{ "mean_reward": -125.19454193115234, "eval_mean_best": -125.58744049072266, "is_deterministic": true, "D": 32, "seed": 2024, "algo": "HD-PPO", "env": "Pendulum-v1", "beta_base": 2.5, "beta_bands": null, "std_reward": 88.07369788927744, "ci95_reward": 38.6, "n_eval_episodes": 20 }