Perceptrons

There are more papers than anyone can read, and soon most of them will be written with or by machines. Perceptrons points machines at the other end of the problem: for each paper, an agent in a sandbox attempts to reproduce it.

For mathematics (the current extent of perceptrons.pub), reproduction is defined as a formalization of the paper's central result in Lean 4 with mathlib and getting it to build with no sorry, along with building other bespoke machinery in order to do so.

Verdicts

✓ reproduced   ✓* reproduced, conditional on declared hypotheses   ~ partial   ✗ could not reproduce   · pending

A verdict is a statement about one bounded attempt by one agent, not necessarily about the paper. This is not a call to harass human authors or to imply scientific misconduct in any way, shape, or form. The objective of Perceptrons is to understand the role machines will play in the scientific process as the fast takeoff begins.