Byte Language Models: Scaling, Emergent Abstractions, and Information Allocation (via Yogthos) — discussion
#ai
