2026
Linear Attention Fundamentals: Deployment
Some Thoughts on Model Sharding, KV Cache, and Inference Acceleration: Compute and Data
Linear Attention Fundamentals: Engineering
2025
Linear Attention Fundamentals: Theory
VLMs and the Evolution of Reasoning
More Thoughts on the Co-Evolution of RL Frameworks and Algorithms
Ray and LLM Reinforcement Learning Framework Design
Ray Data Backpressure Mechanisms
FlashMLA Kernel Analysis