Stories from the gateway
Fundamentals
Attention Mechanisms Explained Without the Linear Algebra
What attention actually computes, why it made transformers work, and why its cost scaling explains almost every practical limit you hit with long context.
· 9 min read
Fundamentals
Batch Size and Throughput: The Trade Behind Every Token Price
Batching is why per-token prices are low and why latency varies under load. How it works, where it stops helping, and what it means for your requests.
· 9 min read
Fundamentals
Beam Search vs Sampling: Why Chat Models Do Not Search
· 9 min read
Fundamentals
Byte-Pair Encoding Explained: How Tokenizers Are Built
· 9 min readFundamentals
Chain of Thought: Why Thinking Out Loud Actually Helps
· 8 min readFundamentals
Constitutional AI: Alignment From Written Rules, Not Labels
· 9 min readFundamentals
Context Length vs Effective Context: The Number You Can Use
· 9 min read
Fundamentals
Continuous Batching: How Servers Keep GPUs Busy
· 9 min readFundamentals
Data Contamination in Training: Why Scores Overstate Skill
· 9 min readFundamentals
DPO vs RLHF: Two Routes From Preferences to Model Behaviour
· 9 min read
Fundamentals
Expert Routing in MoE: How a Token Picks Its Path
· 9 min read
Fundamentals
Flash Attention Explained: Same Maths, Far Less Memory
· 9 min readFundamentals
GPU Memory for Inference: What Actually Fills the Card
· 9 min readFundamentals
Grouped-Query Attention: Why Long Context Got Affordable
· 9 min readReady to start?
Every model, one key. Start free and upgrade when you are ready.