A PRISMA review of six automatic prompt optimization methods suggests they can refine prompts without real data, but an integrated framework is still needed for synthetic data generation.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.HC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Automatic Prompt Optimization Techniques: Exploring the Potential for Synthetic Data Generation
A PRISMA review of six automatic prompt optimization methods suggests they can refine prompts without real data, but an integrated framework is still needed for synthetic data generation.