Policy Optimization with Demonstrations

Authors: Bingyi Kang, Zequn Jie, Jiashi Feng

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We show that POf D induces implicit dynamic reward shaping and brings provable benefits for policy improvement. Furthermore, it can be combined with policy gradient methods to produce state-of-the-art results, as demonstrated experimentally on a range of popular benchmark sparse-reward tasks, even when the demonstrations are few and imperfect.
Researcher Affiliation Collaboration 1Department of Electrical and Computer Engineering, National University of Singapore, Singapore 2Tencent AI Lab, China.
Pseudocode Yes Algorithm 1 Policy optimization with demonstrations
Open Source Code No The paper does not provide an explicit statement or link for open-source code for its methodology.
Open Datasets Yes To comprehensively assess our method, we conduct extensive experiments on eight widely used physical control tasks, ranging from low-dimensional ones such as cartpole (Barto et al., 1983) and mountain car (Moore, 1990) to high-dimensional and naturally sparse environments based on Open AI Gym (Brockman et al., 2016) and Mujoco (Todorov et al., 2012).
Dataset Splits No The paper mentions using
Hardware Specification No The paper does not provide specific details about the hardware used to run the experiments.
Software Dependencies No Implementation Details Due to space limit, we defer implementation details to the supplementary material.
Experiment Setup No Implementation Details Due to space limit, we defer implementation details to the supplementary material.