Prompt engineering
Listed inPrompt EngineeringPrompt Engineeringon
A dense, well-cited tour of zero-shot, few-shot, chain-of-thought, and automatic prompt search.
External
Lil'Log · lilianweng.github.io#prompting
#survey
Getting reliable behaviour out of a model with nothing but the message you send it.
Sign in to track your progress across sections.
12 articles
Listed inPrompt EngineeringPrompt Engineeringon
A dense, well-cited tour of zero-shot, few-shot, chain-of-thought, and automatic prompt search.
Listed inStructured OutputPrompt Engineeringon
Asking for JSON in the prompt, JSON mode, and schema-constrained output give different guarantees. JSON mode promises syntactically valid JSON and not conformance to your schema, which is the gap most applications trip on.
Listed inProviding ContextPrompt Engineeringon
Long documents go above the question, not below it, and the boundary between your instruction and their text is a security control.
Listed inSystem PromptingPrompt Engineeringon
The Messages API takes the system prompt as its own parameter rather than a first message. Putting the date or a user's name in it stops the prefix caching, so you pay the write premium and never collect the read discount.
Listed inChain of ThoughtPrompt Engineeringon
On models that reason by default, asking for step-by-step thinking pays for the steps twice. Carried-over verification instructions cause over-verification, and tag conventions are a weak way to separate reasoning from the answer.
Listed inPrompt EngineeringPrompt Engineeringon
The discipline of specifying a task precisely enough that a model can't drift.
Listed inZero-Shot PromptingPrompt Engineeringon
Asking cold, with no examples, and when that is genuinely enough.
Listed inConstraintsPrompt Engineeringon
Bounding length, format, tone, and scope so output stays inside what you can consume.
Listed inRole & BehaviorPrompt Engineeringon
Assigning a persona and behavioural rules, and how much they actually buy you.
Listed inReAct PromptingPrompt Engineeringon
Interleaving reasoning and tool calls — the loop most agent frameworks implement.
Listed inInput FormatPrompt Engineeringon
Structuring what you send — delimiters, sections, and formats models parse reliably.
Listed inFew-Shot PromptingPrompt Engineeringon
Teaching format and edge cases with a handful of examples in the prompt.