{"id":"ee5f3d97-a740-40d7-bf75-2dc7d94c14a9","arxiv_id":"1908.06959","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A new vector-relation framework on bipartite graphs unifies several discrete integrable systems and proves unique reconstruction from boundary data for plabic graphs.","lead":"This paper builds a geometric model in which vectors at the white vertices of a bipartite graph are constrained by linear relations at the black vertices, and shows it unifies the pentagram map, Q-nets, and discrete Darboux maps. It also proves that for plabic graphs, boundary vectors determine the whole configuration uniquely up to gauge, giving a new geometric proof of invertibility of the boundary measurement map.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The plabic reconstruction theorem is well supported; the load-bearing weakness is the Q-net cluster-algebra claim, whose proof in Proposition 3.10 is a one-sentence reference to a quiver with no mutation path shown.","rationale":"The reader's strongest claim is Theorem 4.3; I traced it through Lemma 4.4 through Proposition 4.23. The only genuinely external input is the Muller-Speyer theorem and the left-twist orthogonality property, both published; the internal steps, including the sign comparison in Proposition 4.9 and the acyclic orientation argument in Proposition 4.19, are sketched but plausible, and the full-rank check via the unique matching is legitimate. Thus I have no load-bearing objection to the plabic theorem itself. The conditional verdict is nevertheless justified by Proposition 3.10: the Q-net cluster structure is announced as a main result, but its proof is a one-sentence appeal to 'the associated cluster algebra.' No mutation path or quiver-mutation verification is provided, and the formulas are nontrivial enough that an indexing slip could go unnoticed. Reference [2] being unpublished compounds the difficulty of checking overlaps, but that is secondary. Since the reader already conditioned on the Q-net derivation gap, the verdict should remain conditional; I would not change it to accept or reject on the basis of this pass.","tokens_in":37507,"tokens_out":26179,"duration_ms":273342,"concrete_test":"Implement the quiver of Figure 12 and the Y-variable mutation rule, then traverse the mutation sequence induced by the five square moves of the gentrification step (Figure 11) on a finite patch covering one elementary cube and its neighbors. Compare the resulting Y-values with the formulas in Proposition 3.10 and their cyclic shifts for at least three consecutive generations; if any monomial, exponent, or denominator differs, the cluster structure for Q-nets is not established. A cheaper partial check is to verify that the quiver transforms into a quiver isomorphic to itself under the full gentrification step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 3.10 is the only place the paper substantiates the abstract's claim that Q-nets admit a cluster structure. Its proof reads 'One simply follows Y-variable dynamics of the associated cluster algebra, whose quiver is shown in Figure 12.' The text does not specify the mutation sequence corresponding to the gentrification move of Proposition 3.7, does not check that the quiver transforms consistently under that sequence, and does not verify that either displayed formula in Proposition 3.10 is actually produced by those mutations. The formulas involve three generations of variables and cyclic shifts; a single indexing or denominator error would invalidate the advertised resolution of the open question. This concern does not touch Theorem 4.3, whose reconstruction argument is detailed and rests on published Muller-Speyer results, but it is load-bearing for a second central claim made in the abstract.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces vector-relation configurations on planar bipartite graphs: a vector at each white vertex and a linear relation at each black vertex, modulo gauge equivalence. It shows that local graph transformations induce dynamics on these configurations that match classical urban renewal and degree-two vertex addition on edge weights (Proposition 2.1), and it defines gauge-invariant face weights that evolve by Y-variable dynamics. It then fits several known projective geometric systems into this framework: Laplace-Darboux dynamics, the pentagram map, Y-meshes, Q-nets, and discrete Darboux maps. The main computational claim for Q-nets is Proposition 3.10, which asserts that the Y-seed of Proposition 3.9 evolves by the cluster Y-dynamics of the quiver in Figure 12, thereby resolving an open question about cluster structures for Q-nets. For plabic graphs, the paper proves Theorem 4.3: the boundary restriction map Phi sends the configuration space C_G to the positroid variety Pi_M, and its restriction to the open subset C^o_G of configurations with all coefficients nonzero is an isomorphism onto T_G, the Muller-Speyer image of the boundary measurement map. This yields unique extension of generic boundary data to the interior. The paper also proves smoothness of C_G (Theorem 5.2) and identifies specializations from resistor and Ising networks.","tokens_in":37680,"tokens_out":6102,"duration_ms":66129,"significance":"If all claims hold, this is a valuable unification: one geometric formalism encompasses the pentagram map, Q-nets, and Darboux maps, and it provides a concrete inverse to Postnikov's boundary measurement map. The plabic graph portion is a genuine and largely self-contained theorem: the reconstruction map Psi defined by the hyperplane intersections L_w in (4.8), the Kasteleyn-sign comparison in Proposition 4.9, the uniqueness proof via acyclic orientations in Section 4.5, and the smoothness atlas in Section 5 are all presented with enough detail to be checked, building explicitly on published Muller-Speyer and Postnikov results. The Q-net cluster statement, if fully proved, would resolve a recognized open question, and Section 6 gives interesting new geometric consequences, including Koenigs nets from resistor networks and CKP maps from Ising networks. The main strength is the concrete, checkable plabic reconstruction machinery and the breadth of examples; the main weakness is the very abbreviated proof of the cluster-dynamics claim for Q-nets and related statements for Darboux maps.","major_comments":[{"comment":"The proof of Proposition 3.10 is a single sentence: 'One simply follows Y-variable dynamics of the associated cluster algebra, whose quiver is shown in Figure 12.' It does not specify the mutation sequence corresponding to the gentrification moves of Figure 11, does not check that the quiver transforms consistently under that sequence, and does not verify either displayed evolution formula. Because the abstract's claim that Q-nets admit a cluster structure rests on this proposition, this is a load-bearing gap, and the formulas involve enough indices and cyclic shifts that a small error would invalidate the advertised resolution. Please provide the explicit mutation sequence, the induced transformation of the quiver, and a derivation of at least one family of displayed Y-variable formulas, or clearly mark the cluster claim as conditional on a proof given elsewhere.","section":"Section 3.2, Proposition 3.10 and Figure 12"},{"comment":"The discrete Darboux case repeats the same pattern: Proposition 3.15 states that the square moves realize the time evolution, but its proof only says 'We verify the sequence of square moves using Proposition 2.4 on each step,' and the Y-dynamics are asserted by saying 'The Y-s evolve according to the Y-dynamics formulas of the associated cluster algebra. The formulas are too long to be written here.' Since the introduction advertises a cluster algebra operating in 'all cases,' this case must receive the same level of verification as the Q-net case, or be explicitly stated as a conjecture with the missing computation deferred to an appendix or a citable companion.","section":"Section 3.3, Proposition 3.15 and the paragraph after Figure 15"},{"comment":"Remark 3.12 cites the authors' own in-progress work [2] ('In progress') as giving independent cluster descriptions of Q-nets and Darboux maps. If this reference is used to support the claim that the open question is resolved, it should be replaced by a citable preprint or the dependence should be removed; the present manuscript should not rely on an unpublished companion for a central claim. This is not the main obstacle to acceptance, but it is a missing-support issue that should be corrected.","section":"Remark 3.12"}],"minor_comments":[{"comment":"The labels a, a', b, b', c, c', e, f, g, h, and G, H are defined only in the proof of Proposition 3.7; adding a caption that explains the labeling would make the figure much easier to read.","section":"Figure 11"},{"comment":"The two displayed evolution formulas in Proposition 3.10 are not numbered, making it awkward to refer to them in the text or in a future erratum; please number them.","section":"Proposition 3.10"},{"comment":"The Q-net quiver in Figure 12 is central to the claimed cluster structure, but the text gives no description of its vertex set or the correspondence between vertices and the Y-variables of Proposition 3.9; this should be stated explicitly even if a full mutation computation is deferred.","section":"Section 3.2, Figure 12"},{"comment":"There are typographical spacing artifacts in the title and abstract (e.g., 'VECTOR-RELA TION' and several words split by line breaks); please proofread the final version.","section":"Abstract and title page"}],"recommendation":"major_revision","confidential_remarks":"The plabic graph theorem is solid and detailed, and the reconstruction map is a real contribution. My recommendation is driven entirely by the abbreviated treatment of the Q-net cluster dynamics, which is the paper's advertised resolution of an open problem; Proposition 3.10 needs a verifiable proof or an explicit withdrawal of the claim at that strength. I would also suggest the editors ensure that the relation to the in-progress work [2] is clarified before publication, since the ordering of claims may matter for priority."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe punchline: the plabic graph half of this paper is solid and worth engaging with; the advertised cluster algebra structure for Q-nets is not actually demonstrated. Theorem 4.3 — the unique extendability of a boundary configuration to the interior for reduced plabic graphs — is a clean geometric reformulation of the invertibility of Postnikov's boundary measurement map, and the proof is detailed and honest about its reliance on Muller–Speyer. The reconstruction map defined by intersecting hyperplanes is elegant, and the smoothness theorem (5.2) for the configuration space is a real new result. The vector-relation framework itself successfully unifies the pentagram map, Q-nets, discrete Darboux maps, and Y-meshes, and the face-weight interpretation via multi-ratios is genuinely useful.\n\nThe soft spot is precisely where the abstract makes its strongest claim. Proposition 3.10, which claims to resolve the open question of cluster structure for Q-nets, is proved by the sentence 'One simply follows Y-variable dynamics of the associated cluster algebra, whose quiver is shown in Figure 12.' No mutation sequence, no verification that the quiver evolves consistently, no derivation of the displayed Y-variable formulas. That is a research announcement, not a proof. It may be true, but as written it does not support the abstract's claim. The dependence on unpublished work [2] for an independent cluster description compounds the issue.\n\nThe stress-test note gets this exactly right: the plabic reconstruction theorem holds up; the Q-net cluster claim is load-bearing and under-supported. I don't think this sinks the paper — the plabic contribution stands on its own — but the abstract oversells. A referee should ask for a complete computation in Section 3.2 or for the claim to be downgraded to a conjecture.\n\nThis paper is for people working in discrete differential geometry, cluster algebras, and positroid theory. It deserves a serious referee: the main theorem is important and largely sound, and the Q-net question is significant even if this treatment is incomplete. I'd send it out, but with a clear expectation that Section 3.2 needs substantial work. My recommendation: engage with it, but don't trust the cluster-algebra claim until you've seen the computation.","headline":"Solid plabic reconstruction theorem and a useful unifying framework, but the Q-net cluster-algebra claim is underproved as written.","tokens_in":38186,"tokens_out":3838,"would_cite":true,"duration_ms":34557,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05E14","13F60","37K10","51A05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a reduced plabic graph, boundary vectors determine the whole vector-relation configuration up to gauge, and the map from configurations to boundary points inverts the boundary measurement map.","keywords":["vector-relation configurations","plabic graphs","boundary measurement map","pentagram map","Q-nets","discrete Darboux maps","cluster algebras","positroid varieties"],"falsifier":"Take a small reduced plabic graph, such as the example in the paper with two boundary dimension equal to four, and compute the boundary restriction map $\\Phi$ on a one-parameter family of configurations in $\\mathcal C_G^\\circ$. If two distinct parameter values give the same boundary point $A$, Theorem 4.3(2) is false; conversely, random boundary points $A\\in T_G$ fed through the reconstruction should always produce a full-rank coefficient matrix $K$ with all edge coefficients nonzero, and the first failure would falsify the theorem.","tokens_in":37303,"feed_emoji":"📐","tokens_out":12860,"duration_ms":124899,"temperature":0.7,"pith_summary":"This paper proposes a single geometric state for any planar bipartite graph: put a nonzero vector at each white vertex and let each black vertex carry a nontrivial linear relation among the vectors of its neighbors. The paper shows that the standard local moves of the dimer model act naturally on these states, so every such system comes with face weights and cluster dynamics. On plabic graphs, the main result is that boundary vectors determine the whole interior configuration uniquely up to gauge whenever all relation coefficients are nonzero; this is a geometric inverse of the boundary measurement map. The same formalism contains the pentagram map, Q-nets, and discrete Darboux maps as special cases, and it gives Q-nets a cluster algebra structure that was previously missing.","feed_headline":"Boundary data uniquely determine plabic configurations","feed_subtitle":"A vector-and-relation state inverts the boundary measurement map and unifies several discrete integrable systems.","key_machinery":"The central object is the vector-relation configuration $(v,R)$: a nonzero vector in $\\mathbb C^k$ at each white vertex and a nontrivial linear relation among the neighboring vectors at each black vertex, modulo gauge transformations. The load-bearing construction is the reconstruction map $\\Psi$: starting from a boundary point $A$, it builds hyperplanes $H_j$ from the basis subsets $I_j\\setminus\\{j\\}$ and defines candidate lines $L_w$ by intersecting the $H_j$ for strands $j$ lying to the left of $w$. What makes the construction valid is the cited description of $T_G$: on $T_G$, all relevant maximal minors of the twisted boundary data are nonzero, so the hyperplanes around each face are in general position and each $L_w$ is a genuine line. A sign comparison identifies $\\Phi$ with the boundary measurement map, transferring uniqueness to and from the classical map.","core_discovery":"The central discovery is that on a reduced plabic graph all the information in a vector-relation configuration is carried by its boundary. For a configuration with every relation coefficient $K_{bw}$ nonzero, the boundary vectors $v_1,\\dots,v_n$ (a point $A=[v_1\\cdots v_n]$ in the space of $k$-dimensional subspaces of $\\mathbb C^n$) determine the configuration uniquely up to gauge at internal vertices. Theorem 4.3 states this sharply: the boundary restriction map $\\Phi$ sends all configurations into the positroid variety $\\Pi_{\\mathcal M}$, its restriction to $\\mathcal C_G^\\circ$ is an isomorphism onto $T_G$, the image of the boundary measurement map, and $\\mathcal C_G^\\circ=\\Phi^{-1}(T_G)$. The proof constructs the inverse explicitly: for $A\\in T_G$, take the hyperplanes $H_j$ spanned by the boundary vectors indexed by $I_j\\setminus\\{j\\}$, where $(I_1,\\dots,I_n)$ is the necklace of basis subsets coming from the matroid, and for each white vertex $w$ form the line $L_w=\\bigcap_{j\\in S_w} H_j$; the vectors on these lines satisfy exactly one linear relation at each black vertex, with all coefficients nonzero.","pith_inferences":["An algorithmic reading the paper leaves implicit is that the reconstruction map gives a practical way to solve geometric extension problems: given boundary data in $T_G$, each internal vector is computable by intersecting hyperplanes indexed by strands, with no need to sum matchings.","Because non-uniqueness can only occur outside $T_G$, probing the fibers of $\\Phi$ on the boundary of the positroid variety—as in the four-boundary, dimension-two example—should describe exactly how the smooth space $\\mathcal C_G$ resolves singularities of $\\Pi_{\\mathcal M}$.","The same vector-relation formalism is a natural search tool for cluster structures in other lattice geometries: any bipartite graph whose local moves are urban renewal carries $Y$-variables, so analogous discrete nets are candidates for integrable cluster dynamics."],"forward_implications":["For any reduced plabic graph, every boundary point in the image $T_G$ of the boundary measurement map extends to exactly one gauge class of configurations in $\\mathcal C_G^\\circ$; internal vectors are determined up to scale.","The reconstruction is explicit: hyperplane intersections built from the basis subsets and strand data recover all internal vectors and relations, giving a geometric algorithm for the inverse boundary measurement map.","Positive edge weights of a totally nonnegative boundary point acquire a geometric reading: after gauging, each internal vector is a convex combination of vectors two steps upstream, so edge weights are barycentric coordinates in a recursive construction.","The local-move dynamics on vector-relation configurations specialize to the pentagram map, Q-nets, and discrete Darboux maps, and the face weights evolve by cluster $Y$-dynamics; in particular Q-nets carry a cluster structure.","Configurations obtained from resistor networks form a special class of conjugate nets, while those obtained from the corresponding spin model satisfy an extra conic condition; the model locates both subvarieties inside the same state space."],"supporting_citations":[{"why":"Supplies Theorem 4.2: the image $T_G$ of the boundary measurement map is described via the twist and nonvanishing of maximal minors, the property used to define the hyperplanes and prove uniqueness.","marker":"[23]"},{"why":"Defines plabic graphs and the boundary measurement map through matchings and path families, the map whose image $T_G$ is the target of the isomorphism.","marker":"[26]"},{"why":"Proves that the positroid variety $\\Pi_{\\mathcal M}$ is exactly the set of points with maximal minors vanishing outside the matroid, used to show $\\Phi$ lands in $\\Pi_{\\mathcal M}$.","marker":"[19]"},{"why":"Gives the dimer-model local moves and face weights that the vector-relation dynamics is shown to reproduce.","marker":"[13]"},{"why":"Provides the dual relation-space model for plabic graphs, to which the boundary restriction statement is compared.","marker":"[21]"},{"why":"Supplies the definitions of Q-nets and discrete Darboux maps and the conjugate-net characterization used in the resistor-network section.","marker":"[3]"}],"fun_headline_variants":["Boundary data uniquely determine plabic graphs","Plabic configs: boundary vectors fix the interior","Inverting Postnikov's map via boundary vectors","Unique extension from boundary for plabic graphs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The uniqueness proof rests entirely on a cited theorem that describes exactly which boundary points can occur; if that description were wrong or incomplete, the inverse map would have no foundation.","fun_headline_variants_meta":{"raw":{"variants":["Boundary data uniquely determine plabic graphs","Plabic configs: boundary vectors fix the interior","Inverting Postnikov's map via boundary vectors","Unique extension from boundary for plabic graphs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000162,"raw_usage":{"total_tokens":1224,"prompt_tokens":918,"completion_tokens":306,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":534,"completion_tokens_details":{"reasoning_tokens":248}},"tokens_in":534,"tokens_out":306,"duration_ms":3443,"temperature":1.0,"reasoning_tokens":248,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:29:38.435536+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a small reduced plabic graph, such as the example in the paper with two boundary dimension equal to four, and compute the boundary restriction map $\\Phi$ on a one-parameter family of configurations in $\\mathcal C_G^\\circ$. If two distinct parameter values give the same boundary point $A$, Theorem 4.3(2) is false; conversely, random boundary points $A\\in T_G$ fed through the reconstruction should always produce a full-rank coefficient matrix $K$ with all edge coefficients nonzero, and the first failure would falsify the theorem.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Theorem 4.2: the image $T_G$ of the boundary measurement map is described via the twist and nonvanishing of maximal minors, the property used to define the hyperplanes and prove uniqueness."},{"cited_title":"Total positivity, Grassmannian s, and networks","cited_arxiv_id":null,"evidence_quote":"Defines plabic graphs and the boundary measurement map through matchings and path families, the map whose image $T_G$ is the target of the isomorphism."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves that the positroid variety $\\Pi_{\\mathcal M}$ is exactly the set of points with maximal minors vanishing outside the matroid, used to show $\\Phi$ lands in $\\Pi_{\\mathcal M}$."},{"cited_title":"Goncharov and Richard Kenyon","cited_arxiv_id":null,"evidence_quote":"Gives the dimer-model local moves and face weights that the vector-relation dynamics is shown to reproduce."},{"cited_title":"Totally nonnegative Grassmannian and Gras smann polytopes","cited_arxiv_id":null,"evidence_quote":"Provides the dual relation-space model for plabic graphs, to which the boundary restriction statement is compared."},{"cited_title":"Bobenko and Yuri B","cited_arxiv_id":null,"evidence_quote":"Supplies the definitions of Q-nets and discrete Darboux maps and the conjugate-net characterization used in the resistor-network section."}],"review_version":1}