Resonant Charge Transport Assisted by the Molecular Vibration in Single-Molecule Junction from Time-Domain ab initio Nonadiabatic Molecular Dynamics Simulations

Using ab initio nonadiabatic molecular dynamics simulation, we study the time-dependent charge transport dynamics in a single-molecule junction formed by gold(Au) electrodes and a single benzene-1,4-dithiol(BDT)molecule. Two different types of charge transport channels are found in the simulation. One is the routine nonresonant charge transfer path, which occurs in several picoseconds. The other is activated when the electronic state of the electrodes and that of the molecule get close in energy, which is referred to as the resonant charge transport. More strikingly, the resonant charge transfer occurs in an ultrafast manner within 100 fs, which notably increases the conductance of the device. Further analysis shows that the resonant charge transport is directly assisted by the B_(2) and A1 molecular vibration modes. Our study provides atomic insights into the time-dependent charge transport dynamics in single-molecule junctions, which is important for designing highly efficient single-molecule devices.

Curricular Robust Reinforcement Learning via GAN-Based Perturbation Through Continuously Scheduled Task Sequence

Reinforcement learning(RL),one of three branches of machine learning,aims for autonomous learning and is now greatly driving the artificial intelligence development,especially in autonomous distributed systems,such as cooperative Boston Dynamics robots.However,robust RL has been a challenging problem of reliable aspects due to the gap between laboratory simulation and real world.Existing efforts have been made to approach this problem,such as performing random environmental perturbations in the learning process.However,one cannot guarantee to train with a positive perturbation as bad ones might bring failures to RL.In this work,we treat robust RL as a multi-task RL problem,and propose a curricular robust RL approach.We first present a generative adversarial network(GAN)based task generation model to iteratively output new tasks at the appropriate level of difficulty for the current policy.Furthermore,with these progressive tasks,we can realize curricular learning and finally obtain a robust policy.Extensive experiments in multiple environments demonstrate that our method improves the training stability and is robust to differences in training/test conditions.