Fetching the paper…

Implicit meta-learning may lead language models to trust more reliable sources · Around