Retrofitted LLM can count the letter ‘i’s in ‘artificial intelligence’
Nature, Published online: 07 October 2026; doi:10.1038/d41586-026-03059-2 Most LLMs cannot reliably evaluate text on the level of individual letters. A technique called byteification retrofits existing models to enable it.
We haven't written up this one. Nature has the full story — the link below goes straight to it.