Research Notes

Axiommachine treats language models as reasoning engines inside a broader cognitive architecture. Notes here can evolve alongside the implementation while still rendering as first-class routes.

Current Questions

• How should a world model remain queryable beyond the context window?
• Which memories need causal structure instead of semantic similarity alone?
• Where should reflection update beliefs, plans, or both?

Editorial Strategy

English notes are the primary publishing target for now. Chinese notes can be added gradually without changing the route structure or UI translation system.