Corrections

How we get things right

Every summary in Fig. 1 is drafted by an AI model and reviewed by a human before anything is sent. As part of that review, each paper’s key result, the headline number and the baseline it is compared against, is checked against the paper itself before the issue goes out.

When we get things wrong

We still will, sometimes. If you spot an error in a summary, a figure, or a concept entry, email hello@fig1.dev with the paper and what is wrong. We correct the web version as soon as the error is confirmed, and material errors, anything that changes what a reader would conclude, are also flagged in the next issue’s email.

Authors of covered papers: if we have mischaracterized your work, the same address reaches a human who has read your paper.