Jul 04, 2026 kNN episodic novelty: the count that can't saturate Jul 03, 2026 ICM: giving Mario curiosity (and the bug that ate it) Jul 01, 2026 PPO: on-policy policy gradients that climb past DQN's ceiling Jun 29, 2026 DQN: value learning that beats the floor (and then hits a ceiling) Jun 28, 2026 rl-factory: an RL codebase where a new algorithm is a plug-in, not a rewrite