are 253 s and 275 s in 304 Å and 171 Å, respectively. A new scenario is proposed to explain the interaction. The equation that interprets the contraction and oscillation of the overlying coronal loops above a flare core can also interpret the expansion and oscillation of ACLs, suggesting that the two phenomena are the same in essence.

astro-ph.SR↗
arXiv2022

Three New Arnoldi-Type Methods for the Quadratic Eigenvalue Problem in Rotor Dynamics

Three new Arnoldi-type methods are presented to accelerate the modal analysis and critical speed analysis of the damped rotor dynamics finite element (FE) model. They are the linearized quadratic eigenvalue problem (QEP) Arnoldi method, the QEP Arnoldi method, and the truncated generalized standard eigenvalue problem (SEP) Arnoldi method. And, they correspond to three reduction subspaces, including the linearized QEP Krylov subspace, the QEP Krylov subspace, and the truncated generalized SEP Krylov subspace, where the first subspace is also used in the existing Arnoldi-type methods. The numerical examples constructed by a turbofan engine low-pressure (LP) rotor demonstrate that our proposed three Arnoldi-type methods are more accurate than the existing Arnoldi-type methods.

math.NA↗
arXiv2022

Coarse-to-Fine Knowledge-Enhanced Multi-Interest Learning Framework for Multi-Behavior Recommendation

Multi-types of behaviors (e.g., clicking, adding to cart, purchasing, etc.) widely exist in most real-world recommendation scenarios, which are beneficial to learn users' multi-faceted preferences. As dependencies are explicitly exhibited by the multiple types of behaviors, effectively modeling complex behavior dependencies is crucial for multi-behavior prediction. The state-of-the-art multi-behavior models learn behavior dependencies indistinguishably with all historical interactions as input. However, different behaviors may reflect different aspects of user preference, which means that some irrelevant interactions may play as noises to the target behavior to be predicted. To address the aforementioned limitations, we introduce multi-interest learning to the multi-behavior recommendation. More specifically, we propose a novel Coarse-to-fine Knowledge-enhanced Multi-interest Learning (CKML) framework to learn shared and behavior-specific interests for different behaviors. CKML introduces two advanced modules, namely Coarse-grained Interest Extracting (CIE) and Fine-grained Behavioral Correlation (FBC), which work jointly to capture fine-grained behavioral dependencies. CIE uses knowledge-aware information to extract initial representations of each interest. FBC incorporates a dynamic routing scheme to further assign each behavior among interests. Additionally, we use the self-attention mechanism to correlate different behavioral information at the interest level. Empirical results on three real-world datasets verify the effectiveness and efficiency of our model in exploiting multi-behavior data. Further experiments demonstrate the effectiveness of each module and the robustness and superiority of the shared and specific modelling paradigm for multi-behavior data.

cs.IR↗
arXiv2022

Neural Architecture Search on Efficient Transformers and Beyond

Recently, numerous efficient Transformers have been proposed to reduce the quadratic computational complexity of standard Transformers caused by the Softmax attention. However, most of them simply swap Softmax with an efficient attention mechanism without considering the customized architectures specially for the efficient attention. In this paper, we argue that the handcrafted vanilla Transformer architectures for Softmax attention may not be suitable for efficient Transformers. To address this issue, we propose a new framework to find optimal architectures for efficient Transformers with the neural architecture search (NAS) technique. The proposed method is validated on popular machine translation and image classification tasks. We observe that the optimal architecture of the efficient Transformer has the reduced computation compared with that of the standard Transformer, but the general accuracy is less comparable. It indicates that the Softmax attention and efficient attention have their own distinctions but neither of them can simultaneously balance the accuracy and efficiency well. This motivates us to mix the two types of attention to reduce the performance imbalance. Besides the search spaces that commonly used in existing NAS Transformer approaches, we propose a new search space that allows the NAS algorithm to automatically search the attention variants along with architectures. Extensive experiments on WMT' 14 En-De and CIFAR-10 demonstrate that our searched architecture maintains comparable accuracy to the standard Transformer with notably improved computational efficiency.

cs.CL↗