D-4RL MuJoCo benchmark

The D4RL MuJoCo benchmark is a dataset used to evaluate offline reinforcement learning algorithms, containing a variety of tasks designed to assess performance in simulated robotic environments.

Papers using D-4RL MuJoCo benchmark (4)

From Robotics to Sepsis Treatment: Offline RL via Geometric Pessimism2026Policy-Driven World Model Adaptation for Robust Offline Model-based Reinforcement Learning2025SALE-Based Offline Reinforcement Learning with Ensemble Q-Networks2025SelfBC: Self Behavior Cloning for Offline Reinforcement Learning2024