Pith. sign in

CONTRACT

Colophon

Who runs Pith, how to verify what it publishes, and where everything lives. Every link on this page used to sit in the site footer; it moved here so the footer could become a provenance stamp instead of a sitemap.

Verify

Pith never asks to be trusted without a re-runnable artifact. Start here to check one.

  • Signing key the Ed25519 public key Pith signs records with; verify any bundle against it
  • Integrity protocol how records are signed, versioned, and challenged
  • Developers APIs and machine-readable record formats

Tools

Instruments that point the review machinery at your own work.

Record

The judgments themselves, and the rules they are made under.

Site

The entity behind Pith

Pith is a project of the Recognition Physics Research Institute, a not-for-profit entity registered in Texas. The institute funds and operates the site; it does not grade its own homework in it. Reviews are produced by a versioned machine pipeline, signed, and open to challenge by anyone, including challenges to papers the institute's own researchers write.

The institute also maintains Recognition Science, a formal research program. On Pith it is a side corpus: browsable, clearly labeled, and nothing more. It is never a review criterion, never a scoring input, and never an arbiter of any judgment published here. No review outcome on Pith depends on agreement with any physics framework.

Jonathan Washburn (ORCID 0009-0001-8868-7497) discovered Recognition Science. Its Lean 4 corpus is public source, as is the foundational layer beneath it, which derives the number tower from a single primitive and names the exact price of the classical continuum. Both can be cloned and built.

He directed the formalization alongside a team of scientists and mathematicians, LLM-based AI models, and Cambrian, an AI system he built that proposes and proves theorems on its own.

He and Emma Tully co-founded the institute and built Pith.

Some of that work is in the corpus Pith reviews. It runs through the same machine pipeline as everyone else's, carries the same verdicts, and is open to the same challenges. Peer-reviewed: Uniqueness of the Canonical Reciprocal Cost (Mathematics 14, 2026); Recognition Geometry (Axioms 15, 2026); Reciprocal Convex Costs for Ratio Matching: Axiomatic Characterization (Axioms 15, 2026).