Kimi K3 Architecture Revealed: KDA and LatentMoE Innovations
Ahmad_Al_Dahle · x · 2026-07-21
While awaiting the official tech report for Kimi K3, Ahmad Al-Dahle highlighted a few key architectural innovations:
- KDA: Swaps Gated DeltaNet's single scalar decay for learned per-dimension forgetting.
- AttnRes: Retrieves selectively across network depth.
- LatentMoE: Activates 16 out of 896 experts, balanced from router score quantiles.
Related event: Moonshot Releases 2.8T Open-Weights Model Kimi K3(14 posts)→
More from Research
- MIT's Ziming Liu: AutoResearch doesn't generalize — blind training tricks cause 'forking' disasters — ZimingLiu11 · 2026-10-26
- Stanford Data-Centric AI Workshop reopens site, extends poster deadline — jure · 2026-10-26
- 'Vibe verification': ~60% of AI-written formal specs prove the wrong property — basedjensen · 2026-10-26
- Ito: a browser that detects browsing intent and auto-organizes tabs into Flows — andrewkuznet · 2026-10-26
- Cornell study says brains process language as linear sequences, challenging Chomsky's hierarchy — anirbanbandyo · 2026-10-26
- Hi3D V1.0 open-sourced with 2M training assets and 90% slicing success rate — kimmonismus · 2026-10-26