Fetching the paper…

Source-Aware Training Enables Knowledge Attribution in Language Models · Around