D-4RL Locomotion

The 'D4RL Locomotion' dataset is a benchmark that contains trajectories for various locomotion tasks used to evaluate offline reinforcement learning methods.

Papers using D-4RL Locomotion (3)

Behavior Preference Regression for Offline Reinforcement Learning2025Diverse Transformer Decoding for Offline Reinforcement Learning Using Financial Algorithmic Approaches2025Finer Behavioral Foundation Models via Auto-Regressive Features and Advantage Weighting2024