Problems/Collatz Conjecture
OpenMathematics

Can two elementary rules force every positive integer to reach the same small cycle?

Collatz Conjecture

Collatz is less about a future industry and more about building better AI-math, formal verification and distributed mathematical-compute infrastructure today.

01

The question

Understand it in two minutes

Collatz is less about a future industry and more about building better AI-math, formal verification and distributed mathematical-compute infrastructure today.

Starting with 6 gives 6 → 3 → 10 → 5 → 16 → 8 → 4 → 2 → 1. Some numbers fall quickly. Others rise far above their starting value before coming down. A counterexample could either grow without bound or enter a different cycle. Computers have checked an enormous finite range, and deep theorems describe the behavior of almost all starting values in specific density senses. Neither result covers every positive integer.

01 · Step 1

The pattern

Starting with 6 gives 6 → 3 → 10 → 5 → 16 → 8 → 4 → 2 → 1.

02 · Step 2

The translation

Some numbers fall quickly. Others rise far above their starting value before coming down. A counterexample could either grow without bound or enter a different cycle.

03 · Step 3

The open gap

Computers have checked an enormous finite range, and deep theorems describe the behavior of almost all starting values in specific density senses. Neither result covers every positive integer.

Why people care

Arithmetic dynamics

The problem studies repeated integer maps, cycles, growth and the boundary between deterministic structure and probabilistic behavior.

Proof versus computation

Its vast computational support makes it a clean lesson in why finite verification is not a universal proof.

Algorithms and formalization

Efficient verification, distributed computing, exact arithmetic and proof assistants all use it as a demanding test case.

02

Possibilities

What it could unlock

What might become possible?

These are evidence-labeled possibilities, not predictions.

AI for MathematicsStrongly supported

Formal Mathematics Benchmarking

Evaluate proof agents on source-linked research statements, known partial results, false-proof traps and machine-checkable subgoals.

Why it connects: Collatz offers finite certificates and graded partial statements.

See the possibility →
03

Ideas

Ideas it inspires

What could someone explore today?

Idea seeds for curiosity—not instructions or promises.

Computational mathematicians and open-science communitiesStrongly supported

Reproducible Counterexample Lab

Package finite mathematical searches as deterministic jobs with exact versions, logs, checkpoints and independently checkable certificates.

Why it is thinkable now

Yes; Collatz and zeta-zero verification provide concrete starting examples.

AI-math labs, proof-assistant teams and research evaluatorsStrongly supported

Formal Proof-Agent Benchmark

A graded benchmark of formal statements, known lemmas, finite certificates, dependency retrieval and adversarial false-proof traps.

Why it is thinkable now

Yes, using existing Lean repositories and source-backed partial results.

04

Keep in mind

The honest boundary

Average downward drift does not control every deterministic orbit

Density results and vast computations can still leave an infinite exceptional set containing a divergent trajectory or a nontrivial cycle.
What we know

Verified below 2^71: A distributed CPU/GPU computation checked convergence for every starting value below 2^71. Almost all orbits attain almost bounded values: Tao proved that, in logarithmic density, almost every starting value eventually reaches a value below any chosen function tending to infinity. No alternate positive cycle is known: The only observed positive cycle is 1 → 4 → 2 → 1, while substantial lower bounds exclude many possible nontrivial cycles.

Latest meaningful signal

Verification extends below 2^71

A peer-reviewed distributed computation confirms convergence for every starting value in a new enormous finite range.

Sources behind this explanation 5