∑ Math Proof PlaygroundResearch › Open campaigns

Open research campaign archive

Substantial investigations that generated reductions, computations, or proof infrastructure—but no complete proof of the famous target problem.

This page exists to prevent progress from being mistaken for a solution. A conditional theorem, a large finite search, or a promising invariant is recorded as progress only.

Goldbach conjecture

Conditional reductionsUnsolved

The campaign isolated the real obstruction as a pointwise, lower-bound-compatible Type-II/Type-III estimate capable of separating primes from semiprimes—the sieve parity problem.

Concrete progress included:

  • the reflected correlation Σ Λ(n) μ(N−n) as a parity-breaking target;
  • a conditional 1+4/3 bridge reducing the residual factor structure;
  • identification of the critical factor geometry near (1/4,3/8,3/8) and modulus scale N^(5/8);
  • a rough-sieving criterion showing how a sufficiently strong weighted distribution theorem would force the residual factor to be prime.

Status: no unconditional estimate strong enough to close the bridge was proved. Goldbach remains open.

Checkerboard Dmono(22)

Exact finite frontier33 or 34 unresolved

The exact finite campaign established

33 ≤ Dmono(22) ≤ 34.

Boundary regimes and hundreds of canonical refinement cases were eliminated, no 34-point witness was found, and no terminal completed branch returned anything other than UNSAT. The remaining central symmetric cases are the hardest.

Status: no complete proof that 34 is impossible and no 34-point construction.

No-three-in-line on the 71×71 grid

Search campaignNo complete construction

Multiple algebraic, modular, SAT, and local-improvement ideas were explored. No independently checkable optimal or target-size configuration was produced, and no upper-bound proof settled the instance.

Collatz conjecture

Invariant explorationUnsolved

The work explored Lyapunov functions, residue-class descent, entropy drift, and ways to strengthen “almost all” results into a pointwise descent mechanism. No global monotone invariant or complete termination proof was found.

P versus NP

Route analysisUnsolved

The campaign compared circuit lower bounds, proof complexity, algebraic methods, meta-complexity, natural-proofs barriers, relativization, and formalization possibilities. It produced no separation theorem.

Hadwiger–Nelson problem

Computational strategyNo improved bound

Potential SAT/graph-generation routes to a 6-chromatic unit-distance graph and improved colorings were mapped, but no new unit-distance graph or coloring bound was certified.

Kaprekar dynamics in general digit lengths

Structural explorationNo new general theorem certified

The work examined fixed points, cycles, digit-count analogues of 6174, and possible finite-state classifications. No historically new full classification was established.

Other named frontiers

ProblemRecorded status
WOWII 133One branch proved; residual case open.
WOWII 314Paper classification candidate; top-level Lean bridge missing.
Full checkerboard all-slope limitFour-direction work complete at draft level; all-slope lower bound open.
Erdős 885Deep arithmetic-geometry progress; rational-point classification incomplete.
Erdős 3669,718,456 3-full candidates checked without a witness; open.
← Research dashboard