arXiv subjects

Yunlong Liu

Publications and source records attributed to Yunlong Liu.

At least 37 records · Page 2Linked to original sources

arXiv2024

Analytic solutions for the motion of spinning particles near brane-world black hole

The general motion of spinning test particles to the leading order approximation of spin in the brane-world spacetime is investigated. Analytical integrations for the equations of motion and linear shifts in orbital frequency are obtained. As a result, we found that both the nodal precession and the periastron precession become larger when the tidal charge $b$ of brane-world spacetime becomes smaller. For the periastron precession, the effect is further amplified as the spin increases. Our work can potentially be applied to the study of gravitational waveforms of Extreme Mass Ratio Inspirals with spin in brane-world spacetime.

gr-qc↗

arXiv2024

Intrinsic Dynamics-Driven Generalizable Scene Representations for Vision-Oriented Decision-Making Applications

How to improve the ability of scene representation is a key issue in vision-oriented decision-making applications, and current approaches usually learn task-relevant state representations within visual reinforcement learning to address this problem. While prior work typically introduces one-step behavioral similarity metrics with elements (e.g., rewards and actions) to extract task-relevant state information from observations, they often ignore the inherent dynamics relationships among the elements that are essential for learning accurate representations, which further impedes the discrimination of short-term similar task/behavior information in long-term dynamics transitions. To alleviate this problem, we propose an intrinsic dynamics-driven representation learning method with sequence models in visual reinforcement learning, namely DSR. Concretely, DSR optimizes the parameterized encoder by the state-transition dynamics of the underlying system, which prompts the latent encoding information to satisfy the state-transition process and then the state space and the noise space can be distinguished. In the implementation and to further improve the representation ability of DSR on encoding similar tasks, sequential elements' frequency domain and multi-step prediction are adopted for sequentially modeling the inherent dynamics. Finally, experimental results show that DSR has achieved significant performance improvements in the visual Distracting DMControl control tasks, especially with an average of 78.9\% over the backbone baseline. Further results indicate that it also achieves the best performances in real-world autonomous driving applications on the CARLA simulator. Moreover, qualitative analysis results validate that our method possesses the superior ability to learn generalizable scene representations on visual tasks. The source code is available at https://github.com/DMU-XMU/DSR.

cs.AI↗

arXiv2024

Gravitational waves for eccentric extreme mass ratio inspirals of self-dual spacetime

In this paper, we calculate the frequencies of geodesic orbits in self-dual spacetime on the equatorial plane and obtain the leading-order effects of loop quantum parameters $P$ on the energy flux and angular momentum flux in eccentric extreme mass ratio inspirals. The gravitational waveform under different eccentricity is carried out by improved "analytic-kludge" method. Through the calculation of waveform mismatches for the LISA detector, the constraints on loop quantum parameters will be improved by 1 to 2 orders of magnitude, compared to the weak field experiments in the solar system, and can reach the level of $10^{-8}$.

gr-qc↗

arXiv2024

Episodic Reinforcement Learning with Expanded State-reward Space

Empowered by deep neural networks, deep reinforcement learning (DRL) has demonstrated tremendous empirical successes in various domains, including games, health care, and autonomous driving. Despite these advancements, DRL is still identified as data-inefficient as effective policies demand vast numbers of environmental samples. Recently, episodic control (EC)-based model-free DRL methods enable sample efficiency by recalling past experiences from episodic memory. However, existing EC-based methods suffer from the limitation of potential misalignment between the state and reward spaces for neglecting the utilization of (past) retrieval states with extensive information, which probably causes inaccurate value estimation and degraded policy performance. To tackle this issue, we introduce an efficient EC-based DRL framework with expanded state-reward space, where the expanded states used as the input and the expanded rewards used in the training both contain historical and current information. To be specific, we reuse the historical states retrieved by EC as part of the input states and integrate the retrieved MC-returns into the immediate reward in each interactive transition. As a result, our method is able to simultaneously achieve the full utilization of retrieval information and the better evaluation of state values by a Temporal Difference (TD) loss. Empirical results on challenging Box2d and Mujoco tasks demonstrate the superiority of our method over a recent sibling method and common baselines. Further, we also verify our method's effectiveness in alleviating Q-value overestimation by additional experiments of Q-value comparison.

cs.LG↗

arXiv2023

Sequential Action-Induced Invariant Representation for Reinforcement Learning

How to accurately learn task-relevant state representations from high-dimensional observations with visual distractions is a realistic and challenging problem in visual reinforcement learning. Recently, unsupervised representation learning methods based on bisimulation metrics, contrast, prediction, and reconstruction have shown the ability for task-relevant information extraction. However, due to the lack of appropriate mechanisms for the extraction of task information in the prediction, contrast, and reconstruction-related approaches and the limitations of bisimulation-related methods in domains with sparse rewards, it is still difficult for these methods to be effectively extended to environments with distractions. To alleviate these problems, in the paper, the action sequences, which contain task-intensive signals, are incorporated into representation learning. Specifically, we propose a Sequential Action--induced invariant Representation (SAR) method, in which the encoder is optimized by an auxiliary learner to only preserve the components that follow the control signals of sequential actions, so the agent can be induced to learn the robust representation against distractions. We conduct extensive experiments on the DeepMind Control suite tasks with distractions while achieving the best performance over strong baselines. We also demonstrate the effectiveness of our method at disregarding task-irrelevant information by deploying SAR to real-world CARLA-based autonomous driving with natural distractions. Finally, we provide the analysis results of generalization drawn from the generalization decay and t-SNE visualization. Code and demo videos are available at https://github.com/DMU-XMU/SAR.git.

cs.LG↗

arXiv2023

Quasinormal modes of the Bardeen black hole with a cloud of strings

We investigate the quasinormal mode and greybody factor of Bardeen black holes with a string clouds by WKB approximation and verify them by Prony algorithm. We found that the imaginary part of the quasinormal modes spectra is always negative and the perturbation does not increase with the time, indicating that the system is stable under scalar field perturbation. Moreover, the string parameter $a$ has a dramatically impact on the frequency and decay rate of the waveforms. In addition, the greybody factor becomes larger when $a$ and $λ$ increase while $q$ and $l$ decreases. The parameter $λ$ and $l$ have a big effect on the tails. Especially, when $l=0$, a de Sitter phase appears at the tail.

gr-qc↗

arXiv2021

Collisional Penrose process with spinning particles in braneworld black hole

The Penrose process of an extremal braneworld black hole is studied. We analyze the Penrose process by two massive spinning particles collide near the horizon. By calculating the maximum energy extraction efficiency of this process, it turns out that the maximal efficiency increases as the tilde charge parameter $d$ of the braneworld blackhole decreases. Interestingly, for the negative value of $d$, the efficiency can be even larger than the Kerr case.

gr-qc↗

arXiv2020

Energy conditions in new model of loop quantum cosmology

Recently, a de-Sitter epoch has been found in the new model of loop quantum cosmology which is governed by the scalar constraint with both of Euclidean and Lorentz terms. The singularity free bounce in the new LQC model and the emergent cosmology constant strongly suggest that the effective stress energy tensor induced by quantum corrections must violate the standard energy conditions. In this paper, we do an explicit calculation to analyze the behaviours of specific representative energy conditions, i.e., average null, strong and dominant energy conditions. It turns out that the averaging null energy condition is violated while the dominant energy condition is violated only at a period around the bounce point. More specifically, the strong energy condition is violated not only at a period around the bounce point, but also the whole period from the bounce point to the classical phase corresponding to the de Sitter period. Our results is expected to shed some lights on construction of the wormhole and time machine which usually need exotic matters violate energy conditions.

gr-qc↗

arXiv2020

Maximal efficiency of the collisional Penrose process with spinning particles in Kerr-Sen black hole

We study the collision of two uncharged spinning particles around an extreme Kerr-Sen black hole and calculate the maximal efficiency of the energy extraction from the Kerr-Sen black hole via super Penrose process. We consider the collision of two massive particles as well as collision of a massless particle with a massive particle. For all the cases, we find that the maximum efficiency decreases as the Kerr-Sen black hole's parameter($b=1-a$) increases.

gr-qc↗

arXiv2019

Collision of spinning particles near BTZ black holes

We study the collision property of spinning particles near a Bañados-Teitelboim-Zanelli (BTZ) black hole. Our results show that although the center-of-mass energy of two ingoing particles diverges if one of the particles possesses a critical angular momentum, however, particle with critical angular momentum can not exist outside of the horizon due to the violation of timelike constraint. Further detailed investigation indicates that only the particle with a subcritical angular momentum is allowed to exist near an extremal rotating BTZ black hole and the corresponding collision center-of-mass energy can be arbitrarily large in a critical angular momentum limit.

gr-qc↗

arXiv2019

Online Learning and Planning in Partially Observable Domains without Prior Knowledge

How an agent can act optimally in stochastic, partially observable domains is a challenge problem, the standard approach to address this issue is to learn the domain model firstly and then based on the learned model to find the (near) optimal policy. However, offline learning the model often needs to store the entire training data and cannot utilize the data generated in the planning phase. Furthermore, current research usually assumes the learned model is accurate or presupposes knowledge of the nature of the unobservable part of the world. In this paper, for systems with discrete settings, with the benefits of Predictive State Representations~(PSRs), a model-based planning approach is proposed where the learning and planning phases can both be executed online and no prior knowledge of the underlying system is required. Experimental results show compared to the state-of-the-art approaches, our algorithm achieved a high level of performance with no prior knowledge provided, along with theoretical advantages of PSRs. Source code is available at https://github.com/DMU-XMU/PSR-MCTS-Online.

cs.AI↗

arXiv2019

Combining Offline Models and Online Monte-Carlo Tree Search for Planning from Scratch

Planning in stochastic and partially observable environments is a central issue in artificial intelligence. One commonly used technique for solving such a problem is by constructing an accurate model firstly. Although some recent approaches have been proposed for learning optimal behaviour under model uncertainty, prior knowledge about the environment is still needed to guarantee the performance of the proposed algorithms. With the benefits of the Predictive State Representations~(PSRs) approach for state representation and model prediction, in this paper, we introduce an approach for planning from scratch, where an offline PSR model is firstly learned and then combined with online Monte-Carlo tree search for planning with model uncertainty. By comparing with the state-of-the-art approach of planning with model uncertainty, we demonstrated the effectiveness of the proposed approaches along with the proof of their convergence. The effectiveness and scalability of our proposed approach are also tested on the RockSample problem, which are infeasible for the state-of-the-art BA-POMDP based approaches.

cs.AI↗

arXiv2018

Kerr-de Sitter and Kerr-anti-de Sitter black holes as accelerators for spinning particles

It is shown that the Kerr black hole could act as accelerators for spinning particles. In principle, it could obtain arbitrarily high energy for extremal Kerr black hole. In this paper, we extend the previous research to the Kerr-(anti-)de Sitter background and find that the cosmological constant plays an important role in the result. For the case of Kerr-anti-de Sitter black holes, like the Kerr background, only the extremal Kerr-anti-de Sitter black holes can have a divergent center-of-mass energy of collision. While, for the case of Kerr-de Sitter black holes, the collision of two spinning particles can take place on the outer horizon of the black holes and the center-of-mass energy of collision can blow up if one of the collision particle takes the critical angular momentum. Hence, non-extremal Kerr-de Sitter black holes could also act as accelerators with arbitrarily high center-of-mass energy for spinning particles.

gr-qc↗

arXiv2018

Conformation Clustering of Long MD Protein Dynamics with an Adversarial Autoencoder

Recent developments in specialized computer hardware have greatly accelerated atomic level Molecular Dynamics (MD) simulations. A single GPU-attached cluster is capable of producing microsecond-length trajectories in reasonable amounts of time. Multiple protein states and a large number of microstates associated with folding and with the function of the protein can be observed as conformations sampled in the trajectories. Clustering those conformations, however, is needed for identifying protein states, evaluating transition rates and understanding protein behavior. In this paper, we propose a novel data-driven generative conformation clustering method based on the adversarial autoencoder (AAE) and provide the associated software implementation Cong. The method was tested using a 208 microseconds MD simulation of the fast-folding peptide Trp-Cage (20 residues) obtained from the D.E. Shaw Research Group. The proposed clustering algorithm identifies many of the salient features of the folding process by grouping a large number of conformations that share common features not easily identifiable in the trajectory.

q-bio.QM↗

arXiv2017

Triple-pulse X-ray spot size measurement of the Dragon-II linear induction accelerator

The Dragon-II linear induction accelerator is able to provide triple-pulse electron beams with an adjustable pulse spacing at the minimum of 300 ns. As a main determinant of the image quality, the X-ray spot size is usually quoted as an evaluation of the resolving power. The pinhole imaging method is applied to measure the two-dimensional spatial distribution of the triple-pulse X-ray source, by which the azimuthal asymmetry of the source can be analyzed. The spot size of each pulse is characterized by both the full-width at half-maximum and the LANL definition. The weighted RMS radius at different sum ratios of photon intensity is also calculated, which directly takes into account the contribution of every single photon within the boundary of the spot.

physics.acc-ph↗

arXiv2016

Selecting Bases in Spectral learning of Predictive State Representations via Model Entropy

Predictive State Representations (PSRs) are powerful techniques for modelling dynamical systems, which represent a state as a vector of predictions about future observable events (tests). In PSRs, one of the fundamental problems is the learning of the PSR model of the underlying system. Recently, spectral methods have been successfully used to address this issue by treating the learning problem as the task of computing an singular value decomposition (SVD) over a submatrix of a special type of matrix called the Hankel matrix. Under the assumptions that the rows and columns of the submatrix of the Hankel Matrix are sufficient~(which usually means a very large number of rows and columns, and almost fails in practice) and the entries of the matrix can be estimated accurately, it has been proven that the spectral approach for learning PSRs is statistically consistent and the learned parameters can converge to the true parameters. However, in practice, due to the limit of the computation ability, only a finite set of rows or columns can be chosen to be used for the spectral learning. While different sets of columns usually lead to variant accuracy of the learned model, in this paper, we propose an approach for selecting the set of columns, namely basis selection, by adopting a concept of model entropy to measure the accuracy of the learned model. Experimental results are shown to demonstrate the effectiveness of the proposed approach.

cs.LG↗

arXiv2015

Magnetic dipolar interaction between correlated triplets created by singlet fission in tetracene crystals

Singlet fission (SF) can potentially break the Shockley-Queisser efficiency limit in single-junction solar cells by splitting one photo-excited singlet exciton (S1) into two triplets (2T1) in organic semiconductors. A dark multi-exciton (ME) state has been proposed as the intermediate connecting S1 to 2T1. However, the exact nature of this ME state, especially how the doubly-excited triplets interact, remains elusive. Here, we report a quantitative study on the magnetic dipolar interaction between SF-induced correlated triplets in tetracene crystals by monitoring quantum beats relevant to the ME sublevels at room temperature. The resonances of ME sublevels approached by tuning an external magnetic field are observed to be avoided, which agrees well with the theoretical predictions considering a magnetic dipolar interaction of ~ 0.008 GHz. Our work paves a way to quantify the magnetic dipolar interaction in organic materials and marks an important step towards understanding the underlying physics of the ME state.

physics.chem-ph↗

arXiv2014

Polarization-dependent exciton dynamics in tetracene single crystals

We conduct polarization-dependent ultrafast spectroscopy to study the dynamics of singlet fission in tetracene single crystals. The spectrotemporal species for singlet and triplet excitons in transient absorption spectra are found to be strongly dependent on probe polarization. By carefully analyzing the polarization dependence, the signals contributed by different transitions related to singlet excitons have been disentangled, which is further applied to construct the correlation between dynamics of singlet and triplet excitons. The anisotropy of exciton dynamics provides an alternative approach to tackle the long-standing challenge in understanding the mechanism of singlet fission in organic semiconductors.

cond-mat.mtrl-sci↗