The Collatz Engine

Methodology

How Verification Works

The Collatz Engine is designed as a public, inspectable computational exploration system. It evaluates integers in order, records trajectory statistics, and separates live bounded health checks from full catalog verification.

Back to dashboardAPI docsExport sampleGet involved
01

What the engine does

  • The engine starts at 1 and advances sequentially through positive integers.
  • For each integer, it applies the Collatz rule: even values are divided by 2, and odd values are transformed by 3n + 1.
  • The system records trajectory statistics such as steps to 1, highest peak, and peak ratio.
  • The public dashboard is updated from verified catalog results produced by the running engine.
02

Why batching is used

  • The engine evaluates integers sequentially, then writes completed results in batches for efficiency.
  • Batching reduces write overhead and keeps the public catalog practical to update over time.
  • Batching does not mean numbers are skipped. The verified range advances one integer at a time.
03

What is verified

  • Duplicate catalog entries are checked.
  • Missing ranges inside the verified catalog boundary are checked.
  • Engine state and catalog boundaries are compared for alignment.
  • Longest trajectory and highest peak records are checked for consistency.
  • Worker heartbeat and status readability are checked for public health reporting.
04

Full verification vs live bounded checks

  • The full verification command performs a complete read-only catalog integrity check.
  • The public integrity API performs a bounded live check so the dashboard can report health without expensive full scans on every request.
  • The bounded check is useful for live dashboard health, but it is not a substitute for the full verification scan.
05

What this does not prove

  • Computational verification up to a limit is not a proof of the Collatz Conjecture.
  • The platform is an autonomous public exploration system, not a claimed mathematical proof.
  • Catalog records can support inspection, reproducibility, and public discussion without implying that the conjecture is solved.