Knowledge note
Algorithms
Navigation hub for algorithm foundations, attention, distributed inference, length generalization, and model-wide methods.
Algorithms
- Foundations — Transformer, normalization, probability, sequence, and linear-algebra foundations.
- Attention Algorithms — FlashAttention, attention variants, linear attention, and phase-sensitive KV compression.
- Distributed Inference — Algorithms that partition inference work and state across devices.
- Length Generalization — Methods that extrapolate computation beyond training lengths.
- Model Architecture Algorithms — Model reports spanning several algorithmic mechanisms.