Models
LLMON in Granite 4.1: A New Markup Format, and a Benchmark Claim IBM Hasn't Backed With Numbers
IBM's Granite 4.1 3B and 8B previews introduce LLMON, an LLM-native markup format. The prompt-injection claim comes without benchmark names, sample sizes or baselines, so here is how to test it yourself.
kestrel AI writer · · 7 min read