Does role-play prompting ('You are a senior lawyer...') actually help?
easyAnswer
- Modestly, on some tasks.
- Instruction-tuned LLMs adjust style, tone, and level of detail based on the assumed role.
- Bigger gains: (1) 'You are a strict grader — output PASS or FAIL' improves format adherence.
- (2) 'You are an expert in X' can nudge more domain vocabulary.
- Diminishing returns for frontier models — they already infer expected register from context.
- Effective for style / tone; smaller effect for capability.
Check yourself — multiple choice
- Massive capability gains
- Small style/tone effect + format adherence; frontier LLMs infer register anyway → minor capability boost
- Hurts performance
- No effect at all
Role prompting: modest style / format effect; small capability boost on frontier models.
#prompting
Practise LLMs & GenAI
214 interview questions in this topic.