A frozen ClinicalT5 embedding with a sliding window and a Transformer-plus-MLP head predicts which trial arm will have more serious adverse events with 77.6% AUC from pre-trial registration text alone.
basic results
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A novel language model for predicting serious adverse event results in clinical trials from their prospective registrations
A frozen ClinicalT5 embedding with a sliding window and a Transformer-plus-MLP head predicts which trial arm will have more serious adverse events with 77.6% AUC from pre-trial registration text alone.