AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Controlling Underestimation Bias in Reinforcement Learning via Quasi-median Operation

Wei Wei, Yujia Zhang, Jiye Liang, Lin Li, Yyuze Li

AAAI 2022 · Reinforcement learning › Policy gradient and actor-critic · approach: Variance-reduced policy gradients · 8 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use quasi-median aggregation of multiple criticsQMD3 · HalfCheetah-v3average return: 9465.3 -> 12221.9Improved
Use quasi-median aggregation of multiple criticsQMD3 · Walker2d-v3average return: 4449.3 -> 4765.3Improved
Use quasi-median aggregation of multiple criticsQMD3 · Humanoid-v3average return: 4719.8 -> 5332.4Improved
Use quasi-median aggregation of multiple criticsQMD3 · IDP-v2average return: 9237.4 -> 9891.3Improved
Add the quasi-median operation (ablation)QMD3 on HalfCheetah · HalfCheetahaverage return: 10576.3 -> 11763.5Improved
Add the quasi-median operation (ablation)QMD3 on Hopper · Hopperaverage return: 3248.9 -> 3663.4Improved
Add the quasi-median operation (ablation)QMD3 on Walker2d · Walker2daverage return: 4103.6 -> 4424.5Improved
Add the quasi-median operation (ablation)QMD3 on Ant · Antaverage return: 4286.3 -> 4786.4Improved
Add Exploration Improvement by updating the actor using multiple critics (ablation)QMD3 on HalfCheetah · HalfCheetahaverage return: 11276.0 -> 11763.5Improved
Add Exploration Improvement by updating the actor using multiple critics (ablation)QMD3 on Hopper · Hopperaverage return: 3472.8 -> 3663.4Improved
Add Exploration Improvement by updating the actor using multiple critics (ablation)QMD3 on Walker2d · Walker2daverage return: 4354.4 -> 4424.5Improved
Add Exploration Improvement by updating the actor using multiple critics (ablation)QMD3 on Ant · Antaverage return: 4534.2 -> 4786.4Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy