Categories
2026
A Better Benchmark Is Not a Better Scientist
Improve LLM Calibration by Moving the Answer to the User Message
Semantic Entropy for LLM Confidence
Model Calibration in Large Language Models
Ray mental model: tasks, actors, objects, scheduling, placement
Phase Separation Patterns in Modern AI Workloads
Tensor Parallelism II: GQA Attention, One Head Group at a Time
Tensor Parallelism I: Rebuilding a Megatron SwiGLU Block Across Two Ranks
