Kimi K3 bounds decay at -5 to keep chunkwise KDA inside BF16 range

suchenzang · x · 2026-07-28

The thread argues that the chunk-wise form of KDA in Kimi Linear remains largely the same, but Kimi K3 introduces a new lower-bounded decay trick to make the math friendlier to finite precision.

What changes

Why it matters

The author notes this is not a totally novel observation about forget gates/weight decay, but it is a useful explanation of how the new parameterization changes the computation path.

Related event: Kimi K3 Optimizes KDA: Prevents Overflow and Upgrades Gate(2 posts)→

Original post →

More from Research

Research channel →