HumanoidBench

HumanoidBench is a benchmark dataset used to evaluate the performance of reinforcement learning algorithms in high-dimensional humanoid control tasks.

Papers using HumanoidBench (6)

Fastdsac: Unlocking The Potential Of Maximum Entropy RL In High-dimensional Humanoid Control2026Scaling Tasks, Not Samples: Mastering Humanoid Control through Multi-Task Model-Based Reinforcement Learning2026MetaWorld: Skill Transfer and Composition in a Hierarchical World Model for Grounding High-Level Instructions2026DoublyAware: Dual Planning and Policy Awareness for Temporal Difference Learning in Humanoid Locomotion2025Non-conflicting Energy Minimization In Reinforcement Learning Based Robot Control2025Fasttd3: Simple, Fast, And Capable Reinforcement Learning For Humanoid Control2025