Framework that uses LLMs to generate study triplets and triplet loss to learn embeddings for clustering observational studies prior to meta-analysis, applied to 58 studies on preterm birth cognitive outcomes.
Why sources of heterogeneity in meta-analysis should be investigated
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning study similarity to investigate heterogeneity in meta-analysis using LLMs and triplet loss
Framework that uses LLMs to generate study triplets and triplet loss to learn embeddings for clustering observational studies prior to meta-analysis, applied to 58 studies on preterm birth cognitive outcomes.