Acrobot

The 'Acrobot' is a benchmark environment used in reinforcement learning that involves a two-link robotic arm designed to swing up and balance, and it is used to evaluate the performance of algorithms under conditions that may violate the Markov property.

Papers using Acrobot (6)

Prediction-Based Markov Violation Scores for Detecting Non-Markovian Observations in Reinforcement Learning2026Sampling Complexity of TD and PPO in RKHS2025Learning to Control Dynamical Agents via Spiking Neural Networks and Metropolis-Hastings Sampling2025Bellman operator convergence enhancements in reinforcement learning algorithms2025Quantifying First-Order Markov Violations in Noisy Reinforcement Learning: A Causal Discovery Approach2025Linear Function Approximation as a Computationally Efficient Method to Solve Classical Reinforcement Learning Challenges2024