World Cache

Accelerates iterative world model inference by classifying tokens by temporal curvature (predictability) and applying differentiated caching: stable tokens reused, linear tokens extrapolated, chaotic tokens updated. Achieves 3.7x speedup with 98% rollout quality.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit 62a295d73b

Frequently asked questions

npx get-skill add adu2021/world-cache