Introducing Lumma-Fev-v2
• 15
AI & ML interests
Sovereign Superintelligence
Papers
Grouped Value Attention: Efficient KV Caching via On-Demand Key Reconstruction
Grouped Query Experts: Mixture-of-Experts on GQA Self-Attention
Articles
Submitted by 
Vishesh Tripathi
Submitted by Vishesh Tripathi