Meta Expert shares MoE experts across layers to cut redundancy
With activation and granularity fixed, it matches vanilla MoE using 70–80% of the experts.
1 stories, newest first · one card per story
With activation and granularity fixed, it matches vanilla MoE using 70–80% of the experts.
Today 1.6K readers · 2.5K views · 778 reads · all time 189K views