Prof to Hand-Calculate Kimi 3 Architecture: Extreme MoE Sparsity
ProfTomYeh · x · 2026-08-02
Prof. Tom Yeh announced a new 'AI by Hand' seminar focusing on frontier LLM architectures. The session will manually calculate and sketch the math behind Kimi 3 (Moonshot AI), including Kimi Delta Attention, attention residuals, and extreme MoE sparsity (activating just 16 of 896 experts in a 2.8T-parameter model). The lecture will also cover architectural optimizations in Google's Gemma 4 and Alibaba's Qwen 3.6.
More from Research
- Simplex Diffusion Models paper talk by EPFL/DeepMind's Justin Deschenaux on Oct 19 — ArnaudDoucet1 · 2026-10-16
- Meituan LongCat team's Data World synthesizes data environments to train data agents — meituan-longcat · 2026-10-16
- FDE-Gym benchmark shows huge gap on real AI delivery tasks: top model scores 0.563 — DeepLethe · 2026-10-16
- ForcingBooster halves computation of autoregressive video diffusion with patch-wise denoising — VAST-AI · 2026-10-16
- Baidu open-sources FlashMask: up to 1.55x faster than FlashAttention-4 for long-context training — BaiduResearch · 2026-10-16
- SemEval-2027 Task 9 opens: multimodal news framing analysis across 4 languages — preslav_nakov · 2026-10-16