GTA reuses attention scores across grouped heads and stores compressed latent values, roughly matching GQA-level quality in sub-1B models while reducing cache and compute.
Hardware-efficient attention for fast decoding, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
GTA: Grouped-head latenT Attention
GTA reuses attention scores across grouped heads and stores compressed latent values, roughly matching GQA-level quality in sub-1B models while reducing cache and compute.