How is the system prompt different from the user prompt?

easy

Answer

$Chat-tuned LLMs use a chat template with roles: 'system' (developer instructions, persona, guardrails, style), 'user' (query), 'assistant' (response).$
$RLHF training teaches the model to weight system instructions higher than user instructions in case of conflict — the primary defense against user jailbreaks and prompt injection.$
$Best practice: put durable, safety-critical, and format instructions in the system prompt; put per-request context in the user prompt.$

Check yourself — multiple choice

No difference
System = developer / persona / durable instructions; RLHF weights it higher than user → put safety and format there
Same weight as user prompt
User overrides system by default

System prompt: durable instructions with higher priority than user (post-RLHF).

#prompting

214 interview questions in this topic.