ATTENTION
1 item
1 post
Blog
Kimi Linear: An Attention Architecture That Outperforms Full AttentionMoonshot AI's Kimi Linear paper introduces KDA, a hybrid linear attention that beats full attention at all scales - 75% less KV cache, 6x decoding at 1M context, and open-source checkpoints.
Get Smarter About AI Dev
New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.
One email per weekReal code, not theoryFree forever