Search

We propose a lightweight deep convolutional neural network (lCNN) to estimate cosmological parameters from simulated three-dimensional dark matter (DM) halo distributions and associated statistics. The training dataset comprises 2000 realizations of a cubic box with a side length of 1000 $h^{-1}{\rm Mpc}$, and interpolated over a cubic grid of $300^3$ voxels, with each simulation produced using $512^3$ DM particles and $512^3$ neutrinos. Under the flat $\Lambda$CDM model, simulations vary standard six cosmological parameters including $\Omega_m$, $\Omega_b$, $h$, $n_s$, $\sigma_8$, $w$, along with the neutrino mass sum, $M_\nu$. We find that: 1) within the framework of lCNN, extracting large-scale structure information is more efficient from the halo density field compared to relying on the statistical quantities including the power spectrum, the two-point correlation function, and the coefficients from wavelet scattering transform; 2) combining the halo density field with its Fourier transformed counterpart enhances predictions, while augmenting the training dataset with measured statistics further improves performance; 3) achieving high accuracy in inferring $\Omega_m$, $h$, and $\sigma_8$ by the neural network model, while being inefficient in predicting $\Omega_b$, { $n_s$}, $M_\nu$ and $w$; 4) { compared to the simple fully connected network trained with three statistical quantities, our CNN yields statistically reduced errors, showing improvements of approximately 23\% for $\Omega_m$, 11\% for $h$, 8\% for $n_s$, and 21\% for $\sigma_8$. Additionally, in comparison with the likelihood-based analysis on $P(k)$ data, our CNN provides much tighter constraints on parameters, especially on $\Omega_m$ and $\sigma_8$.} Our study emphasizes this lCNN-based novel approach in extracting large-scale structure information and estimating cosmological parameters., Comment: v2: matches the version published in PRD,17 pages,12 figures

An unresolved challenge facing electronics at a few-nm scale is that resistive channels start leaking due to quantum tunneling. This affects the performance of nanoscale transistors, with single-molecule devices displaying particularly low switching ratios and operating frequencies, combined with large subthreshold swings.1 The usual strategy to mitigate quantum effects has been to increase device complexity, but theory shows that if quantum effects are exploited correctly, they can simultaneously lower energy consumption and boost device performance.2-6 Here, we demonstrate experimentally how the performance of molecular transistors can be improved when the resistive channel contains two destructively-interfering waves. We use a zinc-porphyrin coupled to graphene electrodes in a three-terminal transistor device to demonstrate a >104 conductance-switching ratio, a subthreshold swing at the thermionic limit, a > 7 kHz operating frequency, and stability over >105 cycles. This performance is competitive with the best nanoelectronic transistors. We fully map the antiresonance interference features in conductance, reproduce the behaviour by density functional theory calculations, and trace back this high performance to the coupling between molecular orbitals and graphene edge states. These results demonstrate how the quantum nature of electron transmission at the nanoscale can enhance, rather than degrade, device performance, and highlight directions for future development of miniaturised electronics., Comment: 11 pages, 4 figures

Quantum interference in nano-electronic devices could lead to reduced-energy computing and efficient thermoelectric energy harvesting. When devices are shrunk down to the molecular level it is still unclear to what extent electron transmission is phase coherent, as molecules usually act as scattering centres, without the possibility of showing particle-wave duality. Here we show electron transmission remains phase coherent in molecular porphyrin nanoribbons, synthesized with perfectly defined geometry, connected to graphene electrodes. The device acts as a graphene Fabry-P\'erot interferometer, allowing direct probing of the transport mechanisms throughout several regimes, including the Kondo one. Electrostatic gating allows measurement of the molecular conductance in multiple molecular oxidation states, demonstrating a thousand-fold increase of the current by interference, and unravelling molecular and graphene transport pathways. These results demonstrate a platform for the use of interferometric effects in single-molecule junctions, opening up new avenues for studying quantum coherence in molecular electronic and spintronic devices., Comment: 14 pages, 3 figures

The interplay between nuclear and electronic degrees of freedom strongly influences molecular charge transport. Herein, we report on transport through a porphyrin dimer molecule, weakly coupled to graphene electrodes, that displays sequential tunneling within the Coulomb-blockade regime. The sequential transport is initiated by current-induced phonon absorption and proceeds by rapid sequential transport via a non-equilibrium vibrational distribution. We demonstrate this is possible only when the vibrational dissipation is slow relative to sequential tunneling rates, and obtain a lower bound for the vibrational relaxation time of 8 ns, a value that is dependent on the molecular charge state., Comment: 8 pages, 7 figures

The application of reinforcement learning (RL) in robotic control is still limited in the environments with sparse and delayed rewards. In this paper, we propose a practical self-imitation learning method named Self-Imitation Learning with Constant Reward (SILCR). Instead of requiring hand-defined immediate rewards from environments, our method assigns the immediate rewards at each timestep with constant values according to their final episodic rewards. In this way, even if the dense rewards from environments are unavailable, every action taken by the agents would be guided properly. We demonstrate the effectiveness of our method in some challenging continuous robotics control tasks in MuJoCo simulation and the results show that our method significantly outperforms the alternative methods in tasks with sparse and delayed rewards. Even compared with alternatives with dense rewards available, our method achieves competitive performance. The ablation experiments also show the stability and reproducibility of our method., Comment: 6 pages, 5 figures

Deep reinforcement learning (DRL) has been proven to be a powerful paradigm for learning complex control policy autonomously. Numerous recent applications of DRL in robotic grasping have successfully trained DRL robotic agents end-to-end, mapping visual inputs into control instructions directly, but the amount of training data required may hinder these applications in practice. In this paper, we propose a DRL based robotic visual grasping framework, in which visual perception and control policy are trained separately rather than end-to-end. The visual perception produces physical descriptions of grasped objects and the policy takes use of them to decide optimal actions based on DRL. Benefiting from the explicit representation of objects, the policy is expected to be endowed with more generalization power over new objects and environments. In addition, the policy can be trained in simulation and transferred in real robotic system without any further training. We evaluate our framework in a real world robotic system on a number of robotic grasping tasks, such as semantic grasping, clustered object grasping, moving object grasping. The results show impressive robustness and generalization of our system., Comment: 9 pages, 10 figures, under review of "Robotics and Autonomous Systems"

Vision-based learning methods provide promise for robots to learn complex manipulation tasks. However, how to generalize the learned manipulation skills to real-world interactions remains an open question. In this work, we study robotic manipulation skill learning from a single third-person view demonstration by using activity recognition and object detection in computer vision. To facilitate generalization across objects and environments, we propose to use a prior knowledge base in the form of a text corpus to infer the object to be interacted with in the context of a robot. We evaluate our approach in a real-world robot, using several simple and complex manipulation tasks commonly performed in daily life. The experimental results show that our approach achieves good generalization performance even from small amounts of training data., Comment: submitted to IROS2020