or

Read a paper

Enter an arXiv ID above, or upload a PDF (you can also drop one anywhere on this page). You get the paper's assumptions (R1) and benchmarks with every reported result (R2). Edit the prompts in the panel on the right before you run.

Saved reports