Model-Architecture

12 pages

• 177: 详解Kimi K3：强到冲击Anthropic估值的模型什么样？

  source
  2026-08-03
• Attention Residues

  concept Updated 2026-08-08
  wiki
• Kimi Delta Attention / KDA

  concept Updated 2026-08-08
  wiki
• Kolmogorov-Arnold Networks

  concept Updated 2026-08-08
  wiki
• Mixture of Experts

  concept Updated 2026-08-08
  wiki
• NoPE / No Position Encoding

  concept Updated 2026-08-08
  wiki
• On-Device Model Hierarchy

  concept Updated 2026-07-09
  wiki
• Per-Head Muon

  concept Updated 2026-08-08
  wiki
• Quantile Balancing

  concept Updated 2026-08-08
  wiki
• Scaling Efficiency

  concept Updated 2026-08-08
  wiki
• Transformer Architecture

  concept Updated 2026-08-08
  wiki
• Zeng Zhiyuan / 曾志远

  entity Updated 2026-08-08
  wiki