Large language models detect methodological errors that human peer review misses in clinical trials
Peer review is the principal quality control of published research, but methodological errors can go undetected. In an earlier experiment with errors inser