The complete mathematics
curriculum for modern AI
From linear algebra to large language models — rigorous derivations, formal proofs, and deep intuition.
Curriculum
12 volumes covering the full mathematical stack of modern AI.
Mathematics for AI
Linear algebra, calculus, probability, statistics, and optimization — the mathematical language of machine learning.
Machine Learning
Supervised and Bayesian learning, generalization theory, and classical statistical inference.
Deep Learning
Neural networks, training dynamics, attention mechanisms, and generative modeling theory.
Transformers & Attention
The complete theory of attention mechanisms, transformer architectures, optimization techniques, and KV-cache systems.
Diffusion Models
Forward and reverse diffusion processes, score matching, and conditional generation.
Large Language Models
Tokenization, pretraining, alignment, fine-tuning, inference, and retrieval-augmented generation.
Systems & Optimization
Memory-efficient attention, quantization, distributed training, and inference optimization.
Flow Models
Normalizing flows, continuous normalizing flows, flow matching, rectified flows, and optimal transport for generative modeling.
Quantization
Reducing model precision for efficient inference: INT8, INT4, GPTQ, AWQ, SmoothQuant, and quantization-aware training.
Model Distillation
Knowledge transfer from large to small models: teacher-student frameworks, logit distillation, feature matching, and self-distillation.
Model Training Methods
Pre-training, fine-tuning, RLHF, DPO, curriculum learning, continual learning, and multi-task training strategies.
Model Optimization
Pruning, sparsity, neural architecture search, efficient architectures, and compute-optimal model design.
Why this textbook?
Built for engineers and researchers who want real mathematical understanding.
Full Derivations
Every proof step-by-step. No "it can be shown" shortcuts.
Clear Structure
Objectives, intuition, formalism, pitfalls, and exercises.
Theory First
Pure mathematics — understand the why, not just the how.
Research Grade
Original papers cited. Connections to frontier research.
Community
Learn together — ask questions, share insights, and contribute.
Ready to build deep intuition?
Start from linear algebra or jump to any topic you need.