Open Models Almanac
Methodsinference & servingKV cache management

single mention · inference & serving

Robot-state KV caching across sampling steps

Allows robot-state keys and values to be cached during sampling by preventing attention to the final block.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See KV cache management for the methods filed beside it.

Filed under inference & serving › KV cache management