1. [ongoing] tracking flip counts in 1 bit llms
2. [dec 2025] kv cache recycling to
expand usable context
capacity
in low parameter llms (published in the ijrsi winter journal 2025)