{"id":"7e78518c-1922-4c36-912e-831e153dd908","arxiv_id":"1908.04727","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The (n-1)-dimensional Hausdorff measure of any k-antichain in [0,1]^n is at most kn, and for n=2 there exist k-antichains attaining exactly 2k.","lead":"A k-antichain in the unit n-cube is a set that meets every chain, a monotone path from one corner to the opposite, in at most k points. The paper proves such sets have (n-1)-dimensional Hausdorff measure at most kn, and constructs examples in the square that reach exactly 2k.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the main upper bound and the n=2 equality argument are coherent; only the unproved asymptotic-sharpness remark is a minor gap, not a load-bearing flaw.","rationale":"I read the full manuscript carefully, focusing on the two main theorems. Theorem 1.5 is proved by a greedy peeling of minimal elements. The only step that could be questioned is whether the set B of minimal elements is nonempty; it is, because a k-antichain cannot contain an infinite strictly descending chain, otherwise that chain would itself be a chain with infinitely many points of A. The argument that A\\B is a (k-1)-antichain is valid: if a chain C contained k points of A\\B, then finiteness of A∩C gives a minimal such point y, and since y is not minimal in A there is x∈A with x<y, forcing x into a chain with the k points and giving k+1 points of A on a chain, a contradiction. Thus Theorem 1.5 is correct conditional on Theorem 1.2, whose statement in the paper covers arbitrary antichains without measurability hypotheses. Lemma 2.1 is also internally consistent: the sequences x_n and y_n are well-defined, the rectangles lie in W(g,h) and touch its boundary, the limits are x=0 and y=1, and the telescoping sums give total length exactly 2. The gluing of the dx,n and dy,n functions produces a strictly decreasing function on (0,1) whose graph is a subset of W(g,h). Theorem 1.6 follows cleanly by applying Lemma 2.1 to k disjoint strips between decreasing bijections. I found no internal inconsistency or hidden assumption that would threaten the central claim. The only stated claim without a full proof is asymptotic sharpness of kn for k>1. The paper says it 'can be shown' by a similar argument, but does not show it. This is a genuine but minor gap; the homothetic-copy construction of k ℓ_p spheres with radii tending to 1 is straightforward and would confirm the claim. Because the gap is expository rather than a correctness risk, I do not raise a load-bearing objection. The reader's conditional verdict is reasonable; I partial-agree with the reader's weakest assumption: the reliance on Theorem 1.2 is real but not a hidden measurability problem, while the asymptotic-sharpness omission is the actual minor gap.","tokens_in":7331,"tokens_out":30976,"duration_ms":294977,"concrete_test":"Verify the claimed asymptotic sharpness for k>1: fix p large and radii r_i = 1-(k-i)ε (0<ε small) and form A_ε = ∪_{i=1}^k {x∈[0,1]^n: ||x||_p = r_i}. Check that each level set is an antichain, any chain in [0,1]^n meets each level set in at most one point (so A_ε is a k-antichain), and H^{n-1}(A_ε) = ∑ r_i^{n-1} H^{n-1}({||x||_p=1}) → kn as p→∞ and ε→0. If this computation holds, the abstract's sharpness claim is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection. Theorem 1.5's peeling argument is sound: every nonempty subset of a k-antichain has a minimal element (otherwise it would contain an infinite descending chain), so B is a nonempty antichain and A\\B is a (k-1)-antichain; the induction plus subadditivity and Theorem 1.2 yields H^{n-1}(A)≤kn. Lemma 2.1's construction of a decreasing function D with H^1(Gr(D))=2 inside any strip g≤·≤h is internally consistent; the sequence construction and the telescoping sums check out. The only gap is the abstract's assertion that kn is asymptotically sharp for k>1, which is stated without proof; a homothetic-copy construction supplies the missing argument. Since the intended construction works, this is a minor exposition issue, not a correctness risk.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies k-antichains in the unit n-cube, i.e. sets A such that every chain meets A in at most k points. Its main results are Theorem 1.5, which gives the upper bound H^{n-1}(A) ≤ kn for every k-antichain A in [0,1]^n, and Theorem 1.6, which constructs, for n=2, a k-antichain with H^1(A)=2k. The upper bound is proved by peeling off minimal elements to partition A into k antichains and then applying the n=1 theorem of Engel et al. The n=2 construction uses a lemma that inserts a strictly decreasing singular graph of Hausdorff measure 2 into any strip between two strictly decreasing bijections, and stacks k such graphs. The paper also asserts that kn is asymptotically sharp and conjectures that equality is attainable in all dimensions.","tokens_in":7487,"tokens_out":24410,"duration_ms":240924,"significance":"If the results are correct, they provide a sharp continuous analogue of Erdős's k-Sperner theorem, with the simple constant kn, exact equality in dimension 2, and asymptotic equality in all dimensions. The upper-bound proof is elegant and checkable, and the singular-function construction in Lemma 2.1 is explicit: the gluing argument is coherent and the telescoping sum correctly yields measure 2. The proofs do not rely on fitted parameters or circular definitions. The only substantive gap is the unproved asymptotic-sharpness remark, which is local and can be supplied by a short construction.","major_comments":[{"comment":"The sentence claiming that the bound kn is asymptotically sharp is stated without proof, even though the abstract advertises this as part of the paper's contribution. Please add a proof. A valid construction is A = ⋃_{j=1}^k {x∈[0,1]^n : ‖x‖_p = r_j} with 0<r_1<...<r_k≤1 chosen close to 1 and p large; each ℓ_p-sphere is an antichain, the union is a k-antichain, and H^{n-1}(A) tends to kn as p→∞ and r_j→1. This gap is local and does not affect the validity of Theorems 1.5 or 1.6.","section":"§1 (remark after Theorem 1.5)"}],"minor_comments":[{"comment":"Please justify explicitly that every nonempty subset of a k-antichain has a minimal element (otherwise it would contain an infinite descending chain); this is used both for the nonemptiness of B and for the choice of y. It would also help to spell out why the set D is a chain, namely that minimality of y in (A\\B)∩C forces all other points of that set to lie above y.","section":"§2, proof of Theorem 1.5"},{"comment":"The construction of the sequence {y_n} and the proof that its limit is 1 are omitted with 'similarly'; the details are indeed analogous, but a brief indication of the symmetric inequalities would improve readability.","section":"§2, Lemma 2.1"},{"comment":"Since Theorem 1.5 imports Theorem 1.2 from [7], it would be helpful to state explicitly that the cited result applies to arbitrary, not necessarily measurable, antichains, or to add a measurability hypothesis to the statement if one is needed.","section":"§1, Theorem 1.2"},{"comment":"In the final display, H1(Gr(Di) is missing a closing parenthesis; it should read H1(Gr(D_i)).","section":"§2, proof of Theorem 1.6"},{"comment":"Reference [3] contains a typo: 'Hausdroff' should be 'Hausdorff'.","section":"References"}],"recommendation":"minor_revision","confidential_remarks":"I found no circularity in the use of Theorem 1.2: it is an external result for the k=1 case and is not equivalent to the k-antichain theorem. The main upper bound and the n=2 equality construction are sound. The only substantive defect is the missing proof of asymptotic sharpness, which is routine and can be fixed locally; I therefore recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read on Pelekis–Vlasák, arXiv:1908.04727. The paper's real content is Lemma 2.1 and the n=2 equality construction; Theorem 1.5 is a clean but easy consequence of Engel et al.'s antichain bound. The induction peeling off minimal elements is valid, and the gluing of singular functions in Lemma 2.1 checks out: the telescoping sums give H^1=2, and the function is well-defined because the two sequences meet at 1/2 with matching values. I don't see a hidden measurability trap in importing Theorem 1.2; Hausdorff outer measure doesn't care, and the chain intersection is finite by the k-antichain condition, so the minimal-element step is legitimate.\n\nThe soft spots are real but minor. The abstract claims kn is asymptotically sharp for general k, but the proof just says 'similarly' and gives no construction. The stress-test note supplies the obvious homothetic-copy argument, so this is an exposition gap, not a correctness risk. Also, the paper does not resolve equality for n≥3, which it clearly states. The alternative direction in [16] is duly noted, and the connection to Problem 3.1 is honest.\n\nWho is this for? People working in continuous analogues of Sperner theory. The upper bound is a one-line corollary for someone who knows [7], but the equality case n=2 is genuinely new and the lemma is a useful tool. The paper is short, readable, and the citation pattern is appropriate. I'd send it to a referee; it deserves a careful look, and the only required fix is to write out the sharpness construction for k>1.","headline":"Solid short paper: the n=2 equality construction is genuinely new, the upper bound is a clean corollary, and the only real gap is the unproved asymptotic-sharpness remark.","tokens_in":7998,"tokens_out":2518,"would_cite":true,"duration_ms":22576,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05D05","28A78","05C35","26A30"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every k-antichain in the unit n-cube has (n−1)-dimensional Hausdorff measure at most kn, and the bound is attained exactly in the plane.","keywords":["k-antichains","Hausdorff measure","unit n-cube","singular function","Erdős k-Sperner theorem","antichain","continuous analogue","extremal set theory"],"falsifier":"Construct a $k$-antichain in $[0,1]^3$ and compute its 2-dimensional Hausdorff measure; if it exceeds $3k$, Theorem 1.5 is false. A simpler check for the sharpness claim: for the pair $g(x)=1-x$, $h(x)=1-x^2$, approximate the graph $D$ produced by Lemma 2.1 by polygonal arcs and sum their lengths; the lemma predicts a limit of exactly 2, so any limit different from 2 would expose an error in the equality construction.","tokens_in":7146,"feed_emoji":"📐","tokens_out":13783,"duration_ms":123584,"temperature":0.7,"pith_summary":"This paper proves a continuous analogue of Erdős's k-Sperner theorem: any $k$-antichain $A$ in the unit $n$-cube $[0,1]^n$ satisfies $H^{n-1}(A) \\le kn$. The proof peels a $k$-antichain into $k$ antichains and applies the known antichain bound $H^{n-1} \\le n$ to each layer. In the plane the bound is attained exactly: the authors build a $k$-antichain as a union of $k$ disjoint singular-function graphs, each of length 2, so $H^1(A)=2k$. The authors conjecture equality holds in every dimension and leave the case $n \\ge 3$ open.","feed_headline":"At most kn: sharp bound for k-antichains in the unit cube","feed_subtitle":"Continuous Erdős k-Sperner: (n−1)-measure of any k-antichain is ≤ kn, with equality in the plane.","key_machinery":"The proof runs on three objects. First, the antichain bound $H^{n-1}(A)\\le n$ (Theorem 1.2) supplies the base case and the per-layer estimate. Second, the peeling argument shows every $k$-antichain decomposes into $k$ antichains by repeatedly discarding minimal elements; this reduces the $k$-antichain problem to the antichain problem. Third, the singular-function graph is the tool that realizes equality: a strictly decreasing function $f$ whose derivative is zero almost everywhere has a graph of length $1+1=2$ in the unit square, and Lemma 2.1 arranges such a graph between any two bounding strictly decreasing bijections. Theorem 1.6 stacks $k$ such graphs between $2k$ ordered bijections, yielding a $k$-antichain of total length $2k$.","core_discovery":"The central claim is that for every $k$-antichain $A \\subset [0,1]^n$, $H^{n-1}(A) \\le kn$, and that this bound is asymptotically sharp; for $n=2$, equality is achieved. The upper bound follows by induction: taking $B$ to be the minimal elements of $A$, $B$ is an antichain and $A\\setminus B$ is a $(k-1)$-antichain, so $A$ is a union of $k$ antichains; Theorem 1.2 gives each antichain measure at most $n$. The equality construction for $n=2$ pairs $2k$ strictly decreasing continuous bijections $f_1>f_2>\\cdots>f_{2k}$ and, using singular functions, inserts between each pair $f_{2i-1}, f_{2i}$ a strictly decreasing graph $D_i$ with $H^1(\\mathrm{Gr}(D_i))=2$; the union $A=\\cup_i \\mathrm{Gr}(D_i)$ is a $k$-antichain with $H^1(A)=2k$.","pith_inferences":["The linear dependence on $k$ in the continuous bound $kn$ stands in contrast to the discrete Erdős theorem, where the extremal size is a sum of $k$ middle binomial coefficients; the continuous analogue loses the binomial profile and treats each layer identically.","The equality construction for $n=2$ is made of singular-function graphs, so if the conjecture for $n\\ge3$ is true, extremal $k$-antichains are likely to be singular or fractal hypersurfaces rather than smooth boundaries; this points toward constructing analogues of the singular graph in higher dimensions.","The paper states asymptotic sharpness in all dimensions without a full proof, referring only to an argument similar to the one after Theorem 1.2; a complete proof would need to exhibit approximating $k$-antichains explicitly and check that no chain meets them more than $k$ times.","If Theorem 1.2 carries hidden regularity conditions on antichains, the peeling argument would inherit them; the authors state no such conditions, so the upper bound is only as unconditional as the antichain theorem it imports."],"forward_implications":["Every $k$-antichain in $[0,1]^n$ has Hausdorff dimension at most $n-1$, since its $(n-1)$-dimensional Hausdorff measure is finite and bounded by $kn$.","The $kn$ bound is asymptotically sharp in every dimension: there are $k$-antichains whose $(n-1)$-dimensional Hausdorff measure is arbitrarily close to $kn$.","In the plane the bound is attained exactly: there exists a $k$-antichain $A\\subset[0,1]^2$ with $H^1(A)=2k$.","The proof reduces the $k$-antichain problem to the antichain case, so any improvement or sharpening of the antichain bound would transfer directly to $k$-antichains.","The conjecture that equality $H^{n-1}(A)=kn$ is attainable for all $n\\ge3$ remains open; the paper settles only $n=2$."],"supporting_citations":[{"why":"Supplies Theorem 1.2, the antichain bound $H^{n-1}(A)\\le n$ that is the base case and the engine of the peeling proof of Theorem 1.5.","marker":"[7]"},{"why":"States the discrete k-Sperner theorem (Erdős) whose continuous analogue the paper formulates and proves; sets the central comparison.","marker":"[8]"},{"why":"Gives the fact that a singular function's graph has length equal to the sum of its coordinate variations, used in Lemma 2.1 to build graphs of Hausdorff length 2.","marker":"[10]"},{"why":"Documents the existence of singular functions, which Theorem 1.4 and Lemma 2.1 require to realize decreasing graphs of full length.","marker":"[18]"},{"why":"Provides the continuity of Hausdorff measure of convex-body boundaries under Hausdorff convergence, used to argue the $kn$ bound is asymptotically sharp.","marker":"[19]"}],"fun_headline_variants":["Sharp bound: (n-1)-measure of k-antichains ≤ kn in cube","K-antichains in unit n-cube: measure ≤ kn, sharp","Hausdorff measure of k-antichains ≤ kn, asymptotically sharp","Unit cube: k-antichains' (n-1)-measure ≤ kn, plane exact","K-antichains in cube: (n-1)-measure ≤ kn, sharp for n=2"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The upper bound assumes the known antichain theorem $H^{n-1}(A)\\le n$ applies to every antichain in the unit cube, including the irregular antichains obtained by repeatedly peeling minimal elements from an arbitrary $k$-antichain; the paper does not state or verify any measurability or regularity hypothesis that this application might require.","fun_headline_variants_meta":{"raw":{"variants":["Sharp bound: (n-1)-measure of k-antichains ≤ kn in cube","K-antichains in unit n-cube: measure ≤ kn, sharp","Hausdorff measure of k-antichains ≤ kn, asymptotically sharp","Unit cube: k-antichains' (n-1)-measure ≤ kn, plane exact","K-antichains in cube: (n-1)-measure ≤ kn, sharp for n=2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000874,"raw_usage":{"total_tokens":3842,"prompt_tokens":1065,"completion_tokens":2777,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":681,"completion_tokens_details":{"reasoning_tokens":2662}},"tokens_in":681,"tokens_out":2777,"duration_ms":19976,"temperature":1.0,"reasoning_tokens":2662,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:35:27.852354+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a $k$-antichain in $[0,1]^3$ and compute its 2-dimensional Hausdorff measure; if it exceeds $3k$, Theorem 1.5 is false. A simpler check for the sharpness claim: for the pair $g(x)=1-x$, $h(x)=1-x^2$, approximate the graph $D$ produced by Lemma 2.1 by polygonal arcs and sum their lengths; the lemma predicts a limit of exactly 2, so any limit different from 2 would expose an error in the equality construction.","supporting_citations":[{"cited_title":"Projection inequalities for antichains","cited_arxiv_id":"1812.06496","evidence_quote":"Supplies Theorem 1.2, the antichain bound $H^{n-1}(A)\\le n$ that is the base case and the engine of the peeling proof of Theorem 1.5."},{"cited_title":"Erd˝ os, On a lemma of Littlewood and Offord, Bull","cited_arxiv_id":null,"evidence_quote":"States the discrete k-Sperner theorem (Erdős) whose continuous analogue the paper formulates and proves; sets the central comparison."},{"cited_title":"Foran, The Length of the Graph of a One to One Function f rom [0, 1] to [0, 1], Real Anal","cited_arxiv_id":null,"evidence_quote":"Gives the fact that a singular function's graph has length equal to the sum of its coordinate variations, used in Lemma 2.1 to build graphs of Hausdorff length 2."},{"cited_title":"Saks, Theory of the Integral , Dover Publications, 1964","cited_arxiv_id":null,"evidence_quote":"Documents the existence of singular functions, which Theorem 1.4 and Lemma 2.1 require to realize decreasing graphs of full length."},{"cited_title":"Schneider, Convex bodies: the Brunn-Minkowski theory , Second expanded edition, Encyclopedia of Mathematics and its Applications, vol","cited_arxiv_id":null,"evidence_quote":"Provides the continuity of Hausdorff measure of convex-body boundaries under Hausdorff convergence, used to argue the $kn$ bound is asymptotically sharp."}],"review_version":1}