{"id":"eb4b3caf-a273-4c7c-be52-97394c0043da","arxiv_id":"2605.25405","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Flag positroid pipe dreams are new diagrams in bijection with Bruhat intervals whose elbow counts are Richardson-cell dimensions and whose row rule characterizes nonnegatively representable elementary positroid quotients.","lead":"This paper introduces flag positroid pipe dreams, a new tiling model for complete flags in nonnegative flag varieties. The diagrams encode every constituent positroid of a flag rank by rank and give a combinatorial classification of consecutive-rank quotients.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Final Richardson-cell identification (Thm 4.43) depends on [9, Lem. 7.7] to recover the full flag positroid from lex-min/max bases u[k], v[k]; that uniqueness is not established and the paper never checks B(P_k(D)) = {w[k]: w∈[u,v]}.","rationale":"The reader's weakest assumption—that the external characterization from [9, Lemma 7.7] carries the final Richardson-cell identification—is exactly the point on which the central claim hinges. The paper's own Example 4.44 shows that the extreme bases do not determine all bases of a constituent in general, so the identification step is not a formality. Although the paper cites credible prior work, the specific implication needed here—that any flag positroid with extreme bases u[k], v[k] equals the Richardson-cell flag positroid—is not independently verified, and the internal standardization theorem (Thm 4.34) has an explicitly omitted reverse case analysis that could undermine the computation of P_k(D). These are correctness risks, not demonstrated errors, so the appropriate verdict remains CONDITIONAL: the architecture is convincing and likely correct, but the missing verification should be supplied before full acceptance. I agree with the reader's identification of the weakest assumption and would not change the verdict, though I would prioritize the concrete equality check above as the decisive test.","tokens_in":42942,"tokens_out":9995,"duration_ms":102271,"concrete_test":"For all n ≤ 6 and all u ≤ v in S_n, compute D = FPP(u,v) via Definition 3.9, then for each k compute B(P_k(D)) from the acyclic directed graph G(D|_k) (Definitions 4.19–4.20). Independently enumerate the Bruhat interval [u,v] and form {w[k] : w ∈ [u,v]}. Check equality B(P_k(D)) = {w[k] : w ∈ [u,v]} for every k. If equality holds in all cases, the encoding is correct and the external lemma is being applied validly; if any counterexample appears, Theorem 4.43 is false and the central Richardson-cell identification fails.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem 4.43 claims that for D=FPP(u,v), the sequence P•(D) is exactly the flag positroid of the positive Richardson cell R_{u,v}^{>0}. The proof only establishes that the lexicographically minimal and maximal bases of each P_k(D) are u[k] and v[k]; then it invokes [9, Lemma 7.7] to conclude that P•(D) is the full flag positroid of R_{u,v}^{>0}. This is the critical load-bearing step: the paper never proves that the full set of bases of P_k(D) equals {w[k] : w ∈ [u,v]}, nor that a nonnegatively representable flag positroid is uniquely determined by the sequence of extreme bases (u[k], v[k]). The latter is not obviously true; in the paper's own Example 4.44, the Gale interval between u[3] and v[3] contains four 3-subsets while P_3 has only two bases, so the flag positroid data is strictly richer than the extreme bases. If [9, Lemma 7.7] is being applied outside its hypotheses, or if two distinct flag positroids share the same sequences u[k], v[k], then Theorem 4.43 and consequently the bijection between maximal chains of (Π_n,⊴_q) and FPP(n) could encode a different flag positroid. A second, supporting gap is that Theorem 4.34 (standardization preserves bases) has only a sketched reverse inclusion; a hidden error there would corrupt the very bases P_k(D) that are fed into the external identification.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces flag positroid pipe dreams (FPPs), a class of reduced Rothe pipe dreams associated to pairs of permutations (u,v). Theorem 3.16 proves that FPP(u,v) exists exactly when u≤v in Bruhat order and that the number of elbows is the length of the interval [u,v]. The authors then develop a rank-by-rank theory: partial FPPs encode rank-k positroids; appending a row in unblocked columns characterizes nonnegatively representable elementary positroid quotients (Theorem 4.26); a standardization operation converts any partial FPP into the L-diagram of the same positroid (Theorem 4.34); and Theorem 4.43 claims that FPP(u,v) encodes the complete flag positroid associated to the positive Richardson cell R_{u,v}^{>0}. Section 5 translates these results to decorated permutations, giving an explicit description of the cyclic-shift freeze sets and proving self-duality of the poset (Π_n,⊴_q).","tokens_in":43347,"tokens_out":10850,"duration_ms":115036,"significance":"The FPP model is a natural and potentially very useful extension of Postnikov's L-diagrams to the flag positroid setting. The paper is constructive: it provides an explicit algorithm for FPP(u,v), an explicit bijection between nonnegatively representable elementary quotients and subsets of unblocked columns, and an explicit formula for the cyclic-shift sets in Theorem 5.12. The final bijection between maximal chains of (Π_n,⊴_q) and FPP(n) gives a clean combinatorial encoding of positive Richardson cells. If the central identification with flag positroids is fully justified, this will be a valuable contribution to the combinatorics of total positivity.","major_comments":[{"comment":"The proof establishes only that the lexicographically minimal and maximal bases of P_k(D) are u[k] and v[k], and then invokes [9, Lem. 7.7] to identify the full flag positroid. The lemma is not stated, so this logical step cannot be checked. If the lemma gives the basis set as {w[k] : w∈[u,v]}, then extreme bases are not sufficient: in Example 4.44, D=FPP(2413,4231), P_3 has bases {124,234}, while the Gale interval between these endpoints also contains 134. To make Thm 4.43 load-bearing, the authors should either state and verify a uniqueness lemma from [9] that applies to arbitrary complete flag positroids, or prove directly that B(P_k(D)) = {w[k] : w∈[u,v]} for each k.","section":"§4.5, proof of Thm 4.43"},{"comment":"The forward inclusion B(D)⊆B(D') is proved by a detailed path surgery, but the reverse inclusion is dismissed with 'a careful case analysis similar to the one above'. Theorem 4.34 is load-bearing: it underlies Definition 4.35 (standardization), Definition 4.41 (the constituents P_k(D)), and Theorem 5.12 (the decorated-permutation characterization). If the reverse direction fails, these constructions encode the wrong bases. The authors should provide a full proof of the reverse inclusion, or replace the argument by an invariant of the directed graph under the local moves of Lemma 4.32.","section":"§4.4, proof of Thm 4.34"},{"comment":"The sentence 'each nonnegatively representable quotient P_k⊴_q P_{k+1} forms an oriented matroid quotient, so by Proposition 7.10 of Boretsky–Eur–Williams, P•(D) is a complete flag positroid' is terse. The authors should spell out which hypotheses of [9, Prop. 7.10] are being verified and why the row-by-row construction gives exactly those hypotheses. This is a smaller gap, but it is part of the chain leading to Theorem 4.43.","section":"§4.5, Definition 4.41"}],"minor_comments":[{"comment":"The text says the two non-LPM positroids on [3] correspond to '321 and 321', but as printed the two decorated permutations look identical. Presumably one should have an overline and one an underline; please fix the notation.","section":"§5.3, Figure 21"},{"comment":"There are typos: 'Riestch' should be 'Rietsch', and 'Bruhar order' should be 'Bruhat order'.","section":"§2.2 and §4.5"},{"comment":"The phrase 'all cross tiles are justified to the southeast direction' is informal. The precise characterization appears later in Proposition 3.23; consider defining the condition directly at Definition 3.4, or moving the definition of Γ-free earlier.","section":"§3.1, Definition 3.4"},{"comment":"The colouring verification in the final paragraph is compressed. Since the decorated permutation includes the colouring data, and the theorem claims equality of decorated permutations, the fixed-point colouring cases should be written out in full rather than condensed into a single paragraph.","section":"§5.2, proof of Thm 5.12"}],"recommendation":"major_revision","confidential_remarks":"The paper is a serious and original contribution, and the main combinatorial constructions are likely correct. The stumbling block is the proof of Theorem 4.43: the appeal to [9, Lem. 7.7] must be made explicit, and the authors need to close the gap between 'extreme bases match' and 'full flag positroid matches'. I also recommend asking for a complete proof of the reverse inclusion in Theorem 4.34. Both issues are fixable within the scope of the manuscript, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The paper is mostly what it claims: a single class of pipe dreams that encodes complete flag positroids, with a row-by-row construction, a standardization operation recovering L-diagrams of constituents, and explicit cyclic-shift sets answering part of Chen et al.'s question. The bijection between FPPs and Bruhat intervals, and the elbow count as cell dimension, are cleanly argued. I believe the main framework is new and useful, and the Le-diagram special case is a nice sanity check rather than a limitation.\n\nThe two soft spots the reader flagged are real but not fatal. Theorem 4.34's reverse inclusion is only sketched ('careful case analysis similar to the one above'). The forward direction has an explicit path-switching algorithm, and the reverse should follow the same line; I'd want the authors to write it out, but this is a standard kind of repair, not a sign the theorem is wrong.\n\nThe bigger one is Theorem 4.43. The proof does not directly show that the bases of P_k(D) are exactly {w[k] : w in [u,v]}. It shows P•(D) is a flag positroid, shows its k-th constituent has lexicographically minimal and maximal basis u[k] and v[k], and then invokes [9, Lemma 7.7] to conclude. The stress-test concern—that the Gale interval between u[k] and v[k] can contain more than the actual basis set—is accurate, but that is precisely what the lemma is meant to handle. The paper should state the lemma explicitly and verify its hypotheses; if the lemma indeed says a flag positroid is uniquely determined by its sequence of extreme bases, then the argument goes through. If not, Theorem 4.43 is underproved. As written, the dependence is legitimate but opaque, and a referee should check it.\n\nThe cyclic-shift section is solid and gives an explicit T(C) construction, which is genuinely more than a restatement of Chen et al. The self-duality result is a nice dividend.\n\nI'd send this to referees. The main construction is significant enough, and the gaps are repairable. For a reader interested in positroid/flag combinatorics, this is worth the time.","headline":"Genuinely new combinatorial model for flag positroids with a clean main construction; two proof gaps—one sketched case analysis, one imported uniqueness lemma—need attention, but nothing here looks broken.","tokens_in":43808,"tokens_out":3952,"would_cite":true,"duration_ms":45797,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05B35","14M15","05A05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Flag positroid pipe dreams encode a complete nonnegative flag rank by rank: each Bruhat interval u ≤ v gets exactly one diagram, whose k-row restriction is the k-th constituent positroid.","keywords":["flag positroid pipe dreams","flag positroids","positroid quotients","Bruhat order","Richardson cells","decorated permutations","nonnegative flag variety","L-diagrams"],"falsifier":"Take a Bruhat interval [u,v] in S_4 or S_5 that is not attached to a Grassmannian permutation (a permutation with at most one descent), build FPP(u,v) with the paper's algorithm, restrict to the first k rows, standardize, and compute the bases of P_k as sink sets of non-intersecting paths; then brute-force enumerate every permutation w in [u,v] and restrict each to its first k values. If the two sets differ for any k — or if the lexicographically maximal basis is not v[k] — Theorem 4.43 is false. A cheaper probe checks the unproved direction of Theorem 4.34: on small partial FPPs with adjacent","tokens_in":42830,"feed_emoji":"🧩","tokens_out":22350,"duration_ms":197708,"temperature":0.7,"pith_summary":"The paper is trying to show that a complete flag positroid — the data attached to a point of the nonnegative flag variety — can be captured by a single grid diagram, the flag positroid pipe dream (FPP), in the way a single positroid is captured by an L-diagram. Its linked claims: FPPs biject with intervals in the Bruhat order, with elbow count equal to the dimension of the corresponding positive Richardson cell; the rank-k constituent positroid is recovered by restricting the diagram to its first k rows and standardizing; and the nonnegatively representable elementary quotients of any positroid are exactly the diagrams obtained by adding one row with elbows in a nonempty subset of the unblocked columns. From this the paper derives an explicit description of the cyclic-shift freeze sets in terms of decorated permutations (marked permutations indexing positroids), partially resolving a problem left open in the literature, plus a self-duality theorem for the poset of nonnegatively representable quotients. A reader should care because the paper converts a piece of real algebraic geometry into a purely combinatorial, row-by-row constructible object whose components are directly readable.","feed_headline":"One pipe-dream diagram encodes an entire nonnegative flag","feed_subtitle":"Flag positroid pipe dreams read off every rank's positroid, its bases, and its L-diagram from a single grid.","key_machinery":"Flag positroid pipe dreams (FPPs): reduced Rothe pipe dreams of u — tiled diagrams of u's coinversions — with exit permutation v, filled by crosses and elbows and Γ-free (no cross with an elbow below it in its column and an elbow to its right whose pipe exits at or below the lower elbow). Key devices: a row-by-row algorithm producing FPP(u,v) exactly when u ≤ v; standardization, swapping adjacent rows and pivots while preserving bases and unblocked columns; the unblocked-column invariant (blocked iff a cross in it has an elbow to its right), governing where new elbows may appear; and the acyclic graph G(D), recovering bases via paths. A transfer map from two-step nonnegative flags to augment","core_discovery":"The central claim: every complete flag positroid has a faithful finite diagram, FPP(u,v) — the Rothe diagram of u tiled with crosses and elbows and required to be Γ-free. For u ≤ v in Bruhat order the diagram is unique, and its elbow count equals the dimension of the positive Richardson cell R^>0_{u,v}. The k-th constituent positroid is read from the first k rows via standardization, its bases being the sink sets of non-intersecting paths in an associated directed graph. Appending a row with elbows in any nonempty subset of the unblocked columns yields exactly the nonnegatively representable elementary quotients. Theorem 4.43 identifies this flag positroid with the one attached to any point","pith_inferences":["A row-by-row generation scheme for complete flag positroids follows naturally but is not pursued in the paper: start from the empty diagram and at each rank choose any nonempty subset of the currently unblocked columns. A testable extension — my inference, not the paper's claim — is that every point of the nonnegative flag variety admits such a build; the paper proves the quotient-to-row correspon","The freeze-set characterization is proved only for nonnegatively representable quotients. Comparing it with the existing characterization that covers all positroid quotients should isolate exactly where nonnegativity fails — plausibly at the greedy construction of T(C), where the monotonicity of unblocked values could break. That diagnosis is an editorial inference.","The paper shows FPPs are simultaneously Bruhat intervals and maximal chains of a self-dual poset; reading the two facts together suggests a mirror traversal of those chains — reversing a maximal chain pairs each Bruhat interval with another via inverse decorated permutations — a duality shadow the authors do not draw explicitly."],"forward_implications":["Every Bruhat interval [u,v] carries a complete flag positroid, and the elbow count of FPP(u,v) is a dimension computation: dim R^>0_{u,v} equals the number of elbows.","Any rank-k positroid P has exactly 2^{|U|} − 1 nonnegatively representable elementary quotients, where |U| is the number of unblocked columns of its partial FPP, and all of them are explicitly constructible by appending one row.","A single FPP contains all constituent data of a flag: the L-diagram of each P_k is the standardization of the k-row restriction, so flag positroids can be decomposed rank by rank without returning to the geometry.","Maximal chains in the poset of nonnegatively representable elementary positroid quotients are in bijection with FPPs and hence with positive Richardson cells; the poset is self-dual under matroid dualization, and its lattice-path-matroid subposet is self-dual as well.","In decorated-permutation terms, nonnegatively representable elementary quotients are exactly the right cyclic shifts of π with the explicitly computed freeze set A(C) = [n] ∖ (C ⊔ T(C)), for nonempty C among the unblocked positions of π."],"fun_headline_variants":["A pipe dream grid holds a full flag's data","Pipe dreams: one grid, all flag positroids","Elbow count gives flag cell dimension","Pipe dreams biject with Bruhat intervals","One diagram maps every nonnegative flag"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The argument leans on two supports not fully proved inside the paper: an imported lemma identifying a complete flag positroid by the lexicographically minimal and maximal bases u[k] and v[k] of its constituents, and the reverse inclusion in the standardization theorem (Theorem 4.34), which is only sketched; if either fails for arbitrary Γ-free pipe dreams, the FPP could encode a different flag positroid than the one attached to the Richardson cell.","fun_headline_variants_meta":{"raw":{"variants":["A pipe dream grid holds a full flag's data","Pipe dreams: one grid, all flag positroids","Elbow count gives flag cell dimension","Pipe dreams biject with Bruhat intervals","One diagram maps every nonnegative flag"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000924,"raw_usage":{"total_tokens":3809,"prompt_tokens":764,"completion_tokens":3045,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":508,"completion_tokens_details":{"reasoning_tokens":2977}},"tokens_in":508,"tokens_out":3045,"duration_ms":23581,"temperature":1.0,"reasoning_tokens":2977,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T13:09:40.406917+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a Bruhat interval [u,v] in S_4 or S_5 that is not attached to a Grassmannian permutation (a permutation with at most one descent), build FPP(u,v) with the paper's algorithm, restrict to the first k rows, standardize, and compute the bases of P_k as sink sets of non-intersecting paths; then brute-force enumerate every permutation w in [u,v] and restrict each to its first k values. If the two sets differ for any k — or if the lexicographically maximal basis is not v[k] — Theorem 4.43 is false. A cheaper probe checks the unproved direction of Theorem 4.34: on small partial FPPs with adjacent","supporting_citations":[],"review_version":2}