Pith. sign in

REVIEW 4 minor 13 references

Random linear configurations in dense sets and primes

T0 review · 0 major / 4 minor · reviewed 2026-07-31 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Polylog-dense sets of integers and primes contain almost every random linear configuration.

desk verdict Solid quantitative additive-combinatorics paper: polylog density forces almost-all random translation-invariant configurations in [N], and relatively dense primes in a shorter range, via a new uniform GvN + degree-lowering to U^{1+}. read the letter →

arxiv 2607.28091 v1 pith:NJS4YR76 submitted 2026-07-30 math.NT math.CO

Lasse Grimmelt , Joni Teräväinen This is my paper
classification math.NTmath.CO MSC 11B3011N1337A45
keywords linearconfigurationsGowersnormsU^{1+}normdegreeloweringgeneralisedvonNeumannprimesdensificationdensityincrement
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper shows that once a subset of the first N integers is denser than a fixed power of 1/log N, it already contains nontrivial patterns of the shape x plus b_i times m for almost every choice of the coefficient vector b of size B, across a very wide range of B. The same conclusion holds for subsets of the primes that are only polylogarithmically dense relative to the primes, though the allowed range of B is shorter. The patterns are the natural translation-invariant linear configurations in two variables; arithmetic progressions are the special case b_i = i. Because the coefficients are allowed to grow, classical Gowers-norm control loses uniformity; the authors restore it by averaging over b, then lower the resulting high-degree control all the way to the U^{1+} norm, which detects long arithmetic progressions. A densification step transfers the bounded result to the primes via a truncated sieve majorant. The upshot is a quantitative existence theorem that is far stronger than what is known for any fixed large coefficient vector.

What carries the argument

A quantitative generalised von Neumann theorem that controls the averaged counting operator R_H by the U^{1+} norm: after iterated Cauchy–Schwarz and concatenation produce U^k control, degree-lowering (dual-difference interchange plus major-arc analysis of phases) reduces the degree to U^{1+}, which is strong enough for a density-increment argument.

What would settle it

Exhibit a subset of [N] denser than N/(log N)^{c} that avoids the configuration x+b_i m for a positive-density set of b in ((B/2,B])^k inside the claimed range of B, or show that the truncated GPY majorant fails the linear-forms condition for some admissible system when B exceeds exp(w^{1/2}).

Watch

Extended reading notes

Core claim

Every subset of [N] denser than N/(log N)^{c_k} contains a nontrivial configuration x+b_1 m,...,x+b_k m with m in [N/B] for all but an O((log N)^{-c_k}) proportion of coefficient vectors b of size B, provided (log N)^{1/c_k} ≤ B ≤ N exp(-(log N)^{c'_k}). The identical statement holds for relatively polylog-dense subsets of the primes when B is at most exp((log N)^{c_k}).

Load-bearing premise

The transfer to the primes needs a truncated sieve majorant that obeys a two-scale linear-forms condition only when the coefficient size B stays below exp of the square root of the small-prime level; larger B creates local Euler-factor obstructions the paper does not remove.

Editorial extensions

If this is right

  • Almost every translation-invariant linear pattern in two variables appears in every polylog-dense set of integers, far beyond the range known for any fixed pattern with large coefficients.
  • The same almost-everywhere statement holds inside the primes once relative density exceeds a power of 1/log N, for coefficients up to exp((log N)^c).
  • The U^{1+} inverse theorem and the averaged concatenation estimates become available as black-box tools for other random or averaged configuration problems.
  • Density-increment arguments that previously required fixed small coefficients now run uniformly for growing random coefficients.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the open 'almost-all forms' pseudorandomness condition suggested in the paper can be verified, the prime result would reach the same coefficient range as the integer result.
  • The same averaging-plus-degree-lowering strategy should apply to other sparse or unbounded settings once a suitable two-scale majorant is available.
  • Quantitative bounds for almost-all configurations may be convertible into effective bounds for a positive-density set of explicit coefficient vectors by a second-moment argument.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

0 major / 4 minor

Summary. The paper proves that every subset A of [N] with density at least (log N)^{-c_k} contains a nontrivial configuration x+b_1 m,...,x+b_k m (m in [N/B]) for all but an O_k((log N)^{-c_k})-proportion of coefficient vectors b in ((B/2,B]\cap Z)^k, whenever (log N)^{1/c_k}\le B\le N exp(-(log N)^{c'_k}). An analogous statement holds for relatively polylog-dense subsets of the primes, but only in the shorter range B\le exp((log N)^{c_k}). The argument proceeds by a quantitative generalised von Neumann theorem giving U^k control of the averaged counting operator R_H (Prop. 3.7), degree lowering of the dual function first to U^2 and then to the U^{1+} norm (Lemmas 4.5–4.7), densification under a two-scale linear-forms condition (Thm. 6.7), verification of that condition for a truncated GPY majorant (Prop. 7.4), and a density-increment iteration (Lem. 8.4).

Significance. The results give a polylogarithmic density threshold for almost all translation-invariant linear configurations in two variables, substantially stronger than the best known bounds for fixed configurations (Kelley–Meka/Bloom–Sisask for 3-APs, Green–Tao for 4-APs, Leng–Sah–Sawhney for longer APs). The same threshold is obtained relatively in the primes, albeit in a shorter coefficient range forced by the sieve. The technical contributions—a polynomial-loss GvN uniform in the coefficient scale B, degree lowering all the way to U^{1+}, and a carefully truncated two-scale GPY majorant—are of independent interest and are tracked with explicit polynomial dependencies throughout. The limitations of the prime range are stated honestly (Rem. 7.5).

minor comments (4)
  1. [Theorems 1.1 and 1.2] The constant c_k is used both as a density exponent and (with a different value) as a range exponent; a brief remark in the statements of Theorems 1.1–1.2 that the same symbol may stand for different positive constants depending only on k would avoid any momentary confusion.
  2. [Lemma 4.7] In the proof of Lemma 4.7 the smoothing parameter A=10 is fixed without comment; a parenthetical that any A>1 works and that 10 is chosen only for convenience would make the dependence clearer.
  3. [Introduction / Remark 7.5] Remark 7.5 already notes that an “almost-all forms” linear-forms condition might enlarge the prime range of B. A one-sentence forward reference in the introduction (after the statement of Theorem 1.2) would help the reader anticipate this limitation.
  4. [Equation (1.9)] Typographical: “Parithmetic progression” in (1.9) should be “P arithmetic progression”; a few other missing spaces appear in the same display.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: self-contained analytic derivation from standard inequalities and external cited inputs

full rationale

The paper is a pure existence/quantitative additive-combinatorics argument. Theorems 1.1–1.2 are obtained from a written chain: iterated Cauchy–Schwarz (Lemma 3.3) plus quantitative concatenation (Lemma 3.5 / Cor. 3.6) give U^k control (Prop. 3.7); degree lowering (Lemmas 4.5–4.7, following Peluse–Prendiville) reduces to U^{1+}; densification (Thm. 6.7) and a truncated GPY majorant (Prop. 7.4, extending Green–Tao) transfer the bound to the primes; density increment (Lem. 8.4) finishes. Constants c_k, C_k are existential “sufficiently small/large” parameters, not fitted to data. Self-citations ([12], [13]) supply methodological templates (densification, relative inverse theorems) whose proofs are re-derived or adapted in-line with explicit linear-forms hypotheses; they do not force the main claims by definition or uniqueness import. No prediction is statistically forced by a fit, and no equation is equivalent to its input by construction. Honest non-finding: score 0.

Assumptions & free parameters 1 free parameters · 5 assumptions · 2 invented entities

The paper is a pure-math existence proof. It relies on standard analytic and combinatorial inequalities, on previously established inverse and concatenation theorems, and on the Green–Tao / GPY sieve infrastructure. No numerical parameters are fitted to data; the only free choices are the usual “sufficiently small/large” absolute constants that absorb polynomial losses. No new physical or combinatorial entities are postulated beyond ordinary counting operators and the already-known U^{1+} norm.

free parameters (1)
  • c_k, c'_k, C_k (existential density and range exponents) = existential; depend only on k
    Chosen sufficiently small or large in terms of k only, to absorb polynomial losses from Cauchy–Schwarz, concatenation, and degree lowering. Not fitted to external data.
assumptions (5)
  • standard math Standard properties of Gowers norms U^s, including the recursive identity and the U² inverse theorem via Fourier analysis.
    Used throughout §§3–5; classical.
  • domain assumption Quantitative concatenation lemma in the spirit of Peluse–Prendiville (Lemma 3.5).
    Proved in the paper by a Fourier-free argument, but the strategy is taken from the cited nonlinear Roth work.
  • domain assumption Green–Tao / GPY linear-forms estimates for sieve weights, extended to two scales (Lemma 7.3, Prop. 7.4).
    The majorant construction and local-factor analysis follow Green–Tao Appendix D, with extra uniformity in coefficients of size B.
  • standard math Shiu’s bound on averages of divisor powers in arithmetic progressions.
    Invoked in Lemma 7.2 to control divisor tails after truncation.
  • standard math Vinogradov’s lemma on simultaneous small fractional parts.
    Used in the major-arc analysis inside degree lowering (Lemmas 4.5, 4.7).
invented entities (2)
  • Counting operators r_H(b; f₁,…,f_k) and R_H(λ; f₁,…,f_k) independent evidence
    purpose: Package the multilinear configuration count and its average over coefficient vectors so that inverse theorems can be stated cleanly.
    Standard multilinear averages rewritten for the paper’s notation; not a new ontological object.
  • Normalised dual function D_μ and the (K,L,η) two-scale linear-forms condition independent evidence
    purpose: Make densification and majorant verification quantitative at scales (N,B,H).
    Technical packaging of existing densification and pseudorandomness ideas; the two-scale formulation is adapted to the paper but not a new physical entity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Random linear configurations in dense sets and primes." pith.science (2026). https://pith.science/paper/NJS4YR76

@misc{pith2026260728091,
  author       = {Pith},
  title        = {Pith review of: Random linear configurations in dense sets and primes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NJS4YR76}},
  note         = {Machine review of arXiv:2607.28091}
}
abstract

We prove that every polylogarithmically dense subset of $[N]$ contains a nontrivial configuration $x+b_1m,\ldots,x+b_km$ for almost all choices of the coefficient vector $(b_1,\ldots, b_k)$ in a wide range of scales. We prove the same statement for polylogarithmically relatively dense subsets of the primes, in a shorter range of scales. The main ingredients are a new quantitative generalised von Neumann theorem, degree lowering to the $U^{1+}$ norm, and densification arguments that transfer the result to the primes.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

13 extracted references · 3 linked inside Pith

  1. [1]

    T. F. Bloom and O. Sisask. An improvement to the Kelley-Meka bounds on three-term arithmetic progressions.arXiv e-prints, page arXiv:2309.02353, September 2023

    arXiv 2023

  2. [2]

    Conlon, J

    D. Conlon, J. Fox, and Y. Zhao. A relative Szemer´ edi theorem.Geom. Funct. Anal., 25(3):733–762, 2015

    2015

  3. [3]

    Green and T

    B. Green and T. Tao. The primes contain arbitrarily long arithmetic progressions.Ann. of Math. (2), 167(2):481–547, 2008

    2008

  4. [4]

    Green and T

    B. Green and T. Tao. Linear equations in primes.Annals of Mathematics, 171(3):1753–1850, 2010

    2010

  5. [5]

    Green and T

    B. Green and T. Tao. The quantitative behaviour of polynomial orbits on nilmanifolds.Ann. of Math. (2), 175(2):465–540, 2012

    2012

  6. [6]

    Green and T

    B. Green and T. Tao. New bounds for Szemer´ edi’s theorem, III: a polylogarithmic bound forr 4(N). Mathematika, 63(3):944–1040, 2017

    2017

  7. [7]

    Kelley and R

    Z. Kelley and R. Meka. Strong bounds for 3-progressions, 2023. In2023 IEEE 64th Annual Symposium on Foundations of Computer Science (FOCS), pp. 933–973.arXiv:2302.05537

    arXiv 2023

  8. [8]

    J. Leng, A. Sah, and M. Sawhney. Improved Bounds for Szemer´ edi’s Theorem.arXiv e-prints, page arXiv:2402.17995, February 2024

    arXiv 2024

Show all 13 references
  1. [9]

    Peluse and S

    S. Peluse and S. Prendiville. Quantitative bounds in the nonlinear Roth theorem.Invent. Math., 238(3):865–903, 2024

    2024

  2. [10]

    P. Shiu. A Brun-Titchmarsh theorem for multiplicative functions.J. Reine Angew. Math., 313:161–170, 1980

    1980

  3. [11]

    T. Tao. The Gowers uniformity norm of order 1+. Blog post,https://terrytao.wordpress.com/ 2021/07/25/the-gowers-uniformity-norm-of-order-1/, 2021

    2021

  4. [12]

    Tao and J

    T. Tao and J. Ter¨ av¨ ainen. Quantitative bounds for Gowers uniformity of the M¨ obius and von Mangoldt functions.Journal of the European Mathematical Society, 27(4):1321–1384, 2025

    2025

  5. [13]

    Ter¨ av¨ ainen and M

    J. Ter¨ av¨ ainen and M. Wang. On the Green-Tao theorem for sparse sets.arXiv e-prints, page arXiv:2603.09281, March 2026. 58 LASSE GRIMMELT AND JONI TER ¨AV ¨AINEN Department of Pure Mathematics and Mathematical Statistics, University of Cambridge, Cambridge CB3 0WB, UK Email...

    2026

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.