Sample Efficiency Comparisons Across Major RL Algorithms on Atari 100K
Atari 100K shows which algorithms learn efficiently from minimal experience.
Ingrid Ford
Section
Features
3 stories in Features.
NetHack Learning Environment as a Frontier Benchmark for General Agents
Why a decades-old roguelike remains unsolved and what that teaches us about AI.
Simone Novak
Zero-Shot Transfer of Game Policies Across Procedurally Varied Levels
Policies trained on narrow conditions collapse on unfamiliar levels without retraining.
Ingrid Yamada