Wouldn’t it be safer to have independent estimates to make sure the model owners did not make « mistakes »?

Mathias is awesome both scientifically and personally. Don’t hesitate for this fantastic opportunity!

You must have a very efficient upstream filter before spending time to study and run code. How do you choose which of those paper deserve such an investment? (Besides your network that you mention in this thread). Do you trust authors, labs, institutions… or even your instinct based on abstracts?

Thanks for sharing! My timeline is now filled with good stuff!