Skip to content
mangeshraut712Public

About

Erdős problem 142 research notebook — OPEN (single final obstruction). Audits, verifiers, roadmap. © Mangesh Raut

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

40 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Erdős 142 research notebook

Status: OPEN. Research notes, proofs of auxiliary statements, finite verifiers, and a record of failed approaches toward the local target

$$r_4(N)=o(N/\log N).$$

This repository does not prove that target or solve official Erdős #142. The official problem asks for asymptotic formulae for the progression-free extremal functions; the local target is a related, weaker question.

By Mangesh Raut. Research through 24 October 2026, including correction of the October 21 proof claims.

Current status · Proof correction · LaTeX manuscript · Peer review guide · Proof dependencies · Reproduce checks · Next steps

Corrected research status — 24 October 2026

The October 21 closure and explicit-rate claims are withdrawn. The proof reversed a variance inequality, dropped an uncontrolled cubic term, and assumed spectral concentration without proving it. The target, DIRECT-Q-GAIN, and conditional P12/W8 remain open because of these gaps.

Read the proof correction, which includes exact rational counterchecks and a corrected tight-frame certificate. Finite diagnostics are not eligible asymptotic counterexamples.

The October 23 continuation proves a sharp completion-mean interval and valid variance upper bound, and constructs an abstract zero-mean obstruction. The arithmetic low-mean branch remains open.

The overlap continuation improves the completion bound to q<=2aw and proves a consistent local-law obstruction through floor(1/a)+1 vertices. Neither supplies the missing uniform phase gain.

The October 24 continuation proves a weighted incidence-operator bound and constructs actual zero-completion subsets in growing patterns. The missing transfer from arbitrary triples to weighted arithmetic triples remains open.

The bilinear pair incidence audit (BP24) proves the universal ceiling |E_Omega[psibar Kf]|<=6w a^(3/2) and exact spatial decoupling <u, Kf> = -a <g, f>, showing that pairing with K cannot provide the diverging factor F_epsilon(a)->infinity needed by DIRECT-Q-GAIN.

The multi-scale Bohr overlap continuation (MS24) proves universal density discrepancy invariance on overlaps and extends spatial decoupling <u_1, K_{1->2} f> = -a_1 <g_{1->2}, f> across arbitrary multi-scale hosts, proving that multi-scale host coupling cannot bypass the hole-support obstruction.

Retained result Scope
Boolean expansion and quartic covariance identity Exact identities; the W8 inequality is still missing
Twelve-slot terminal theorem 7/11 ceiling under the stated M13 regime
Positive-density branch reduction Uses classical qualitative Szemerédi; no quantitative rate
October 19 covariance obstruction Contains a 4-AP; does not refute P12
Tight-frame reconstruction and partial-spectrum certificate No automatic concentration, coefficient bound, or phase selection
Universal Boolean chirp fourth moment Holds even for AP-rich sets; does not imply the missing cross-spectrum concentration
Universal bilinear pair incidence ceiling (BP24) Universal 6w a^(3/2) ceiling; spatial decoupling on hole support
Multi-scale Bohr overlap decoupling (MS24) Universal density discrepancy invariance and cross-host hole annihilation

The manuscript and disputed drafts are clearly marked withdrawn, with previous versions preserved. Lean formalization remains incomplete and contains an admitted lemma and additional axioms.

Reproduce the checks

Use Python 3.10+. The recent Fourier checkers require NumPy; the thirteen original September checkers use the standard library.

python3 -m venv .venv
source .venv/bin/activate
python -m pip install -e ".[verification,test]"
python verification/check_repository.py
python verification/generate_dashboard.py --check
python -m unittest discover tests
python verification/run_current.py --suite all --output .verification-runs/latest

The runner executes in a temporary copy so saved JSON evidence is preserved. current selects twenty-five current suites and diagnostics, legacy selects the thirteen original suites, and all runs both (38 scripts). Earlier dated experiments remain available separately. A passing run checks finite identities and examples; it does not prove an asymptotic theorem. See VERIFYING.md for exact versus floating checks, optional SciPy searches, and limitations.

Reading and repository map

The interactive status dashboard displays live verifier suite passes and metrics across the research pack. Older card snapshots in docs/screenshots/ are preserved as historical references.

Contributing, attribution, and citation

Contributions should state exact hypotheses, preserve failed attempts, and include a relevant verifier or explain why a claim is purely analytic. Follow CONTRIBUTING.md, including signed-off commits.

Cite: Mangesh Raut, Erdős problem 142 research pack, 2026, GitHub repository. Machine-readable metadata: CITATION.cff.

Copyright 2026 Mangesh Raut. Apache-2.0, with NOTICE and attribution guidance. Support the author.

About

Erdős problem 142 research notebook — OPEN (single final obstruction). Audits, verifiers, roadmap. © Mangesh Raut

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Sponsor this project

Packages

Used by

Contributors

Languages