Ant

The 'Ant' dataset/benchmark is used to evaluate reinforcement learning algorithms in a simulated environment that presents challenges related to high-dimensional policies and non-stationary rewards.

Papers using Ant (3)

ARISE: Adaptive Reinforcement Integrated with Swarm Exploration2026A Comparison Between Decision Transformers and Traditional Offline Reinforcement Learning Algorithms2025An Advantage-based Optimization Method for Reinforcement Learning in Large Action Space2024 · 1 cites