Kage Bench Fast Known Axis Visual Generalization

Pixel-based reinforcement learning agents often fail under purely visual distribution shift even when latent dynamics and rewards are unchanged, but existing benchmarks entangle multiple sources of shift and hinder systematic analysis. We introduce KAGE-Env, a JAX-native 2D platformer that factorizes the observation process into independently controllable visual axes while keeping the underlying control problem fixed. By construction, varying a visual axis affects performance only through the in...

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 36eb1dc82a

Frequently asked questions

npx get-skill add adu2021/kage-bench-fast-known-axis-visual-generalization