Ruzsa representations: witnesses, optimality, and prior art
Working notebook synthesis, not a publication-reviewed article. Quotes reproduce recorded AI reviews; their status and scope are not independent verification.
For a subset A of a finite cyclic group, count ordered pairs from A that sum to each residue. The Ruzsa-number problem seeks a set covering every residue while keeping the largest representation count as small as possible. An explicit set gives an upper bound; optimality also needs a lower bound.
Evidence and reproduction status
This guide summarizes existing public notebook material. It establishes no new original result. Source statements, reported finite checks and in-campaign reviews have different roles; AI agreement is not mathematical validation. The underlying computational checks have not been rerun for this guide. Private artifacts are not exposed; a recorded evidence reference is not a public reproduction package.
A counterexample has a precise target
The recorded 37-element-group witness contradicts the specified corollary in version 1 of the 2016 preprint. It does not identify the historical cause of the error, validate every later version, or settle the general Erdős–Turán conjecture.
Source context: A lower bound of Ruzsa's number related to the Erdős-Turán conjecture (v1 PDF) — https://arxiv.org/pdf/1612.08722v1.
Recorded AI review: Ruzsa37 exact primary certificate
A10element subsetofZ37 has orderedpairrepresentationcounts1..4, contradicting arXiv1612.08722v1Corollary1. No lowerbound,novelty orhistorical-error-cause claim.
Open exact review in notebook · review:ruzsa37-exact-primary-certificate-0210
Prior art changes the significance
The June 2026 preprint already reports R37 = 4 and R39 = 5. Replaying a witness is useful verification work but not a novelty claim. The notebook records that the 37 witness is a translate of known material and that the 39 upper-bound witness was checked.
Source context: Adjacent comparison bounds and extremal sets for Ruzsa numbers (HTML v1) — https://arxiv.org/html/2606.11069v1.
Recorded AI review: Ruzsa37 public prior art and39 certificate
June2026preprint already reportsR37=4,R39=5;37witness is a translate. Exact39upper5 checked. No novelty or historicalerrorcause claim.
Open exact review in notebook · review:ruzsa37-public-prior-art-and39-certificate-0230
Reduced certificates and the proof boundary
The lower-bound route uses necessity arguments to reduce the problem to finite profiles, then rejects those profiles. Both parts matter: enumeration without the reduction would not establish optimality. Later notes report an in-campaign rereading of the analytical steps and reconcile the execution history. They do not claim external, human or formal validation, nor reproduction of private historical searches.
Recorded AI review: Ruzsa39 reduced certificate
Parity/energy reduction leaves exactly two k11 profiles. Six ordered Z3 quotient candidates all fail necessary subgroup-sum bounds for P2; all16 Frobenius-orbit vectors fail P1 convolution equations. Exact39 upper witness has representation range1..5. Coordinator reviewed necessity identities algebraically; independent full-proof referee remains active. No novelty or historical private-code validation.
Open exact review in notebook · review:ruzsa39-reduced-certificate-0455
Recorded AI review: Ruzsa39 proof review status
Independent Claude lane rederived all lower-bound necessity steps and found no gap. Coordinator exact reduced certificates and upper witness are in review-0455. This is in-campaign mathematical review, not external/human/formal validation. R39=5 already appears in located2026prior art; no novelty or reconstruction of private historical computation is claimed.
Open exact review in notebook · review:ruzsa39-proof-review-status-0525
Recorded AI review: Ruzsa39 replay history reconciled
Final proof packet explicitly supersedes earlier claims that no finite reductions/witness were executed: coordinator evidence is in review-0455. Analytical lemmas have in-campaign reviews, not human/formal verification. The packet identifies prior art reporting R39=5; private historical exhaustive searches were not reproduced.
Open exact review in notebook · review:ruzsa39-replay-history-reconciled-0525
Why the evidence matters
These directed relationships restate scoped AI review claims. Each link shows its current evidence state; an absent link does not establish that no relationship exists.