RAG that beats fine-tuning, and the cases where it doesn't
- John Doe
- Architecture , RAG
- 15 May, 2026
RAG won the early-deployment war for good reasons: ...
Self-consistency sampling: cheap reliability when you need the right answer
- William Jacob
- Reliability , Sampling
- 14 May, 2026
Self-consistency sampling sounds like the kind of ...
Function calling: when the schema is more important than the prompt
- Sam Wilson
- Tools , Function Calling
- 13 May, 2026
Function calling looks like a free win the first t ...
Context window management when 128k still isn't enough
- John Doe
- Architecture , Context
- 12 May, 2026
Larger context windows were supposed to make conte ...
Defending against prompt injection without breaking your prompt
- William Jacob
- Security , Prompt Injection
- 11 May, 2026
Prompt injection is the SQL injection of LLM appli ...
System prompts that survive long sessions
- Sam Wilson
- Prompting , System Prompts
- 10 May, 2026
Every team writes a careful system prompt and forg ...