Unverified50% confidenceFactExact time
GQA is an attention optimization scheme proposed by Google in 2023 that groups query heads so each group shares a set of KV projections, typically reducing KV cache VRAM by 4–8×.
1
Sources
50%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGQA由Google在2023年提出，将多个查询头共享同一组Key-Value头，可降低KV Cache内存消耗50%-75%77% similarVerifiedGQA (Grouped Query Attention) is an efficient variant of standard multi-head attention, proposed by Google in 2023, and has been adopted by mainstream models such as LLaMA 3 and Mistral.76% similarUnverifiedGQA was proposed by Google in 2023 as a compromise between MHA and MQA; the 70B version of LLaMA 2 uses 8-group GQA, reducing the KV cache to 1/4 of that of MHA.70% similarUnverifiedGQA由Google在2023年提出，将多个查询头共享同一组Key-Value头，已被LLaMA 3、Mistral等主流模型采用66% similarUnverified2020年谷歌推出第四代版本GA4，从基于会话的分析模型转向基于事件的数据模型，并引入机器学习预测功能60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/914950API
curl https://kongchang.com/api/v1/knowledge/claims/914950MCP
get_claim(id=914950)