{"id":"596d93ff-e3da-487a-82d7-e63f7d2093a7","arxiv_id":"1908.04241","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For n hard disks in a strip of width w, the j-th Betti number grows polynomially like n^{2j} when w >= j+2 and exponentially like (q+1)^n n^{qw+2r} when 2 <= w <= j+1.","lead":"Mathematicians determined exactly how the topology of a strip filled with hard disks changes as the strip widens. They show when the homology matches point-particle configuration spaces, and when it grows exponentially with the number of disks.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.5's vector-field construction only addresses short measurements between consecutive blocks; the proof that τ increases at the claimed rate—and hence the deformation retraction Theorem 3.3—is not complete, and the central upper and lower bound machinery depends on it.","rationale":"The reader identified the same weakest assumption: the sketched vector-field construction in Lemma 3.5. I agree that this is the most load-bearing concern, because Theorem 3.1 and everything built on it depend on the deformation retraction of Theorem 3.3. The proof as written does not handle short measurements between non-consecutive blocks, and it omits continuity and boundary-exit checks for the flow. These are fillable by an expert, but they are not merely cosmetic: without a complete proof of Lemma 3.5, the homotopy equivalence is unsupported. I also found an independent concrete error in the lower-bound count in Section 4: the displayed formula overcounts special symbols by a factor of q!. For n=6, w=3, j=4 the formula gives 160, while the paper's own Table 1 lists β4=80. This does not change the asymptotic Ω((q+1)^n n^{qw+2r}), since the extra factor is constant in n, but it should be corrected. Because the main asymptotic claim is very likely true and the gaps appear repairable, I would not reject; I would request a rigorous proof of Lemma 3.5 (plus the small counting correction) before final acceptance.","tokens_in":26547,"tokens_out":39228,"duration_ms":413874,"concrete_test":"Independently re-derive Lemma 3.5 and check, for every symbol α and every p in Uα with τ(p)<1/w(α), that the prescribed vector field vα(p) satisfies Dτ_p(vα(p)) ≥ 2√(1/w(α)−τ(p)). A concrete computational check: for n=3,4 and w=2,3, sample Uα on a fine grid, construct vα exactly as in §3.2, and verify the inequality at every short measurement, including those between non-consecutive blocks; also integrate the ODE to confirm the flow reaches C_{1/w}(n,1) in time ≤√(1/w) without leaving U(n,w). If the inequality fails on any sample, Lemma 3.5 as stated is false and Theorem 3.3 requires repair.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem 3.1 reduces C(n,w) to cell(n,w) via Theorem 3.3, which rests on Lemma 3.5. In the proof of Lemma 3.5, the construction of x(p) chooses horizontal separations for consecutive blocks only, based on short measurements between those blocks. However, τ(p) is the minimum over all point-pair and boundary distances, and a configuration can have its active short measurement realized by two points in non-consecutive blocks; the proof does not show such a measurement is increased at the required rate by the resulting vector field. The text also does not prove that λ(p) and the chosen separations are continuous, nor that the resulting flow stays in U(n,w) until it reaches C_{1/w}(n,1). If any of these steps fails, the deformation retraction of Theorem 3.3 is not established, so the homotopy equivalence Theorem 3.1 and the subsequent upper bounds via discrete Morse theory and the gas-regime isomorphism lose their foundation. This is a gap in proof completeness rather than a demonstrated counterexample, but it is exactly the load-bearing point of the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the configuration space C(n,w) of n unit-diameter hard disks in an infinite strip of width w. It proves an asymptotic description of the Betti numbers beta_j[C(n,w)] for fixed j and w as n tends to infinity: in the 'gas' regime w >= j+2 the Betti numbers are those of the plane configuration space and grow polynomially, in the 'liquid' regime j >= w-1 they grow exponentially with an explicit base and polynomial correction, and in the 'solid' regime the homology is trivial. The paper also proves a finite-n phase portrait. The proof strategy is to identify C(n,w) with a subcomplex of the Salvetti complex via a nerve cover and a deformation retraction, obtain lower bounds by explicit intersecting torus cycles, and obtain upper bounds by a discrete Morse matching; an appendix computes Betti numbers for n <= 8.","tokens_in":1485,"tokens_out":1722,"duration_ms":96164,"significance":"If the main theorems are correct, this is a significant contribution to the topology of configuration spaces of hard disks, giving the first systematic asymptotic picture and a finite-n phase portrait that are both conceptually clean and quantitatively explicit. The methods are largely self-contained, the arguments contain no fitted parameters, and the computational appendix with PHAT provides useful independent evidence for small n. The strengths are real: the nerve-theoretic comparison to the Salvetti complex, the explicit torus cycles, and the skyline-counting upper bounds are elegant ideas. However, several load-bearing steps in the written proofs are sketched rather than completed, especially the construction of the deformation retraction in Section 3.2 and the transversality check in Section 4. These gaps do not appear to point to a counterexample, but they need to be filled before the central claims are fully established.","major_comments":[{"comment":"The construction of x(p) is described only for pairs of consecutive blocks, and the proof verifies the required increase rate only for short measurements realized by points in those consecutive blocks. A short measurement can be realized by points in two non-consecutive blocks, and the increase in their horizontal separation is only an indirect sum of the chosen increases in the intervening consecutive gaps. The text does not prove that this sum achieves the required rate 2*sqrt(1/w(alpha) - m(p)) for every active short measurement, especially when some intermediate consecutive gap is left unchanged. Since Lemma 3.5 is the core of Theorem 3.3, this is a load-bearing gap.","section":"Section 3.2, Lemma 3.5"},{"comment":"The proof defines v(p) as a partition-of-unity combination of the local vector fields v_alpha(p) and then uses an ODE argument to conclude that the flow reaches C_{1/w}(n,1) in time at most 1. This requires continuity of lambda(p), x(p), and hence v_alpha(p); lambda(p) is defined as a minimum over a set of short measurements that changes as p varies, and the text does not prove that this minimum varies continuously. The assertion that the directional derivative of tau is always well-defined is also stated without proof. Without these facts, the existence of the flow and the claim that it stays inside U(n,w) until reaching the target are not fully justified.","section":"Section 3.2, proof that Lemma 3.5 implies Theorem 3.3"},{"comment":"The lower-bound argument relies on Lemma 4.1, whose condition (4) requires that Z_alpha intersects Z*_alpha transversely in a point. The proof of Theorem 4.4 asserts the single-point intersection and then says 'We omit the details' for the tangent-space direct-sum check. This transversality check is essential for linear independence of the cycles and therefore for the lower-bound half of Theorem 1.1(2); it should either be supplied or replaced by a precise reference.","section":"Section 4, Theorem 4.4"},{"comment":"The algorithm producing A_p imposes the condition y_{sigma_l} > y_{sigma_l'} + 1, which is incompatible with the ambient condition 0 < y_k < 1 and with Definition 2.3, where the same-block condition is simply y_{sigma_l} > y_{sigma_l'}. The same impossible '+1' appears later in the chain-to-point construction. This is probably a typographical error, but because the algorithm is load-bearing for the identification of the nerve with the barycentric subdivision of cell(n,w), the correct strict inequality should be stated and the surrounding argument checked.","section":"Section 3.2, Theorem 3.4, Step 2"}],"minor_comments":[{"comment":"Several inclusions are printed with a missing arrow, for example 'C(n,w) contained in C(n,w+1)' in Theorem 3.1 and 'C(n,w) contained in C(n,R^2)' in the proof of Theorem 1.1(1); these should be typeset with the inclusion arrow.","section":"Throughout, Theorem 3.1 and Section 3.3"},{"comment":"The lower-bound formula contains malformed brace markup in the multinomial coefficient and is hard to read; please typeset the multinomial coefficient properly.","section":"Section 4, proof of Theorem 1.1(2)"},{"comment":"The first sentence of the proof says '0 <= j <= n - ceil(n/w)', while the theorem statement requires 'w-1 <= j <= n - ceil(n/w)'; the proof later uses j >= w-1, so the displayed range should be corrected.","section":"Section 5, proof of Theorem 1.2(2)"},{"comment":"Reference [12] appears to cite Davis's book as a chapter in 'Introduction to modern mathematics' with volume 33; please verify the bibliographic data.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The reader's report recommends accept with high confidence. I agree that the main results are likely correct and are supported by the computational appendix, but my own reading identifies two places where the written proof is incomplete at exactly the load-bearing points: the deformation retraction of Theorem 3.3 depends on a vector-field construction whose rate estimate is only shown for consecutive blocks, and the transversality check in Theorem 4.4 is explicitly omitted. These are not presentation issues and should be fixed before publication; hence I recommend major revision rather than accept. I see no concerns about novelty or citation practice."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper delivers something genuinely new: for n hard disks of unit diameter in an infinite strip of width w, it pins down the asymptotic growth of each Betti number beta_j[C(n,w)] as n grows, with polynomial growth in the gas regime (w >= j+2), exponential growth of the specific form (q+1)^n n^{qw+2r} in the liquid regime (w <= j+1), and trivial homology in the solid regime. The phase portrait for finite n is a nice bonus. The proof strategy is also new in this setting: a subcomplex of the Salvetti complex, a nerve-theoretic homotopy equivalence, explicit torus cycles for lower bounds, and a discrete Morse matching with skyline counting for upper bounds. No parameters are fitted; the comparison with Arnold's computation of C(n,R^2) is independent. The small-n Betti table in the appendix gives independent computational support. This is a serious paper with a real result.\n\nThe main soft spot is exactly where the stress-test note points: Lemma 3.5 constructs a vector field v_alpha on each open set U_alpha, and the proof only explicitly shows how to increase short measurements between consecutive blocks. But tau(p) is the minimum over all point-pair distances, including pairs in non-consecutive blocks. The text asserts that choosing the horizontal gaps between consecutive blocks handles all short measurements, but it does not prove that a short measurement between, say, blocks 1 and 3 is increased at the required rate by the sum of the chosen gaps. The construction of x(p) also does not address continuity of lambda(p) and the chosen separations, nor that the flow stays inside U(n,w) until it reaches C_{1/w}(n,1). This is not a demonstrated counterexample; the construction is plausible and probably repairable by choosing the gaps to satisfy all short measurements simultaneously and then checking continuity. But as written, the deformation retraction of Theorem 3.3 — and therefore the homotopy equivalence C(n,w) ≃ cell(n,w) and everything built on it — is not fully proved. A second, smaller gap: Theorem 4.4 explicitly says \"we omit the details\" for the transversality check. That is a standard sort of verification and unlikely to fail, but it is still an omitted detail in a proof of a lower bound.\n\nWho should read this: anyone working on configuration spaces of disks, hard-sphere phase transitions, or discrete Morse theory on Salvetti complexes. The main theorem is the kind of clean asymptotic statement that deserves to be in the literature, and the techniques will likely be reused. My recommendation: send it to peer review, and in the report ask the authors to fill the gap in Lemma 3.5 and to provide the transversality details. The result is probably right, but the proof needs another pass.","headline":"New asymptotic growth rates for Betti numbers of disk configuration spaces in a strip, with a real but likely repairable gap in the deformation-retraction lemma.","tokens_in":27263,"tokens_out":3539,"would_cite":true,"duration_ms":41761,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55R80","82B26"],"pacs":[],"model":"deepseek-v4-flash","headline":"Hard-disk configuration spaces in a strip have Betti numbers that grow exponentially in the liquid regime, at rate $(q+1)^n n^{qw+2r}$.","keywords":["configuration spaces","hard disks","Betti numbers","infinite strip","Salvetti complex","discrete Morse theory","homological phase portrait","phase transitions"],"falsifier":"A direct persistent-homology computation of $\\beta_1[C(9,3)]$ must give 36, since the gas-regime isomorphism identifies it with $\\beta_1[C(9,\\mathbb{R}^2)]$; a simultaneous computation of $\\beta_6[C(9,4)]$ must give a nonzero value, as the liquid regime requires. A violation at either value would pinpoint a failure in the homotopy-equivalence chain behind the asymptotic claims.","tokens_in":26360,"feed_emoji":"⚪","tokens_out":19590,"duration_ms":194679,"temperature":0.7,"pith_summary":"This paper studies the configuration space $C(n,w)$ of $n$ non-overlapping unit-diameter disks in an infinite strip of width $w$, as $n$ grows. It establishes that each homology group $H_j[C(n,w)]$ falls into one of three regimes: a gas regime where the homology agrees with the configuration space of labeled points in the plane, a liquid regime where the Betti numbers grow exponentially in $n$, and a solid regime where the homology is zero. The central quantitative result is that for $w \\ge 2$ and $j \\ge w-1$, writing $j = q(w-1) + r$ with $0 \\le r < w-1$, one has $\\beta_j[C(n,w)] \\asymp (q+1)^n n^{qw+2r}$. This gives a sharp asymptotic description of hard-disk configuration-space Betti numbers in the liquid regime and a complete homological phase portrait for every finite $n$.","feed_headline":"Betti numbers of hard disks in a strip go exponential","feed_subtitle":"Gas-regime homology matches point particles; liquid-regime growth is $(q+1)^n n^{qw+2r}$.","key_machinery":"The central objects are $\\mathrm{cell}(n,w)$, a subcomplex of the Salvetti complex whose cells are permutations with bars and with no block of more than $w$ entries, and a good cover $U_\\alpha$ used to prove the homotopy equivalence $C(n,w) \\simeq \\mathrm{cell}(n,w)$ through the nerve theorem. The deformation retraction from $U(n,w)$ to the disk configuration space is driven by a vector field that raises the tautological function $\\tau$ at a controlled rate. For lower bounds, the paper constructs embedded tori $Z_\\alpha$ labeled by special symbols and closed dual submanifolds $Z^*_\\alpha$, then uses the intersection pairing between homology and homology with closed supports to show these cycles are linearly independent, yielding the exponential lower bound. For upper bounds, a discrete gradient vector field collapses $\\mathrm{cell}(n,w)$ to a complex whose critical cells are coded by skylines; counting those cells gives the matching exponential upper bound.","core_discovery":"The paper's central claim is that the homology of hard-disk configuration spaces in a strip changes character at a sharp threshold: once $j \\ge w-1$, the Betti number $\\beta_j[C(n,w)]$ stops being polynomial and grows exponentially in $n$, at the rate $(q+1)^n n^{qw+2r}$, where $j = q(w-1) + r$ and $0 \\le r < w-1$. In the complementary range $0 \\le j \\le w-2$, the inclusion of $C(n,w)$ into the point configuration space $C(n,\\mathbb{R}^2)$ induces an isomorphism on $H_j$, so $\\beta_j[C(n,w)]$ is the unsigned Stirling number $\\left[{n \\atop n-j}\\right]$ and grows as a polynomial in $n$ of degree $2j$. For $w=0$, or $w=1$ with $j \\ge 1$, the homology vanishes; for $w=1$ and $j=0$, $\\beta_0 = n!$. The paper also proves a complete finite-$n$ phase portrait in the $(w,j)$-plane, dividing degrees into homological solid, liquid, and gas regimes.","pith_inferences":["The formula has a statistical-mechanics flavor: the factor $(q+1)^n$ counts placements of the leftover disks in the gaps between maximal vertical blocks, so the exponential growth rate may be interpreted as a free-energy-like quantity for which a generating function could exist.","The numerical coincidence the paper observes between $H_1(C(n,2))$ and the first homology of the $(n+1)$-point k-equal manifold suggests there may be a natural map between those spaces; finding it would connect disk configuration spaces to the well-developed homology theory of k-equal arrangements.","The skyline coding that makes the upper bound efficient may extend to related hard-particle configuration spaces, such as mixtures of disk sizes or disks in higher-dimensional slabs, giving analogous exponential Betti bounds."],"forward_implications":["In the gas regime, hard disks are indistinguishable from point particles for homology in degrees $j \\le w-2$: the inclusion induces $H_j[C(n,w)] \\cong H_j[C(n,\\mathbb{R}^2)]$ for every $n$.","In the liquid regime, the exponential base is $q+1$, so for a fixed strip width, larger homology degree comes with a faster exponential growth rate, and the polynomial correction is $n^{qw+2r}$.","For every finite $n$, the phase portrait gives exact boundaries: $H_j[C(n,w)] = 0$ for $j \\ge n - \\lceil n/w \\rceil + 1$, and $H_j[C(n,w)] \\ne 0$ but not isomorphic to point-particle homology for $w-1 \\le j \\le n - \\lceil n/w \\rceil$.","The space $C(n,2)$ is aspherical: it admits a locally CAT(0) cube complex structure, so its higher homotopy groups vanish."],"supporting_citations":[{"why":"Computes the Poincaré polynomial of the point configuration space $C(n,\\mathbb{R}^2)$; supplies the gas-regime baseline that $H_j[C(n,w)]$ is compared with.","marker":"[2]"},{"why":"Defines the tautological function $\\tau$, whose controlled growth under a vector field is the mechanism for the deformation retraction $U(n,w) \\simeq C_{1/w}(n,1)$.","marker":"[3]"},{"why":"Provides the extended nerve theorem used to make the homotopy equivalences commute with inclusions into wider strips, needed for the inclusion-induced isomorphism in the gas regime.","marker":"[4]"},{"why":"Introduces the Salvetti complex and its homotopy equivalence to $C(n,\\mathbb{R}^2)$; the paper's cell complex $\\mathrm{cell}(n,w)$ is the subcomplex restricting block width.","marker":"[24]"},{"why":"Supplies discrete Morse theory, which bounds Betti numbers by the number of critical cells and is the backbone of the upper bound.","marker":"[16]"}],"fun_headline_variants":["Disk-strip Betti numbers turn exponential when j ≥ w−1","Strip hard-disk homology: sharp threshold to exponential growth","Hard disks in strip: homology jumps to exponential at j = w−1","Disk configuration spaces in strip: exponential Betti numbers","Homology of hard disks in a strip: polynomial vs exponential"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on a geometric flow that moves any configuration with no vertical stack of $w+1$ points so that its minimal disk diameter grows at a guaranteed speed; if that flow cannot be constructed as claimed, the homotopy equivalence and the bounds built on it would need to be reworked.","fun_headline_variants_meta":{"raw":{"variants":["Disk-strip Betti numbers turn exponential when j ≥ w−1","Strip hard-disk homology: sharp threshold to exponential growth","Hard disks in strip: homology jumps to exponential at j = w−1","Disk configuration spaces in strip: exponential Betti numbers","Homology of hard disks in a strip: polynomial vs exponential"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000239,"raw_usage":{"total_tokens":1585,"prompt_tokens":1087,"completion_tokens":498,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":703,"completion_tokens_details":{"reasoning_tokens":410}},"tokens_in":703,"tokens_out":498,"duration_ms":5870,"temperature":1.0,"reasoning_tokens":410,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:46:17.231025+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct persistent-homology computation of $\\beta_1[C(9,3)]$ must give 36, since the gas-regime isomorphism identifies it with $\\beta_1[C(9,\\mathbb{R}^2)]$; a simultaneous computation of $\\beta_6[C(9,4)]$ must give a nonzero value, as the liquid regime requires. A violation at either value would pinpoint a failure in the homotopy-equivalence chain behind the asymptotic claims.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Computes the Poincaré polynomial of the point configuration space $C(n,\\mathbb{R}^2)$; supplies the gas-regime baseline that $H_j[C(n,w)]$ is compared with."},{"cited_title":"Mi n-type Morse theory for configu- ration spaces of hard spheres","cited_arxiv_id":null,"evidence_quote":"Defines the tautological function $\\tau$, whose controlled growth under a vector field is the mechanism for the deformation retraction $U(n,w) \\simeq C_{1/w}(n,1)$."},{"cited_title":"Salvetti","cited_arxiv_id":null,"evidence_quote":"Introduces the Salvetti complex and its homotopy equivalence to $C(n,\\mathbb{R}^2)$; the paper's cell complex $\\mathrm{cell}(n,w)$ is the subcomplex restricting block width."},{"cited_title":"A user’s guide to discrete Morse theory","cited_arxiv_id":null,"evidence_quote":"Supplies discrete Morse theory, which bounds Betti numbers by the number of critical cells and is the backbone of the upper bound."}],"review_version":1}