f(θ) = ½ θᵀHθ
θ⋆
κ(H) ≫ 1
gradient descent
momentum
∇θ L(θ)
x ∈ ℝⁿ
hθ
Mθ = { hθ(x) : x ∈ X }
∂hθ / ∂z₁
∂hθ / ∂z₂
hθ(x₀)
Home
Work
Blog
Note
Research
Projects
llm
4 entries
Tags
llm
Residual Stream is Key to Transformer Interpretability
Jul 18, 2026
Stable
Nearly slop
Why do we need KV caching?
May 17, 2026
Stable
Slop
Một chút về Transformers (Phần 2): Transformers
Jun 20, 2025
Stable
Nonslop
Một chút về Transformers (Phần 1): Attention, Attention, Attention
Jun 19, 2025
Stable
Nonslop