{"id":"52046dad-490f-4557-9626-803c03f18a5a","arxiv_id":"2508.18559","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For free Borel Z^d-grids, every grid has a Borel (2^d-1)-polychromatic coloring, while ergodic grids admit no Borel 2^d-polychromatic coloring, so the Borel threshold is 2^d-1.","lead":"This paper proves that every grid from a free Borel action of the integer group Z^d admits a Borel coloring in which every unit cube contains 2^d-1 colors, and that no ergodic action admits a coloring with all 2^d colors. The pair of results fixes the Borel polychromatic number of such grids at exactly 2^d-1 under ergodicity.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sharpness claim is false: a free ergodic Z^d-action can have a Borel 2^d-polychromatic coloring, so the universal \"exactly 2^d-1\" statement is overclaimed.","rationale":"The reader's weakest assumption concerned the toast lemma's parameterization in the upper-bound proof, but the load-bearing flaw I find is in the sharpness claim. The upper-bound construction of Theorem 3.3 may well be correct; my concern is that the paper's advertised characterization, 'the Borel polychromatic number of a free Z^d-action is exactly 2^d-1', is false as stated. The explicit product construction is a free ergodic action that admits a Borel 2^d-coloring, so the abstract's assertion that ergodic generators preclude such a coloring is incorrect. This is not a matter of missing detail or an overly compressed proof; it is a counterexample to the central claim. The paper could be revised to state only the upper bound for all actions and the lower bound for actions with a coordinate subaction lacking a Borel 2-coloring, but as submitted the main result is overstated. I therefore recommend rejection of the current version.","tokens_in":10060,"tokens_out":37298,"duration_ms":358930,"concrete_test":"Instantiate the counterexample for d=2: take Y = {0,1}^(Z^2) with Bernoulli shifts T_0,T_1, X = Y x (Z/2)^2, e_0(y,eps)=(T_0 y, eps+(1,0)), e_1(y,eps)=(T_1 y, eps+(0,1)). Verify (1) the action is free, (2) each of e_0,e_1 and the whole Z^2 action is ergodic for the product measure, and (3) c(y,eps)=eps assigns all four colors to every unit square. If all three checks pass, the abstract's universal ergodic obstruction is refuted; the same construction works for all d.","verdict_should_be":"REJECT","load_bearing_attack":"The abstract and Section 5 claim that every free Z^d-action has Borel polychromatic number exactly 2^d-1, and specifically that any action whose generators act ergodically admits no Borel 2^d-polychromatic coloring. Section 4, however, only proves the obstruction under a stronger condition: some generator's Z-subaction must fail to admit a Borel proper 2-coloring (Lemma 4.1). Ergodicity of a generator does not imply this. A concrete counterexample exists for every d. Let Y = {0,1}^(Z^d) with the free Bernoulli shift action T_i, and let X = Y x (Z/2)^d. Define e_i(y, eps) = (T_i y, eps + delta_i), where delta_i is the ith standard basis vector in (Z/2)^d. This is a free Borel action of Z^d. Each generator e_i is ergodic: if A is e_i-invariant, decomposing A by the eps coordinate gives A_eps = T_i^{-1}(A_{eps+delta_i}), so A_eps is T_i^2-invariant; since T_i^2 is ergodic, each A_eps has measure 0 or 1, and the coupling across eps forces A to be null or conull. The full action is ergodic by the same argument. Yet c(y, eps) = eps is a Borel coloring, and on every unit cube {e_0^{a_0} ... e_{d-1}^{a_{d-1}} x : a_i in {0,1}} the second coordinate ranges over all 2^d elements of (Z/2)^d. Thus a Borel 2^d-polychromatic coloring exists, directly contradicting the claimed sharpness for ergodic generators. The valid theorem would be: every free action admits a Borel (2^d-1)-coloring, and no Borel 2^d-coloring exists when some coordinate subaction has no Borel 2-coloring.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Borel polychromatic colorings of the Schreier graph of a free Borel action of Z^d, where every unit d-cube must contain all colors. The positive result, Theorem 3.3, asserts that every such grid admits a Borel (2^d-1)-polychromatic coloring, constructed from a Borel toast decomposition and a local interpolation lemma. The paper further claims sharpness: it asserts in the abstract and in Section 5 that the Borel polychromatic number of every free Z^d-action is exactly 2^d-1, and in particular that no Borel 2^d-coloring exists when the generators act ergodically. Section 4 gives a more limited obstruction, Lemma 4.1, showing that absence of a Borel proper 2-coloring of some generator's Z-subaction implies absence of a Borel 2^d-polychromatic coloring, and it analyzes the d=2 case in Theorem 4.5.","tokens_in":10461,"tokens_out":20785,"duration_ms":213546,"significance":"If the positive existence statement is correct, it is a genuinely interesting contribution to Borel combinatorics: it shows that a local polychromatic constraint can be met with 2^d-1 colors in a Borel way for every free Z^d-action, using the highly nonconstructive toast machinery in a clever way. The local interpolation lemma is elegant and potentially reusable. The paper also correctly identifies a rigidity phenomenon for 2^d-colorings, namely that they force Borel 2-colorings of the coordinate subactions. However, the advertised sharpness theorem is false as stated, and this reduces the significance of the paper unless the claims are corrected. The positive theorem remains valuable even after the sharpness overclaim is removed.","major_comments":[{"comment":"The claimed sharpness is false. Consider the action e_i(y,epsilon) = (T_i y, epsilon + delta_i) on X = {0,1}^{Z^d} x (Z/2)^d, where T_i is the Bernoulli shift and delta_i is the i-th standard basis vector. This action is not free on all of X, but its free part F is a conull invariant Borel subset, so the restricted action is free and each generator is ergodic on F. The map c(y,epsilon)=epsilon is a Borel 2^d-polychromatic coloring on F, because every unit cube {a.x : a in {0,1}^d} has second coordinate ranging over all of (Z/2)^d. Thus ergodicity of the generators does not prevent a Borel 2^d-coloring. Section 4 only proves the obstruction under the strictly stronger hypothesis of Lemma 4.1, namely that some generator's Z-subaction has no Borel proper 2-coloring; the abstract and Section 5 must replace the universal 'exactly 2^d-1' statement with this conditional statement and should discuss the counterexample.","section":"Abstract and Section 5"},{"comment":"The proof invokes Lemma 2.2 to take a Borel r-toast with r = (2^{d+3}+1)d, but Lemma 2.2 as stated merely says that 'a toast decomposition' exists and does not specify a radius. The existence of toasts with sufficiently large radius is load-bearing for the whole construction. Please cite the precise quantitative statement from [6] or include a proof that free Z^d-actions admit Borel r-toasts for this value of r.","section":"Theorem 3.3, choice of r"},{"comment":"The verification step is incomplete. The text asserts that every unit cube C contained in K is contained either in E_K or in S_{i,t} union S_{i,t+1}, but this omits cubes contained entirely in an internal piece L_i and cubes crossing from L_i into P_i. Also, a cube straddling two shells does not literally have c restricted to C equal to one of the two adjacent templates; the mixed assignment can be surjective because consecutive templates differ at at most one vertex, but this needs to be stated and proved. The intended argument appears repairable, but the proof as written does not cover all cases.","section":"Theorem 3.3, induction verification"}],"minor_comments":[{"comment":"The index arithmetic in Lemma 3.1 is unclear: the sequence appears to have 2^{d+1}+1 labelings, not '2d+1', and the bounds on i in the induction should be made explicit (i runs from 0 to 2^d-1, with final labeling c_{2^{d+1}}).","section":"Lemma 3.1"},{"comment":"The notation [K]_G is used without definition; please define it as the G-connected component of K.","section":"Theorem 3.3, Step B"},{"comment":"The sentence 'The choice r = 2R + d ensures...' is asserted without derivation; a short argument that the R-thickenings of distinct maximal internal pieces are separated by more than d would improve readability.","section":"Theorem 3.3, Step B"},{"comment":"The phrase 'factor through to a quotient Z_2^d of Z^d' should specify that the quotient is by (2Z)^d; this is implicit but could be stated explicitly.","section":"Definition 3.2"},{"comment":"The claim that every unit cube of G is contained in some toast piece needs a brief proof; it follows from r > d and the nesting property of toasts, but the manuscript should spell this out.","section":"End of Section 3"}],"recommendation":"major_revision","confidential_remarks":"The sharpness overclaim is the main issue: the advertised 'exactly 2^d-1' result is false, and the paper needs a careful revision of the abstract, the introduction, and Section 5. The positive existence theorem and the d=2 rigidity theorem are worth preserving. I do not recommend rejection because the false claim is separable and the corrected statements appear within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the upper-bound half of this paper is a real contribution, but the sharpness theorem as stated is false. The abstract's claim that ergodic generators prevent Borel 2^d-polychromatic colorings is refuted by a concrete free action.\n\nThe Borel polychromatic number is a sensible new notion for grid actions, and the proof of the (2^d-1)-coloring uses toast technology in a genuinely new way. Lemma 3.1, the local path of surjective labelings, is neat and likely reusable. The d=2 rigidity analysis in Theorem 4.5(2) is also a nice structural result.\n\nThe construction in Theorem 3.3 is compressed in the spots you'd expect—the Borel recursion over the toast and the claim that each cube lies in one exterior or gap region—but those look like minor presentational gaps, not actual holes.\n\nThe real problem is Section 4 and the abstract. Lemma 4.1 is fine: a Borel 2^d-polychromatic coloring forces a Borel proper 2-coloring of each generator's subaction. But ergodicity of a generator does not imply absence of such a 2-coloring. Here is the counterexample. Let Y={0,1}^{Z^d} with the Bernoulli shift in each coordinate, and let X=Y x (Z/2)^d with e_i(y,eps)=(T_i y, eps+delta_i). The action is free. Each e_i is ergodic: an invariant set A decomposes as A_eps, and the invariance gives A_eps = T_i^{-1}(A_{eps+delta_i}), so each A_eps is T_i^2-invariant; since T_i^2 is ergodic, each A_eps has measure 0 or 1, and the coupling forces all of them to have the same measure. Yet c(y,eps)=eps is a Borel 2^d-polychromatic coloring: on any unit cube, the second coordinate runs through all 2^d values. So the universal 'exactly 2^d-1' claim is wrong. What the paper actually proves is a correct but weaker dichotomy: if some coordinate subaction has no Borel proper 2-coloring, then there is no Borel 2^d-polychromatic coloring. There exist free actions, like the one above, that do admit Borel 2^d-colorings.\n\nNet: the upper bound and the new definition are worth a referee's time. The authors need to correct the sharpness statement, and the open questions in Section 5 should be revised to ask which actions admit 2^d-colorings rather than implying none with ergodic generators do. I would send it to review, with the expectation of major revision.\n\nWho is it for? People in Borel combinatorics working on hyperfinite graphs and definable colorings. The construction technique will be cited, but as it stands I wouldn't cite the sharpness claim.","headline":"New upper bound and a nice local lemma, but the advertised sharpness is false: ergodic generators do not block Borel 2^d-colorings, and a concrete product action gives a counterexample.","tokens_in":11000,"tokens_out":13670,"would_cite":false,"duration_ms":137117,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03E15","05C15","37A20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Borel polychromatic number of a free Z^d grid is exactly 2^d − 1: every such grid admits a Borel coloring using all colors on every unit cube, and one more is impossible for ergodic actions.","keywords":["Borel polychromatic coloring","grid graphs","Borel combinatorics","Z^d actions","Borel toast decomposition","ergodic actions","polychromatic number","descriptive set theory"],"falsifier":"Exhibit a free Borel action of $\\mathbb{Z}^d$ whose grid graph has no Borel $(2^d-1)$-polychromatic coloring, or exhibit a Borel $2^d$-polychromatic coloring of an ergodic action such as the Bernoulli shift; either outcome would contradict the claimed sharp value $2^d-1$.","tokens_in":9858,"feed_emoji":"🎨","tokens_out":17076,"duration_ms":144232,"temperature":0.7,"pith_summary":"This paper asks how many colors a Borel labeling of a grid graph can use while still guaranteeing that every unit $d$-dimensional cube sees every color at least once, and it answers the question for grids coming from free Borel actions of $\\mathbb{Z}^d$. In the classical setting every $\\mathbb{Z}^d$ grid admits a polychromatic coloring with all $2^d$ colors available on its $2^d$ vertices; the paper shows that imposing Borelness costs exactly one color. The main theorem constructs a Borel $(2^d-1)$-polychromatic coloring for every free Borel $\\mathbb{Z}^d$ action, and the sharpness theorem shows that when the generators act ergodically no Borel $2^d$-polychromatic coloring exists. The upshot is that the Borel polychromatic number of a free ergodic $\\mathbb{Z}^d$ grid is exactly $2^d-1$.","feed_headline":"Borel grids use 2^d−1 colors, never 2^d","feed_subtitle":"Making grid colorings Borel costs exactly one color per unit cube; ergodic grids cannot do better.","key_machinery":"The construction runs on a Borel $r$-toast: a Borel collection of finite nested pieces of the grid such that any two pieces are either $r$-apart or one is contained in the $r$-thickening of the other. Each toast piece's exterior is colored by a repetitive template, a coloring that factors through the quotient $(\\mathbb{Z}/2\\mathbb{Z})^d$ and therefore carries all $2^d-1$ colors on every cube, and the paper interpolates between the inner template of a piece and the outer template through shells around each internal piece. The interpolation step rests on Lemma 3.1, which states that any two surjective labelings of the $d$-cube with $2^d-1$ colors can be connected by a sequence of surjective labelings in which consecutive labelings differ on at most one vertex; the shells are spaced so that every unit cube meets at most two consecutive steps of such a sequence. The toast radius $r=(2^{d+3}+1)d$ is chosen large enough that the shells of distinct pieces never interfere, so the local interpolations assemble into a single Borel coloring of the whole grid.","core_discovery":"The central claim, Theorem 3.3, is that for every free Borel action of $\\mathbb{Z}^d$ on a standard Borel space, the induced Schreier grid graph carries a Borel polychromatic coloring with $2^d-1$ colors: a Borel labeling with that many colors in which every unit hypercube $\\{0,1\\}^d\\cdot x$ contains all colors. The bound is sharp: if the generators act ergodically, in particular for the Bernoulli shift, no Borel $2^d$-polychromatic coloring exists, so the Borel polychromatic number of a free ergodic $\\mathbb{Z}^d$-grid is exactly $2^d-1$. The paper also shows that any Borel $2^d$-polychromatic coloring forces a Borel proper 2-coloring of each coordinate $\\mathbb{Z}$-subaction, and it characterizes the $d=2$ case completely: a Borel 4-polychromatic coloring exists exactly when the action admits a '1-fold invariant' pair of Borel 2-colorings, and every 4-coloring is invariant under $e_0^2$ or $e_1^2$ on each orbit.","pith_inferences":["The same toast-and-shell recipe suggests a general pattern for other translation-invariant shapes $T$: when a tile admits a surjective labeling with $|T|-1$ colors and a homotopy lemma like Lemma 3.1, the Borel polychromatic number may fall to $|T|-1$ under ergodicity, a concrete route into the paper's Open Question 5.2.","The paper's dichotomy between 0-fold and $(d-1)$-fold invariant tuples raises the testable conjecture that for $d\\ge 3$ any Borel $2^d$-coloring forces orthogonally invariant 2-colorings in all but at most one direction; a finite periodic-grid search could look for a 0-fold-invariant tuple that still yields a $2^d$-coloring, which would refute the conjecture.","Because Lemma 4.1 pins the obstruction at a single generator, the sharpness threshold may extend beyond globally ergodic actions: any action whose coordinate subactions all admit Borel 2-colorings is the natural candidate class in which the existence of a Borel $2^d$-coloring should be re-examined."],"forward_implications":["The Borel polychromatic number of a free ergodic $\\mathbb{Z}^d$ grid is exactly $2^d-1$, one less than the classical value.","Any Borel $2^d$-polychromatic coloring yields a Borel proper 2-coloring of each coordinate $\\mathbb{Z}$-subaction, so actions with a non-2-colorable generator, such as irrational rotations of the circle, admit no such coloring.","For $d=2$, Borel 4-polychromatic colorings exist exactly when the action admits a 1-fold invariant pair of Borel 2-colorings, giving a complete structural picture in two dimensions.","The sharpness mechanism is rigidity: a Borel $2^d$-coloring forces a type of invariance along at least one generator, which ergodic actions cannot support."],"supporting_citations":[{"why":"Supplies the original notion of face-polychromatic colorings of plane graphs that this paper generalizes to the Borel setting.","marker":"[1]"},{"why":"Provides the Borel $r$-toast decomposition of $\\mathbb{Z}^d$ grids (Lemma 2.2) on which the entire $(2^d-1)$-coloring construction rests.","marker":"[6]"},{"why":"Establishes ergodicity of the Bernoulli shift action, the canonical example for which no Borel $2^d$-polychromatic coloring exists.","marker":"[10]"},{"why":"Shows Borel chromatic numbers can differ from classical ones, the core comparison motivating the paper's sharpness question.","marker":"[11]"}],"fun_headline_variants":["Ergodic Borel grids: exactly 2^d−1 colors","Borel grid colorings drop one color from classical bound","Sharp bound: Borel grids admit 2^d−1 colors","For Borel grids, 2^d−1 colors is the optimal universal bound","Borel polychromatic grids: one color less than classical"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction of the $(2^d-1)$-coloring rests on the imported lemma that every Borel graph induced by a Borel action of $\\mathbb{Z}^d$ admits a Borel $r$-toast with nesting radius as large as $(2^{d+3}+1)d$; if such large-radius toasts failed to exist for some action, the claimed coloring would not follow from the given proof.","fun_headline_variants_meta":{"raw":{"variants":["Ergodic Borel grids: exactly 2^d−1 colors","Borel grid colorings drop one color from classical bound","Sharp bound: Borel grids admit 2^d−1 colors","For Borel grids, 2^d−1 colors is the optimal universal bound","Borel polychromatic grids: one color less than classical"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001261,"raw_usage":{"total_tokens":5158,"prompt_tokens":931,"completion_tokens":4227,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":547,"completion_tokens_details":{"reasoning_tokens":4132}},"tokens_in":547,"tokens_out":4227,"duration_ms":35131,"temperature":1.0,"reasoning_tokens":4132,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:57:35.618793+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit a free Borel action of $\\mathbb{Z}^d$ whose grid graph has no Borel $(2^d-1)$-polychromatic coloring, or exhibit a Borel $2^d$-polychromatic coloring of an ergodic action such as the Bernoulli shift; either outcome would contradict the claimed sharp value $2^d-1$.","supporting_citations":[{"cited_title":"Polychromatic colorings of plane graphs","cited_arxiv_id":null,"evidence_quote":"Supplies the original notion of face-polychromatic colorings of plane graphs that this paper generalizes to the Borel setting."},{"cited_title":"Topics in orbit equivalence","cited_arxiv_id":null,"evidence_quote":"Establishes ergodicity of the Bernoulli shift action, the canonical example for which no Borel $2^d$-polychromatic coloring exists."},{"cited_title":"Borel chromatic numbers","cited_arxiv_id":null,"evidence_quote":"Shows Borel chromatic numbers can differ from classical ones, the core comparison motivating the paper's sharpness question."}],"review_version":1}