Walker-2D-v-5

'Walker2D-v5' is a benchmark environment in the MuJoCo simulation suite used to evaluate locomotion control tasks in deep reinforcement learning by analyzing the motion phase structures of trained policies.

Papers using Walker-2D-v-5 (3)

Visualizing Latent Phase Structures in Locomotion Policies: A Multi-Environment Study with Temporal Feature Extension2026Monte Carlo Beam Search For Actor-critic Reinforcement Learning In Continuous Control2025Monte Carlo Beam Search for Actor-Critic Reinforcement Learning in Continuous Control2025