matlok 's Collections
LMM

Papers - Encoders - Compression

Emergence with scale is unlikely Given the recent findings of [55], we anticipate that continuing to scale models beyond 2 billion parameters is unlik