Interventions · Entity Annotation Prototype (beta)

1articles mentioning interventions

1intervention mentions

1distinct interventions extracted

0%of mentions human-reviewed

Active filtersEntity: constrained output ×

constrained output

This intervention is selected. Its supporting articles are shown below.

constrained output1

Articles mentioning constrained output

1 articles · page 1 of 1

Impact of prompt engineering on large language models for risk of bias assessment: a comparative study

ebmed · abstract · 2026-05-17 · 1 matching annotations

83.9%

Prompt engineering DeepSeek-V3 formula-constrained output constrained output direct command GPT-5

…pSeek-V3: 0.0765 vs 0.1239; GPT-5: 0.0548 vs 0.1068).Conclusion Prompt engineering substantially improved the performance of LLMs in risk of bias …

Browse interventions

constrained output

Articles mentioning constrained output

Impact of prompt engineering on large language models for risk of bias assessment: a comparative study