The Library
Study Deeply, Not Widely
A deliberately small canon read in full — with annotations, citations and a written analysis for every paper.
Mathematics
- Mathematics for Computer ScienceLehman, Leighton and Meyer
- Introduction to Linear AlgebraGilbert Strang
- Linear Algebra Done RightSheldon Axler
- Introduction to ProbabilityBlitzstein and Hwang
- Understanding AnalysisStephen Abbott
- Convex OptimizationBoyd and Vandenberghe
Computer Science
- Introduction to AlgorithmsCormen, Leiserson, Rivest and Stein
- Computer Systems: A Programmer’s PerspectiveBryant and O’Hallaron
- Introduction to the Theory of ComputationMichael Sipser
- Designing Data-Intensive ApplicationsMartin Kleppmann
Artificial Intelligence
- Understanding Machine LearningShalev-Shwartz and Ben-David
- Probabilistic Machine LearningKevin Murphy
- Deep LearningGoodfellow, Bengio and Courville
- Reinforcement LearningSutton and Barto
- Elements of Information TheoryCover and Thomas
Philosophy
- MeditationsMarcus Aurelius
- Nicomachean EthicsAristotle
- Causal Inference: What IfHernán and Robins
The Paper Program
The reading of papers begins with the works that shaped modern language models, then becomes a rolling syllabus — refreshed every six months rather than frozen in one year.
- Attention Is All You Need
- Scaling Laws for Neural Language Models
- Training Compute-Optimal Large Language Models
- Training Language Models to Follow Instructions with Human Feedback
- Direct Preference Optimization
- Switch Transformers
- LoRA
Paper-analysis template
- Claim
- Assumptions
- Method
- Strongest evidence
- Weakest evidence
- Missing baseline
- Confounder
- Follow-up experiment
Written for every paper read. Kept in the scholar’s library record inside Frontier Campus.