AI Paper Atlas

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

Wasserstein Gradient Flows for Optimizing Gaussian Mixture Policies

Hanna Ziesche, Leonel Rozo

NeurIPS 2023 · Reinforcement learning › Policy gradient and actor-critic · approach: Trajectory reuse in policy gradients · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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 WGF with explicit Bures-Wasserstein updates instead of PPOGMM policy optimization · Collision-avoidance taskenvironment steps to success rate 1: 6 times more environment stepsImproved
Use WGF instead of SAC (no trust region)GMM policy optimization · Collision-avoidance tasksuccess rate: 0.8 -> 1Improved
Use WGF instead of SACGMM policy optimization · Reaching tasksuccess rateImproved
Use WGF instead of PPOGMM policy optimization · Reaching taskenvironment steps to success: 280000 -> 80000Improved
Use WGF instead of PPOGMM policy optimization · Multiple-goal taskenvironment steps at convergence: 130000 -> 95000Improved
Replace Bures-Wasserstein Riemannian updates with implicit Euclidean scheme (non-BW) (ablation)WGF · Reaching and collision-avoidance taskssuccess rateHurt
Replace Bures-Wasserstein Riemannian updates with implicit Euclidean scheme (non-BW) (ablation)WGF · Three planar tasksvariance of success rateHurt
Constrain policy updates via trust region (WGF, PPO) versus unconstrained SACpolicy optimization · Reaching tasksuccess rateImproved

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