Log in to access more pages.
Create an account or log in to continue reading more pages.
Log in
Table of contents
Modern Neural Networks
From first principles to large language models and frontier research
Read each section in order. Every title can be opened as a TheoryTrace document.
- Cover1
- Copyright2
- How to read this book3
- Introduction4
- Chapter 1: The Computational Neuron5
- Chapter 2: Multilayer Networks and Representation Learning6
- Chapter 3: Optimization by Gradient Descent7
- Chapter 4: Generalization, Regularization, and Model Capacity8
- Chapter 5: Initialization, Normalization, and Training Dynamics9
- Chapter 6: Convolutional Neural Networks10
- Chapter 7: Sequence Models Before Transformers11
- Chapter 8: Attention and the Transformer12
- Chapter 9: Probabilistic Modeling and Neural Generative Models13
- Chapter 10: Self-Supervised Learning and Foundation Models14
- Chapter 11: Large Language Models15
- Chapter 12: Scaling Laws and Emergent Capabilities16
- Chapter 13: Instruction Tuning, Alignment, and Preference Optimization17
- Chapter 14: Retrieval, Tools, and Agentic Systems18
- Chapter 15: Efficient Training and Inference19
- Chapter 16: Multimodal Neural Networks20
- Chapter 17: Robustness, Interpretability, and Mechanistic Understanding21
- Chapter 18: Evaluation, Benchmarks, and Scientific Validity22
- Chapter 19: Safety, Ethics, and Societal Impact23
- Chapter 20: Frontier Research and Open Problems24
- Conclusion25