Weird that no one explicitly added embedded ascii/utf-8 directly to LLM training data for compression. Given that high dimensional spaces are built as vector sp

Weird that no one explicitly added embedded ascii/utf-8 directly to LLM training data for compression. Given that high dimensional spaces are built as vector spaces fully describable by basis vectors, I would assume somewhere these characters got added. Perhaps it's an activation issue (i.e. broken after all) and it just needs an occasional change of basis.