LLM Research & News MiniCPM5-2B Won a 2.8-Point Average. That Is Not a Server
OpenBMB's 2.52B model posted 53.9 vs 51.1 across 34 benches in a Sep 25–27 small-model cluster. Liquid's draft model only speeds decode. An SAE paper flags answers.
Articles tagged with sparse autoencoders
1 article