Recent Developments in LLM Architectures: KV Sharing, MHC, Compressed Attention

From Gemma 4 to DeepSeek V4, How New Open-Weight LLMs Are Reducing Long-Context Costs

Discussion

0 comments

Replying to cancel

GIF Post

Powered by

More from Substack

No, an AI cannot know the future and never will..

Substack·42

Rethinking Databases for Humans and AI Agents

Substack·64

Semantic transactions: securing untrusted AI agent workflows at the OS boundary

Substack·19

AI's Biggest Unlock Isn't Productivity. It's Access to Expertise

Substack·23

Understanding AI Memory the Basics

Substack·54

How to Spot AI Maps

Substack·44