mqa

Tag

Cards List
#mqa

@che_shr_cat: 1/ We have spent years optimizing KV cache via head-sharing (GQA/MQA), but we ignored a fundamental assumption: why do …

X AI KOLs Timeline · 2026-06-09 Cached

This thread challenges the fundamental assumption that Transformers require separate Q, K, and V projections, proposing that merging them can yield massive memory savings for KV cache.

0 favorites 0 likes
← Back to home

Submit Feedback