LCLipeng Chen

Lipeng Chen

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Lipeng Chen has a total of 102 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Reinforcement Learning-Assisted Quantum Simulation of Many-Body Excited States and Real-Time Dynamics

    by Jiaji Zhang, Lipeng Chen, Carlos L. Benavides-Riveros

    The computation of electronic excited states and real-time quantum dynamics of many-Fermion systems is among the most promising applications of near-term quantum computing. In this work, we generalize the reinforcement learning contracted quantum eigensolver (RL-CQE), previously developed for ground-state problems, to electronic excited states and real-time quantum dynamics, in which a deep Q-network agent adaptively selects the two-body operators at each iteration, yielding more compact ansätze and improved robustness with respect to critical hyperparameters. A key feature of the algorithm is a scalable state representation based on the ACSE residuals, whose dimension grows with the one-particle basis but remains independent of the number of targeted excited states. We also show the equivalence of sign-free qubit operators in the excited-state and time-evolution settings, extending a result previously established for ground-state problems. Our RL-CQE for time evolution derives from a constant-scaling ansatz that represents the wave function with a fixed number of unitary transformations independent of simulation time, enabled by the shared unitary structure of the purified ensemble treatment of excited states. Benchmarks on chemical systems demonstrate chemical accuracy with minimal operator counts across a range of bond lengths.

  • Ab initio strong-pump strong-probe doorway-window simulation protocol

    by Xiuqing Zhang, Kewei Sun, Luis Vasquez, Lipeng Chen, Sebastian V. Pios, Zhenggang Lan, Maxim F. Gelin

    We have developed an on-the-fly methodology for the simulation of strong-field pump-probe (PP) spectra using semiclassical trajectory methods. This methodology generalizes the weak-field semiclassical doorway-window (DW) simulation protocol to pump and probe pulses of any intensity. The strong-field DW functions are defined by explicit analytical formulas which are fully determined by the molecular transition dipole moments and electronic energies evaluated along the semiclassical trajectories, so that the weak-field and strong-field PP signals can be simulated at roughly the same computational cost. The developed methodology has been applied for the simulation of strong-field PP spectra of pyrazine. The obtained results vividly demonstrate the advantage of using strong pump and (notably) probe pulses for increasing the information content of PP measurements. In particular, the strong probe pulse significantly enhances the weak transitions originating from the lowest excited electronic states of pyrazine, thereby rendering the nonadiabatic wavepacket dynamics in these weakly allowed states visible.

  • Ab Initio Simulation of Femtosecond Time-Resolved Multipulse Spectroscopies Applied to the Heptazine···H 2 O Complex

    by Sebastian V. Pios, Maxim F. Gelin, Wolfgang Domcke, Lipeng Chen

    In multidimensional time-resolved spectroscopic experiments, multiple (more than two) short laser pulses with variable pulse delay times are employed for the time-resolved exploration of the photoinduced dynamics of molecular chromophores. In the present work, the quasi-classical doorway-window (DW) methodology recently developed for transient absorption pump–probe (PP) spectroscopy [ Gelin, M. F. ; J. Chem. Theory Comput. 2021, 17, 2394] has been generalized to multipulse spectroscopies. Pump-push-probe (PPP) spectroscopy (involving three laser pulses) and pump-stimulated two-dimensional (P-2D) spectroscopy (involving five laser pulses) are considered as specific examples. The quasi-classical DW approximation results in conceptually simple and computationally efficient simulation protocols which are suitable for implementation with ab initio on-the-fly electronic-structure calculations. Simulations of PPP and P-2D spectra performed for the hydrogen-bonded heptazine···H2O complex illustrate that pump-stimulated experiments provide rich information on the ultrafast radiationless electronic relaxation dynamics of the excited electronic states of the heptazine···H2O complex which complements the information obtained by conventional PP or 2D experiments.

  • WaveMixings.jl: A Julia Package for Computing Time-Resolved Nonlinear Electronic Spectra from on-the-Fly Quasi-Classical Trajectories

    by Luis Vasquez, Maxim F. Gelin, Sebastian Pios, Lipeng Chen, Zhenggang Lan, Wolfgang Domcke

    We present an efficient numerical implementation of the quasi-classical doorway-window approximation, specifically designed for on-the-fly simulations of time-resolved nonlinear spectroscopic signals, in the Julia package WaveMixings.jl. The code takes as input quasi-classical trajectory data (energies of electronic states and transition dipole moments between elecronic states) as a function of discretized time, which may be provided by any nonadiabatic quasi-classical dynamics package. The output of WaveMixings.jl are raw spectroscopic data, that is, spectral intensities as functions one or more frequencies and pump-probe delay time. The package contains modules that facilitate standard tasks such as input/output data handling, data filtering, and postprocessing, among others. WaveMixings.jl includes implementations of various signals, including integral and dispersed transient absorption pump-probe signals, time- and frequency-resolved fluorescence spectra, and two-dimensional electronic spectra. By developing WaveMixings.jl we aim to create a versatile platform to perform simulations and develop new methodologies within the quasi-classical doorway-window framework. WaveMixings.jl differs from other existing codes for the simulation of nonlinear time-resolved spectra by the explicit inclusion of the shapes and durations of the laser pulses.


  • Neural network solution of non-Markovian quantum state diffusion and operator construction of quantum stochastic process

    by Jiaji Zhang, Carlos L. Benavides-Riveros, Lipeng Chen

    Non-Markovian quantum state diffusion provides a wavefunction-based framework for modeling open quantum systems. In this work, we introduce a novel machine learning approach based on an operator construction algorithm. This algorithm employs a neural network as a universal generator to reconstruct the stochastic time evolution operator from an ensemble of quantum trajectories. Unlike conventional machine learning methods that approximate time-dependent wavefunctions or expectation values, our operator-based approach offers broader applicability to stochastic processes. We benchmark the algorithm on the spin-boson model across diverse spectral densities, demonstrating its accuracy. Furthermore, we showcase the operator’s utility in calculating absorption spectra and reconstructing reduced density matrices at extended timescales. These results establish a new paradigm for the application of machine learning in quantum dynamics.


  • Open Access

    Neural Network Solution of Non-Markovian Quantum State Diffusion and Operator Construction of Quantum Stochastic Process

    by Jiaji Zhang, Carlos L. Benavides-Riveros, Lipeng Chen

    Non-Markovian quantum state diffusion provides a wavefunction-based framework for modeling open quantum systems. In this work, we introduce a novel machine learning approach based on an operator construction algorithm. This algorithm employs a neural network as a universal generator to reconstruct the stochastic time evolution operator from an ensemble of quantum trajectories. Unlike conventional machine learning methods that merely approximate time-dependent wavefunctions or expectation values, our operator-based approach yields broader applications and enhanced interpretability of the stochastic process. We benchmark the algorithm on the spin-boson model across diverse spectral densities, demonstrating its accuracy. Furthermore, we showcase the operator's utility in calculating absorption spectra and reconstructing reduced density matrices at extended timescales. These results establish a new paradigm for the application of machine learning in quantum dynamics.

  • Quasiclassical Doorway–Window Simulation of Femtosecond Transient-Absorption Pump–Probe Signals Beyond the Weak-Pump Limit

    by Hefen Guan, Kewei Sun, Luis Vasquez, Lipeng Chen, Sebastian V. Pios, Zhenggang Lan, Maxim F. Gelin

    We developed an ab initio on-the-fly protocol for the simulation of transient-absorption strong-pump weak-probe signals by combining a trajectory surface hopping method with the classical doorway-window approximation. With this protocol, transient-absorption strong-pump weak-probe signals can be evaluated at about the same computational cost as weak-pulse transient-absorption pump-probe signals. The methodology is illustrated by the simulation of transient-absorption strong-pump weak-probe spectra of pyrazine, the weak-pulse pump-probe spectra of which were computed with the on-the-fly trajectory surface hopping─doorway-window method by Gelin et al., J. Chem. Theory Comput. 2021, 17, 2394. The effect of the intensity of the pump pulse on the transient-absorption strong-pump weak-probe spectra is investigated. In particular, it is demonstrated that the relative intensities of the ground-state bleach, stimulated emission, and excited-state absorption signals can be modified by variation of the strength of the pump pulse.