Share your thoughts, 1 month free Claude Pro on usSee more

Reinforcement Learning on Craftax

48.8Mean Return

PPO-Hybrid

Updated 1mo ago

Evaluation Results

Method	Links
PPO-Hybrid 2026.03		48.8
PPO-Hybrid 2026.03		45.6
PPO-Hybrid 2026.03		45
PPO-Hybrid 2026.03		43.9
PPO-Hybrid 2026.03		43.9
PPO-Hybrid 2026.03		43.2
PPO-RNN 2026.03		39.2
PPO-RNN 2026.03		38.3
PPO-RNN 2026.03		38.1
PPO-RNN 2026.03		36.7
PPO-RNN 2026.03		36.7
PPO-RNN 2026.03		36.4
PPO-MLP 2026.03		30.4
PPO-MLP 2026.03		28.7
PPO-MLP 2026.03		28.6
PPO-RNN 2026.03		27.5
PPO-MLP 2026.03		26.9
PPO-MLP 2026.03		26.9
PPO-MLP 2026.03		24.6
PPO-Hybrid 2026.03		21.2
PPO-MLP 2026.03		13.7
PPO-MLP 2026.03		-0.9
PPO-RNN 2026.03		-0.9
PPO-Hybrid 2026.03		-0.9