{"id":"b2e7e535-dc4e-4546-baae-914132fd63ce","arxiv_id":"2509.06348","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors conjecture that every LOCC-monotone multi-invariant is labeled by a finite Coxeter group and prove this for the A_n, B_n, D_n and I_n families.","lead":"The paper studies quantum entanglement measures for many-particle states and claims these measures are fully labeled by a family of symmetry groups called finite Coxeter groups. It proves this for most cases and leaves six exceptional cases open.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.1's bridge to Marc's mirror-graph classification is asserted, not proved; the 'only if' direction of Conjecture 1.1 depends on it.","rationale":"The reader's weakest_assumption already identifies the asserted identification with Marc's mirror graphs as the key unproved step, and I agree that this is the most load-bearing point: without a complete proof of the equivalence in Theorem 3.1, the necessity direction of the main conjecture is not established. I also share the reader's secondary concern about the unproved extension of reflecting planes in Lemma 3.1, which supports the positive results. Since the reader's verdict is CONDITIONAL and my reading does not reveal a counterexample or a contradiction, I do not move the verdict. The requested concrete check is therefore to write out the full dictionary between the two notions of mirror graph; this would settle whether the bridge to Marc's theorem actually holds or whether additional edge-label and automorphism conditions must be imposed.","tokens_in":11909,"tokens_out":19125,"duration_ms":172387,"concrete_test":"Supply a complete translation between reflecting cuts (Definition 2.3) and the cuts of Marc's mirror graphs used in [5, Theorem 2.8], verifying that the edge-label-preserving odd automorphism k corresponds exactly to the graph isometry in Marc's classification and that 'every edge lies in a reflecting cut' is identical to Marc's mirror condition. As a finite warm-up, list all reflecting cuts of Z_{F_4} and check them against Marc's cuts; if any reflecting cut fails to be a Marc cut (or vice versa) because of edge labels or vertex colors, Theorem 3.1(3) and the only-if direction of Conjecture 1.1 are false.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing issue is the unproved bridge in Theorem 3.1. The 'only if' direction of Conjecture 1.1 runs through Remark 2.8 and Theorem 3.1, and Theorem 3.1(2)⇔(3) is dispatched in Appendix B by the sentence 'Our definition of mirror ψ-graph is equivalent to the definition of mirror graphs given in [5,9]' plus a sketch. Because Definition 2.3 requires an edge-label-preserving odd automorphism, whereas Marc's classification is stated for unlabelled mirror graphs, the correspondence is not automatic; a mismatch here would break the necessity of the Coxeter characterization. The associated proof of 2⇒1 is also compressed: the geodesic/odd-cycle argument is not fully written, and the claim that a reflecting cut containing a geodesic edge contains neither e1 nor e2 is asserted. Separately, Lemma 3.1 assumes without proof that a reflecting cut of Cay(H,K) extends uniquely to Cay(G,S); this extension underpins Proposition 1.1 and Theorem 1.2. These are gaps in proof, not known counterexamples, but they leave the central classification conditional.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies graph-theoretic local unitary invariants of pure multipartite quantum states, called multi-invariants, which are labeled by bipartite uniformly edge-labeled graphs ('psi-graphs'). The main object is the combinatorial condition of edge-convexity introduced by the authors in [1], which guarantees that the associated normalized invariant defines a pure-state entanglement monotone. The paper conjectures that a connected psi-graph is edge-convex if and only if it is the Cayley graph of a finite Coxeter group with its standard involutive generators. The 'only if' direction is claimed via the necessary edge-reflecting condition and an external classification of mirror graphs, while the 'if' direction is proved for the families A_n, B_n(=C_n), D_n, and for direct products, leaving the six exceptional Coxeter diagrams E_6,E_7,E_8,F_4,H_3,H_4 open. The main technical novelty is a reduction to a vertex-convexity condition for Schreier coset graphs, together with Lemma 3.1, which transfers edge-convexity from a subgroup to the full group.","tokens_in":12132,"tokens_out":9287,"duration_ms":82833,"significance":"If the conjecture is correct, it gives a clean and surprising classification of a natural class of entanglement monotones in terms of finite Coxeter groups, and it reduces the remaining cases to a finite list. The paper's partial results are valuable: Proposition 1.1 reduces the conjecture to connected Coxeter-Dynkin diagrams, and Theorem 1.2 verifies three infinite families. The strategy of passing to coset graphs and vertex-convexity is elegant and likely to be reusable. However, several load-bearing steps in the proofs are only sketched or asserted, in particular the translation between the paper's reflecting cuts and the external mirror-graph classification, and the extendibility of reflecting cuts used in Lemma 3.1. As it stands, the classification is conditional on filling these gaps.","major_comments":[{"comment":"Lemma 3.1 is load-bearing for Proposition 1.1 and Theorem 1.2, but its proof relies on the unproved assertion that a reflecting cut of Cay(H,K) extends uniquely to a reflecting cut of Cay(G,S). In Section 3.1 this is introduced with 'First note', and the proof in Appendix B repeats the statement without argument. The definition of vertex-convexity (Definition 3.2, Eq. (7)) also depends on which reflecting planes are extendible, so without a proof of uniqueness and existence of extensions, both the lemma and the examples built on it are not fully established.","section":"Section 3.1 and Appendix B, Lemma 3.1"},{"comment":"The vertex-convexity verifications for the orthoplex and the demi-hypercube are incomplete. For the n-orthoplex, the reflecting plane transverse to the first axis produces a matrix only for the pair (+-1,0,...,0); it does not cover arbitrary pairs such as (1,0,...,0) and (0,1,...,0), which are not separated by that plane. For the demi-hypercube, the single pi/4 co-dimension-one plane similarly addresses only one mirror pair. The hypercube discussion works for pairs related by the chosen hyperplane reflection, but the paper does not explain how the remaining pairs are handled. Since Eq. (7) requires a solution for every pair of vertices, these sketches do not yet prove edge-convexity of Z_{B_n} and Z_{D_n}.","section":"Section 3.2, vertex-convexity of B_n and D_n coset graphs"},{"comment":"The equivalence between the paper's mirror psi-graphs and the unlabelled mirror graphs of [5] is asserted rather than proved. Definition 2.3 requires an edge-label-preserving odd automorphism that flips vertex colors, whereas the classification in [5] is stated for unlabelled mirror graphs. The sentence 'Our definition of mirror psi-graph is equivalent to the definition of mirror graphs given in [5,9]' and the subsequent sketch are not a proof of this dictionary. A mismatch between the two notions would invalidate the 'only if' direction of Conjecture 1.1. The sketch also asserts, without derivation, that every pair of neighboring edges lies on a unique convex cycle of the form (s_A s_B)^{m_{AB}}; this is a substantive claim that needs to be justified.","section":"Appendix B, Theorem 3.1, 2)<=>3)"},{"comment":"The proof of the implication 2)=>1) in Theorem 3.1 is too compressed. The step 'Also it contains neither e1, nor e2' is asserted without proof, and the construction of the odd cycle p-e1-p' is not fully written out. Because this implication is needed for the equivalence between mirror psi-graphs and edge-reflecting graphs, the gap affects the claimed classification and not merely the exposition. Please provide a complete argument, including a proof that a reflecting cut containing a geodesic edge of p cannot contain e1 or e2.","section":"Appendix B, Theorem 3.1, 2)=>1)"}],"minor_comments":[{"comment":"The phrase 'E_6,7,8, F_4, H_3,5' should read 'E_6, E_7, E_8, F_4, H_3, H_4'.","section":"Introduction"},{"comment":"The cross-references 'the first figure of ??' and 'the second figure of ??' are unresolved; they should point to the actual figure in the manuscript.","section":"Section 3.2"},{"comment":"The matrices M^{(k)}(e,e') and M^{(k)}(v,v') are used in Eqs. (6) and (7) but are not defined precisely before first use. Please define them explicitly or refer to the definitions in [1].","section":"Definition 2.4 and Definition 3.2"},{"comment":"The proof of 4)=>1) in Theorem 2.1, especially the induction showing that the alternating 4-cycle property forces a hypercube, is very terse. Since this theorem is invoked in the proof of Theorem 3.1, please expand the induction step or cite a full proof.","section":"Appendix B, Theorem 2.1"},{"comment":"The labels '4 5 5' and 'nIn' appear to be embedded in the displayed Coxeter diagrams in a confusing way; the figure should be redrawn so that edge labels and diagram names are clearly separated from the graph edges.","section":"Figure 5"},{"comment":"The argument for the inequality |Z(|psi>)| <= 1 is stated for the unnormalized invariant Z, while the Cauchy-Schwarz justification is cleanest for the normalized invariant; please clarify the normalization in the statement.","section":"Remark 2.2"}],"recommendation":"major_revision","confidential_remarks":"The core issue is the unproved bridge between the labelled reflecting-cut definition of this paper and the unlabelled mirror-graph classification imported from [5]. The authors should be asked to prove a precise translation lemma; without it, the 'only if' direction of Conjecture 1.1 is not established. The paper also needs a rigorous proof of the extendibility claim in Lemma 3.1 and completed vertex-convexity constructions for the orthoplex and demi-hypercube examples. If these gaps are fixed, the paper would be a solid contribution to the graph-theoretic study of entanglement monotones."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this paper introduces a genuinely new conjecture -- edge-convex psi-graphs are exactly Cayley graphs of finite Coxeter groups -- and backs it with a partial proof for the A_n, B_n, D_n families plus a useful new technique. It deserves serious peer review, but the referee should press hard on the bridge to Marc's mirror-graph classification, because the 'only if' direction of the conjecture runs through it and it is currently asserted rather than fully demonstrated.\n\nThe new content is real: the conjecture itself, the vertex-convexity method for coset graphs, and Theorem 1.2 for A_n, B_n, D_n. I_n was already edge-convex from the authors' earlier paper, and the six exceptional cases (E_6, E_7, E_8, F_4, H_3, H_4) are openly left open. The paper is honest about the boundary of what it proves.\n\nThe main soft spot is exactly what the stress-test flags: the equivalence (2) iff (3) in Theorem 3.1. Appendix B dismisses the correspondence between the paper's mirror psi-graphs and Marc's mirror graphs with one sentence and a sketch. But the paper's reflecting cut requires an edge-label-preserving odd automorphism, while Marc's classification is stated for unlabelled mirror graphs. The correspondence is plausible, but it is not automatic, and the 'only if' direction of Conjecture 1.1 rests on it. A mismatch here would break the Coxeter characterization, not just weaken a minor lemma.\n\nTwo smaller gaps sit nearby: Lemma 3.1 assumes without proof that a reflecting cut of Cay(H,K) extends uniquely to Cay(G,S), and the 2=>1 direction of Theorem 3.1 compresses the geodesic/odd-cycle argument. These are gaps in proof, not known counterexamples. My reading is that the partial theorem is very likely true, but the proof as written leaves the central classification conditional.\n\nThe citation pattern is fine: the authors cite their own earlier work for background, Marc's and Bresar et al.'s independent results for the mirror-graph side, and the new classification claim is not present in those sources. The manuscript also has unresolved '??' figure cross-references; minor, but sloppy.\n\nWho this is for: people working on entanglement monotones, multi-invariants, or Cayley-graph/Coxeter combinatorics. If the conjecture survives, it is a tidy unification. I would send it to peer review and would cite it for the conjecture and the vertex-convexity technique. My own verdict is conditional: the referee should demand a rigorous proof of the Theorem 3.1 bridge before the 'only if' direction is considered established.","headline":"A clean conjecture connecting edge-convex psi-graphs to finite Coxeter groups, with a credible partial proof; the load-bearing bridge to Marc's mirror-graph classification is asserted more than proved and deserves real referee attention.","tokens_in":12670,"tokens_out":2298,"would_cite":true,"duration_ms":19782,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05C25","20F55","81P45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper conjectures a complete classification of LOCC-monotone multi-invariants: a psi-graph is edge-convex exactly when it is the Cayley graph of a finite Coxeter group, and proves the classification for all but six exceptional diagrams.","keywords":["entanglement monotones","LOCC","multi-invariants","psi-graphs","edge-convexity","Coxeter groups","Cayley graphs","mirror graphs"],"falsifier":"Take the Cayley graph of each of the six exceptional Coxeter groups $E_6,E_7,E_8,F_4,H_3,H_4$, enumerate all reflecting cuts, and check whether the linear system (6) admits a positive-semidefinite solution; any failure would disprove the conjecture's 'if' direction, while an explicit edge-convex $\\psi$-graph whose automorphism group is not a finite Coxeter group would disprove the 'only if' direction.","tokens_in":11699,"feed_emoji":"🔗","tokens_out":12134,"duration_ms":100276,"temperature":0.7,"pith_summary":"Pure-state entanglement measures that are monotone on average under local operations and classical communication can be built from multi-invariants, polynomials in a multipartite state and its conjugate that are invariant under local unitary transformations. Each multi-invariant is labeled by a bipartite, edge-colored graph, and the authors' earlier work turned monotonicity into a graph condition called edge-convexity. The paper's main conjecture says that a connected graph of this kind is edge-convex exactly when it is the Cayley graph of a finite Coxeter group with standard involutive generators. The authors prove the 'only if' direction in full, and prove the 'if' direction for the Coxeter families $A_n$, $B_n=C_n$, and $D_n$, leaving the six exceptional connected diagrams $E_6,E_7,E_8,F_4,H_3,H_4$ open. If the conjecture is right, the multi-invariants that give LOCC-monotone pure-state entanglement measures are completely classified by finite Coxeter groups.","feed_headline":"Edge-convex graphs match finite Coxeter groups","feed_subtitle":"The graph condition behind LOCC-monotone entanglement measures coincides with Coxeter Cayley graphs; A_n, B_n, D_n proven.","key_machinery":"The load-bearing object is the $\\psi$-graph of a multi-invariant: a bipartite graph whose white and black vertices are copies of the state and its conjugate, and whose edge labels record which party's index is contracted. Edge-convexity asks that for every pair of same-label edges there exist reflecting cuts, which are odd automorphisms that swap two sides of the graph, together with positive-semidefinite matrices $M^{(k)}$ satisfying equation (6); this is exactly the condition that $\\hat\\nu(Z)=1-\\hat Z$ does not increase under LOCC. The proof machinery is the equivalence of Theorem 3.1, which identifies edge-reflecting $\\psi$-graphs with finite-Coxeter Cayley graphs, plus Lemma 3.1, which lifts edge-convexity of a subgroup Cayley graph and vertex-convexity of a coset graph to edge-convexity of the full Cayley graph. This lift is what powers the inductive proofs for $A_n$, $B_n$, and $D_n$.","core_discovery":"The paper's central claim is Conjecture 1.1: a connected $\\psi$-graph $Z$ is edge-convex if and only if it is a Cayley graph of a finite Coxeter group with standard involutive generators. Here $\\psi$-graphs are the contraction graphs of multi-invariants, and edge-convexity is the property that makes the normalized invariant $\\hat\\nu(Z)=1-\\hat Z$ a pure-state entanglement monotone. The authors prove the 'only if' direction by showing edge-convexity implies the weaker edge-reflecting condition, then using a graph-theoretic classification theorem to conclude that edge-reflecting $\\psi$-graphs are exactly Cayley graphs of finite Coxeter groups. They prove the 'if' direction for the Coxeter-Dynkin diagrams $A_n$, $B_n(=C_n)$, and $D_n$ by induction, using a lemma that lifts edge-convexity from a subgroup to a larger group; since $I_n$ was already known and disconnected diagrams follow from connected ones, the full conjecture is reduced to the six exceptional connected diagrams $E_6,E_7,E_8,F_4,H_3,H_4$.","pith_inferences":["If the six exceptional cases are settled, quantum information theory gains entanglement monotones indexed by exceptional Coxeter groups; the $E_8$ case would be an eight-party monotone with unusually large symmetry.","Because edge-convexity is purely a property of the contraction graph, the same classification likely extends to resource theories whose monotones are built by pairing tensors with their duals, not just quantum entanglement.","A finite computation, namely solving the positive-semidefinite program (6) for the six remaining diagrams, could settle the 'if' direction without new conceptual machinery.","The logical hygiene of the 'only if' direction sits in the identification between reflecting cuts in colored graphs and the mirror-graph classification it invokes; a rigorous formal check of that identification would strengthen the proof."],"forward_implications":["Every connected edge-convex multi-invariant is labeled by a finite Coxeter group, so the classification reduces the search for pure-state entanglement monotones to a finite list of Coxeter-Dynkin diagrams.","The earlier examples $E^{(2)}$, $E^{(3)}$, and $C_n$ fit the conjecture as $Z_{A_1\\sqcup A_1}$, $Z_{A_1\\sqcup A_1\\sqcup A_1}$, and $Z_{I_n}$, confirming the proposal on known cases.","Edge-convexity of a disconnected Coxeter diagram follows from edge-convexity of its components by Proposition 1.1, so only the six exceptional connected diagrams need checking to finish the classification.","The coset-graph machinery yields explicit certificates: each reflecting cut contributes a positive-semidefinite matrix $M^{(k)}$, so a proved edge-convex graph comes with a checkable verification of monotonicity."],"supporting_citations":[{"why":"The authors' earlier paper: it defines multi-invariants' psi-graphs, introduces edge-convexity, and proves Theorem 1.1 that edge-convex graphs yield pure-state entanglement monotones.","marker":"[1]"},{"why":"Classification of mirror graphs as Cayley graphs of finite Coxeter groups; Theorem 3.1 invokes it to prove the equivalence of edge-reflecting and Coxeter Cayley graphs.","marker":"[5]"},{"why":"Jointly with [5], it supplies the definition of mirror graphs used to identify reflecting cuts in the paper's colored graphs.","marker":"[9]"},{"why":"Structural characterization of Cayley graphs used in Theorem 2.1 to equate vertex-transitive psi-graphs with Cayley graphs on involutive generators.","marker":"[7]"},{"why":"The cited fixed-point-free automorphism theorem used inside the proof of the Cayley-graph characterization.","marker":"[8]"}],"fun_headline_variants":["Coxeter groups classify LOCC monotones","Edge-convex graphs match Coxeter Cayley","LOCC monotones: Coxeter classification near complete","Six exceptions: edge-convex to Coxeter","Coxeter graphs unlock entanglement monotones"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the colored bipartite graphs whose edges are covered by reflecting cuts are exactly the unlabelled mirror graphs in the cited classification; if the edge-label-preserving odd automorphisms do not match that classification's symmetries, the Coxeter characterization of the necessary condition collapses.","fun_headline_variants_meta":{"raw":{"variants":["Coxeter groups classify LOCC monotones","Edge-convex graphs match Coxeter Cayley","LOCC monotones: Coxeter classification near complete","Six exceptions: edge-convex to Coxeter","Coxeter graphs unlock entanglement monotones"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00035,"raw_usage":{"total_tokens":1898,"prompt_tokens":922,"completion_tokens":976,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":538,"completion_tokens_details":{"reasoning_tokens":905}},"tokens_in":538,"tokens_out":976,"duration_ms":9175,"temperature":1.0,"reasoning_tokens":905,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:16:45.642370+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the Cayley graph of each of the six exceptional Coxeter groups $E_6,E_7,E_8,F_4,H_3,H_4$, enumerate all reflecting cuts, and check whether the linear system (6) admits a positive-semidefinite solution; any failure would disprove the conjecture's 'if' direction, while an explicit edge-convex $\\psi$-graph whose automorphism group is not a finite Coxeter group would disprove the 'only if' direction.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Classification of mirror graphs as Cayley graphs of finite Coxeter groups; Theorem 3.1 invokes it to prove the equivalence of edge-reflecting and Coxeter Cayley graphs."},{"cited_title":"Breˇ sar, S","cited_arxiv_id":null,"evidence_quote":"Jointly with [5], it supplies the definition of mirror graphs used to identify reflecting cuts in the paper's colored graphs."},{"cited_title":"Structural characterization of Cayley graphs","cited_arxiv_id":"1609.08272","evidence_quote":"Structural characterization of Cayley graphs used in Theorem 2.1 to equate vertex-transitive psi-graphs with Cayley graphs on involutive generators."},{"cited_title":"Sabidussi,On a class of fixed-point-free graphs, Proceedings of the American Mathematical Society9(5), 800 (1958)","cited_arxiv_id":null,"evidence_quote":"The cited fixed-point-free automorphism theorem used inside the proof of the Cayley-graph characterization."}],"review_version":2}