Definition
Plain language
A stretch of the character list that has shadow copies of ordinary letters which screens usually draw as nothing.
As stated in the literature
A deprecated Unicode block (U+E0000 range) containing tag characters that mirror ASCII; valid and decodable by tokenizers but unassigned in mainstream fonts, so they render invisibly while still reaching the model.
Also called: tag-block, Unicode tag block
Why it matters: It matters because invisible characters can smuggle instructions into an AI that a human reviewer would never see.
For example, a sentence can carry hidden copies of ordinary letters from this range that your screen draws as nothing while the AI still reads them.
Heard on the show
“And buried up in that unused territory is a block of about a hundred and twenty-eight characters — a deprecated set called the tag block — where every ordinary character has a shadow twin.”Episode 208 — The Blank Space in Your AI Approval Box That Isn't Empty