New benchmark suite AMEND++ predicts clinical trial eligibility criteria amendments

Researchers have introduced AMEND++, a new benchmark suite designed to predict amendments to clinical trial eligibility criteria. This suite includes two datasets: AMEND, which tracks historical edits and amendment labels from public trials, and AMEND_LLM, a curated subset focused on substantive changes. They also developed Change-Aware Masked Language Modeling (CAMLM), a pretraining strategy that uses historical edits to improve amendment prediction, showing consistent gains across various baseline models. AI

IMPACT Could streamline clinical trial processes by predicting costly amendments.

RANK_REASON The cluster contains an academic paper introducing a new benchmark suite and methodology for a specific NLP task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• AMEND
• AMEND_LLM
• arXiv
• Change-Aware Masked Language Modeling
• clinical trials
• eligibility criteria
• Hugging Face
• Trisha Das

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →