Methodenbewijsdossier

Prompt Engineering

Prompt engineering is the practice of crafting structured natural-language instructions — prompts — to elicit targeted outputs from large language models (LLMs). Formalised by Brown et al. (2020) in the context of GPT-3 and extended by Wei et al. (2022) with chain-of-thought prompting, it encompasses four main strategies: zero-shot, few-shot, chain-of-thought, and tree-of-thought. Rather than re-training a model, the analyst shapes the model's behaviour entirely through the design of the input text.

Sources recorded, not reviewed

Bronrecord

Citaten letterlijk overgenomen uit het bronrecord van de methode. Hieruit wordt geen verificatie op claimniveau afgeleid.

Prompt Engineering (Instruction Design for Large Language Models)

Taxonomisch methodendossier · process-pipeline / text-mining

Volledige methode openen

Gecureerde claims

Claims opgeslagen in het bewijsregister, elk met zijn eigen beoordeling.

Nog geen gecureerde claims

Deze weergave verzint geen claimbeoordeling als het register er geen heeft.

Gerelateerde methoden

Gegenereerd uit de methodegraaf en getoond als machinaal voorgestelde relaties — er wordt geen bewijsclaim afgeleid.

Same method familyFew-Shot Text Classificationmachine-suggested · Relational suggestion, not evidence.See alsoGPT Fine-Tuningmachine-suggested · Relational suggestion, not evidence.See alsoLoRA and PEFTmachine-suggested · Relational suggestion, not evidence.Same method familyNatural Language Generationmachine-suggested · Relational suggestion, not evidence.Same method familyRetrieval-Augmented Generationmachine-suggested · Relational suggestion, not evidence.Same method familyText Classificationmachine-suggested · Relational suggestion, not evidence.Same method familyZero-Shot Classificationmachine-suggested · Relational suggestion, not evidence.