REVIEW
Golden-ratio growth of Conway's subprime closure
Number Theory · Submitted 12 Sep 2026 · 11 references
Pith's verdict: The combinatorial machinery is elegant, but the central analytic lemma (Lemma 3.2) claims an error bound far beyond what the cited references support, and the entire inductive proof depends on it.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
Conway's subprime closure grows at the golden ratio
The reading
What carries the argument
The central mechanism is a pair of analytic lemmas: the Goldbach filling lemma (Lemma 3.2), which asserts that if all primes up to X are present in C_n then at most O_A(X/log^A X) integers up to X are missing from C_{n+1}, and the interval-restricted binary lemma (Lemma 3.3), which guarantees representations of odd numbers as 2p+q with p and q in prescribed prime intervals except for a sparse exceptional set. These lemmas allow the induction to extend the complete prime prefix by approximately a Fibonacci step, and a refined gap-closing argument forces the asymptotic constants for M_n and Q_n to coincide.
Extended reading notes
Core claim
The paper establishes Theorem 1.1: there is a positive constant c such that the maximum element M_n and the largest prime Q_n completely present in the n-th closure both satisfy M_n ∼ Q_n ∼ c φ^n, and the cardinality |C_n| ∼ c $φ^{{n-1}}$. It follows that lim_{n→∞} |C_{n+1}|/|C_n| = φ and lim_{n→∞} |C_n|/M_n = 1/φ. The proof uses a Goldbach-type filling lemma to extend complete prime prefixes, a contracted recurrence argument to extract asymptotic constants, and a comparison of the two constants to force equality, ultimately deriving the golden-ratio growth.
Load-bearing premise
The Goldbach filling lemma assumes that known results on representing even numbers as sums of two nearly equal primes (Baker-Harman and Coppola-Laporta) apply to the specific intervals and with the uniform error bounds required by the induction, and that those results are correct as stated.
Editorial extensions
If this is right
- The growth rate of Conway's subprime closure is exactly the golden ratio, providing a deterministic analogue of Fibonacci growth in a number-theoretic setting.
- The set C_n is nearly complete below its maximum: the number of missing integers up to M_{n-1} is o(M_{n-1}), so the closure fills almost all integers up to its current bound.
- The result confirms that the subprime operation produces a robust combinatorial structure with precise asymptotics, opening the possibility of similar analyses for other divisor-based iterations.
- Because the proof relies on deep Goldbach-type estimates, it ties an elementary iterative process to advanced analytic number theory, establishing a new bridge between the two areas.
Reading between the lines
- The proof does not compute the constant c, so its numerical value remains open; a future calculation or an effective bound would sharpen the asymptotic picture.
- The reliance on known results about representations of even numbers as sums of two nearly equal primes means that any improvement in those estimates would directly sharpen the error bounds in the filling lemma and potentially the rate of convergence.
- The Lean 4 formalization (cited in the paper) indicates that the entire proof, including the analytic estimates, can be checked mechanically, which may encourage formal verification of other results using the circle method.
- If the Goldbach-type estimates were to fail for the specific intervals used, the entire induction would collapse, making the paper's claim conditional on those deep number-theoretic results.
Original abstract
Let $s(m)$ be the Conway subprime function define the binary operation on the natural numbers $x \circ y= s(x + y)$, and denote by $C_n$, $n \ge 0$, the sequence of subsets of natural numbers defined by $C_0 = \{1\}$, and $C_{n+1} = C_n \cup (C_n \circ C_n)$. We prove the conjecture by Caragiu, Vicol and Zaki that $$\lim_{n\to \infty} \frac{ | C_{n+1}| }{ | C_n |}= \frac{1+\sqrt{5}}{2}.$$ The underlying mathematical proof in this paper was constructed with some algorithmic assistance from GPT-6 Astra and its correctness has been formally verified using the Lean 4 proof assistant.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Let $s(m)$ be the Conway subprime function define the binary operation on the natural numbers $x \circ y= s(x + y)$, and denote by $C_n$, $n \ge 0$, the sequence of subsets of natural numbers defined by $C_0 = \{1\}$, and $C_{n+1} = C_n \cup (C_n \circ C_n)$. We prove the conjecture by Caragiu, Vicol and Zaki that $$\lim_{n\to \infty} \frac{ | C_{n+1}| }{ | C_n |}= \frac{1+\sqrt{5}}{2}.$$ The underlying mathematical proof in this paper was constructed with some algorithmic assistance from GPT-6 Astra and its correctness has been formally verified using the Lean 4 proof assistant.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Let $s(m)$ be the Conway subprime function define the binary operation on the natural numbers $x \circ y= s(x + y)$, and denote by $C_n$, $n \ge 0$, the sequence of subsets of natural numbers defined by $C_0 = \{1\}$, and $C_{n+1} = C_n \cup (C_n \circ C_n)$. | background 0.97 | 0.27 | 0.03 | asserted 0.54 |
| s1 | We prove the conjecture by Caragiu, Vicol and Zaki that $$\lim_{n\to \infty} \frac{ | C_{n+1}| }{ | C_n |}= \frac{1+\sqrt{5}}{2}.$$ The underlying mathematical proof in this paper was constructed with some algorithmic assistance from GPT-6 Astra and its correctness has been formally verified using the Lean 4 proof assistant. | result 0.82 | 0.90 | 0.95 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.