Jinwei Xing

I'm last-year PhD student at University of California, Irvine, working on Brain-Inspired Artificial Intelligence with Prof. Jeff Krichmar and Prof. Emre Neftci. Before that, I received the B.S. degree in Computer Science from Sichuan University.

I was fortunate to work at Google Cloud AI as a software engineer intern and research scientist intern at Amazon. I'm interested in applying machine learning in real-world scenarios and have experience in reinforcement learning, natural language processing, computer vision, recommender systems, mobile robotics and neuromorphic computing.

Email / Google Scholar / Github / LinkedIn

Selected Publications

	Policy Distillation with Selective Input Gradient Regularization for Efficient Interpretability Jinwei Xing, Takashi Nagata, Xinyun Zou, Emre Neftci, Jeffrey L. Krichmar Neural Networks, 2023 Conducting input gradient regularization along with policy distillation allows us to generate highly interpretable gradient-based saliency maps and produce more robust policies.
	Linking global top-down views to first-person views in the brain Jinwei Xing, Elizabeth R. Chrastil, Douglas A. Nitz, Jeffrey L. Krichmar PNAS, 2022 To investigate the underlying computations needed to implement the important cognitive function of linking between a first-person experience and a global map in our brain, we used variational autoencoders to reconstruct the top-down images from a robot’s camera view, and vice versa, and conducted analysis.
	Domain Adaptation In Reinforcement Learning Via Latent Unified State Representation Jinwei Xing, Takashi Nagata, Kexin Chen, Xinyun Zou, Emre Neftci, Jeffrey L. Krichmar AAAI, 2021 [code] By separating domain-general and domain-specifc information and conduct RL training based on domain-general embedding, RL policies could achieve almost zero performance loss when depolyed in new domains.
	Neurorobots as a Means Toward Neuroethology and Explainable AI Stewart, Kenneth, Kexin Chen, Tiffany Hwu, Hirak J. Kashyap, Jeffrey L. Krichmar, Jinwei Xing, and Xinyun Zou Frontiers in Neurorobotics, 2020 Using neurorobots as a form of computational neuroethology can be a powerful methodology for understanding neuroscience, as well as for artificial intelligence and machine learning.
	Neuromodulated Patience for Robot and Self-Driving Vehicle Navigation Jinwei Xing, Xinyun Zou, Jeffrey L Krichmar IJCNN, 2020 Using semantic segmentation and deep Q learning to train a mobile robot that can drive in wilder environments.
	CARLsim 4: an open source library for large scale, biologically detailed spiking neural network simulation using heterogeneous clusters Ting-Shuo Chou, Hirak J Kashyap, Jinwei Xing , Stanislav Listopad, Emily L Rounds, Michael Beyeler, Nikil Dutt, Jeffrey L Krichmar IJCNN, 2018 CARLsim is an efficient, easy-to-use, GPU-accelerated library for simulating large-scale spiking neural network (SNN) models with a high degree of biological detail.

Other Projects

Reinforcement Learning Algorithms for Visual Continuous Control

[code]

This project provides a modularized and unified implementation of a series of visual/image-based reinforcement learning algorithms (e.g. SAC, SAC+AE, CURL, RAD, DrQ and ATC) for continuous control tasks.

RLCodebase: PyTorch Codebase For Deep Reinforcement Learning Algorithms

[code]

RLCodebase is a modularized codebase for deep reinforcement learning algorithms based on PyTorch. This project aims to provide an user-friendly reinforcement learning codebase for beginners to get started and for researchers to try their ideas quickly and efficiently.