{"id":"e8868f8f-835d-4eb5-8199-3a02c65f09be","arxiv_id":"2607.03939","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Optimized two-site Clifford disentanglers for the Haldane phase realize the generalized Kramers-Wannier map, proven optimal for the AKLT state and converting SPT order into Z2 spontaneous symmetry breaking.","lead":"The optimal local Clifford disentanglers for the Haldane phase of spin-1 chains implement the generalized Kramers-Wannier transformation. This both reduces entanglement for classical simulation and unitarily maps the SPT phase onto a Z2 symmetry-breaking phase, distinct from the Kennedy-Tasaki route.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The strongest claim is carefully scoped to the sequential two-site Clifford ansatz used by CAMPS-DMRG. Within that scope the analytic argument (canonical-form recurrence a_{j+1}=(2-a_j)/3 staying inside [4/9,2/3], majorization of all other Schmidt spectra by the SUM-type spectrum, and the final right-to-left check that identity is optimal after the KW sweep) is complete and does not rely on unstated assumptions. The numerical CAMPS-DMRG results recover the same circuit on two independent models, and the Z2-SSB diagnostics (long-range ⟨S^z_j S^z_{j+r}⟩=1/4 and the order-parameter ⟨S^z_j⟩=±1/2 depending on the left edge) are derived explicitly for the KW-transformed AKLT state. The boundary-initiated character of the disentangler is a genuine limitation of the ansatz class, but it is acknowledged by the authors and does not constitute a correctness risk for the claim as stated. Consequently the reader's ACCEPT verdict with high confidence remains appropriate; no adjustment is warranted.","tokens_in":39019,"tokens_out":576,"duration_ms":4828,"concrete_test":"Independently recompute the characteristic polynomials of F(\nu,a) for the 90 two-qutrit Clifford classes (Table S2) at a=1/2 and confirm that type T0 (containing U_j,j+1) is the unique majorizing spectrum; if any other class yields a strictly smaller binary entropy, the optimality claim for the AKLT sweep fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest-assumption note correctly flags that optimality is proved only inside the sequential two-site Clifford ansatz that must start from an open boundary (SM §S4, Prop. 1, Lem. 3, Thm. 1). That restriction is essential: the paper itself shows that no two-site Clifford reduces entanglement between two bulk AKLT tensors. However, the restriction is stated openly, is the natural setting of CAMPS-DMRG, and does not undermine the central claim as formulated. The analytic optimality proof for AKLT is self-contained (canonical-form propagation + majorization of Schmidt spectra over the 90 inequivalent two-qutrit Cliffords), the numerical recovery of the same circuit on Heisenberg and BLBQ is consistent, and the Z2-SSB mapping is verified both by order-parameter calculation and by explicit action on edge spins. No internal inconsistency or hidden assumption that would falsify the strongest claim was found.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper extends CAMPS-based DMRG to qutrit (spin-1) systems and shows that the optimized sequential two-site Clifford disentanglers for the Haldane phase of the Heisenberg and bilinear-biquadratic models implement the generalized Kramers–Wannier (KW) transformation U_KW = ∏ U_{j,j+1} with U_{j,j+1} = X_{j+1}^2 U_SUM_{j,j+1}. For the AKLT state the authors introduce a (u,v;a)-canonical form, prove that U_{j,j+1} preserves it while minimizing the Schmidt spectrum among the 90 inequivalent two-qutrit Cliffords whenever a lies in [4/9,2/3], and show that the recurrence a_{j+1}=(2-a_j)/3 keeps the parameter inside that interval (SM §S4, Prop. 1, Lem. 3, Thm. 1). They further demonstrate that the same unitary maps the Haldane SPT phase onto a phase with spontaneously broken Z_2 symmetry generated by X=∏ e^{iπ S^x_j}, distinct from the full Z_2×Z_2 breaking uncovered by the Kennedy–Tasaki transformation, and support the claim with order-parameter calculations and explicit edge-spin actions.","tokens_in":39275,"tokens_out":828,"duration_ms":6420,"significance":"The work supplies a rare analytic understanding of why a particular Clifford circuit emerges as the optimal disentangler inside the CAMPS-DMRG ansatz, rather than treating the circuit as a purely numerical black box. The optimality proof for AKLT is self-contained (canonical-form propagation + majorization over the 90 Clifford classes) and the numerical recovery of the same circuit on Heisenberg and BLBQ is consistent. The partial Z_2 SSB characterization, verified both by long-range order of S^z and by the action of X and Z_R on edge spins, offers a new unitary route from SPT order to symmetry breaking that is distinct from the Kennedy–Tasaki map. These results strengthen the physical interpretability of hybrid Clifford–tensor-network methods and are of clear interest to the quantum many-body and tensor-network communities.","major_comments":[],"minor_comments":[{"comment":"The restriction of optimality to sequential two-site Clifford gates that must start from an open boundary is stated clearly in SM §S4 and the End Matter, but a short explicit sentence in the main-text Discussion would help readers who do not consult the supplement.","section":"Discussion"},{"comment":"Figure 2 insets showing E(N-1)=0 are useful; a brief cross-reference to Lemma 1 of the SM in the caption would make the vanishing immediately transparent.","section":"Fig. 2"},{"comment":"The comparison with the Kennedy–Tasaki transformation is accurate, yet a single sentence noting that U_KT is non-Clifford (already shown in SM §S6.C) would sharpen the contrast for readers of the main text.","section":"Characterization of the KW-transformed Haldane phase"},{"comment":"A few typographical inconsistencies appear in the arXiv version (e.g., “eH” versus “\\tilde H”, occasional missing spaces around operators); a light copy-edit pass would improve readability.","section":null}],"recommendation":"accept","confidential_remarks":"The manuscript is technically solid and the analytic optimality proof is a genuine strength. The sequential two-site restriction is essential but openly acknowledged and natural for CAMPS-DMRG; I do not regard it as a load-bearing flaw. Fit for a high-quality quant-ph or condensed-matter journal is good."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The real news is twofold. First, they extend CAMPS-DMRG to qutrits and show that the optimized local disentanglers for the Haldane phase are the sequential product of U_j,j+1 = X_{j+1}^2 U_SUM, i.e., the generalized Kramers–Wannier circuit. Second, they prove this is optimal for the AKLT state under left-to-right sweeps and show that the same map sends the SPT phase to a pure Z2 SSB phase generated by X = ∏ e^{iπ S^x_j}, not the full Z2×Z2 that Kennedy–Tasaki uncovers.\n\nWhat they do well is the analytic part. They introduce a (u,v;a)-canonical form for the site tensors, show that U_j,j+1 preserves it while updating a → (2-a)/3, and prove by majorization of the Schmidt spectra that this gate is minimal among the 90 inequivalent two-qutrit Cliffords whenever a stays in [4/9,2/3]. The recurrence keeps it there, so the whole sweep is optimal (SM §S4, Prop. 1, Lem. 3, Thm. 1). Numerics on Heisenberg and BLBQ recover the same circuit and give a clear energy and entanglement advantage; the order-parameter calculation for the transformed AKLT state is clean. The comparison with KT is careful: both expose the global X, but KW realizes the second edge action only by a right-boundary operator, so only one bulk Z2 is broken.\n\nThe soft spot is real but openly stated: optimality holds only inside the sequential two-site Clifford ansatz that must start from an open boundary. They themselves show that no two-site Clifford reduces entanglement between two bulk AKLT tensors. That is the natural setting of CAMPS-DMRG, so it does not undercut the claim as formulated; it just means the result is about this class of disentanglers, not about arbitrary unitaries.\n\nMath and data look solid; citations are appropriate. This is for people who care about tensor-network simulation of SPT phases or about circuit-level dualities. It deserves a serious referee. I would engage.","headline":"Solid analytic result: sequential two-site Clifford disentangling of the Haldane phase recovers the generalized KW map, proved optimal for AKLT and shown to expose only a single Z2 of the hidden SSB, distinct from KT.","tokens_in":39840,"tokens_out":560,"would_cite":true,"duration_ms":5921,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Optimized Clifford disentanglers for the Haldane phase are the generalized Kramers–Wannier map, turning SPT order into Z2 symmetry breaking.","keywords":["Haldane phase","AKLT state","Clifford circuits","CAMPS","Kramers-Wannier transformation","symmetry-protected topological order","Z2 spontaneous symmetry breaking","matrix product states"],"falsifier":"Show that some other two-site Clifford gate (or a non-sequential arrangement of Clifford gates) yields strictly lower entanglement entropy than U_KW for the exact AKLT MPS under the same left-to-right sweep, or that the long-distance correlator of S^z after the KW map fails to approach 1/4.","tokens_in":39938,"feed_emoji":"⚛️","tokens_out":678,"duration_ms":5598,"temperature":0.7,"pith_summary":"Disentangling circuits make quantum many-body ground states cheaper to simulate classically by stripping out entanglement before a matrix-product-state representation is built. This paper asks what those circuits actually look like for the Haldane phase of spin-1 chains, a textbook symmetry-protected topological phase. Extending the Clifford-augmented DMRG method to qutrits, the authors find that the gates selected by the optimizer are precisely the successive factors of the generalized Kramers–Wannier transformation. For the exactly solvable AKLT point they prove that this choice is optimal among all two-site Clifford gates when the circuit is built sequentially from an open boundary. The same unitary maps the Haldane phase onto a phase with spontaneous Z2 symmetry breaking, revealing a hidden order that is distinct from the better-known Kennedy–Tasaki duality. The result therefore supplies both a practical speed-up for classical simulation and a new circuit-level bridge from topological order to ordinary symmetry breaking.","feed_headline":"Clifford disentanglers for Haldane phase are Kramers–Wannier","feed_subtitle":"They cut entanglement and turn topological order into ordinary Z2 symmetry breaking.","key_machinery":"The generalized Kramers–Wannier unitary U_KW formed by the ordered product of two-site gates U_j,j+1 = X_{j+1}^2 U_SUM; it simultaneously minimizes bipartite entanglement of the AKLT (and Haldane) MPS and conjugates the Haldane Hamiltonian into a local model whose only bulk on-site product symmetry is a spontaneously broken Z2.","core_discovery":"Within the sequential two-site Clifford disentangling ansatz, the product of gates U_j,j+1 = X_{j+1}^2 U_SUM is an optimal local disentangler for the Haldane phase; it is rigorously optimal for the AKLT state and maps that phase onto a Z2 spontaneous-symmetry-breaking phase generated by the global operator X = ∏ e^{iπ S^x_j}.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Clifford disentanglers for Haldane implement generalized KW","Optimal local Clifford disentangler in Haldane is KW map","KW via Clifford circuits maps Haldane SPT to Z2 breaking","Generalized KW optimal for AKLT in Clifford CAMPS-DMRG","Clifford disentanglers turn Haldane order into Z2 SSB"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"The optimality claim holds only inside the restricted class of sequential two-site Clifford gates applied from an open boundary; no two-site Clifford gate can reduce entanglement between two bulk AKLT tensors.","fun_headline_variants_meta":{"raw":{"variants":["Clifford disentanglers for Haldane implement generalized KW","Optimal local Clifford disentangler in Haldane is KW map","KW via Clifford circuits maps Haldane SPT to Z2 breaking","Generalized KW optimal for AKLT in Clifford CAMPS-DMRG","Clifford disentanglers turn Haldane order into Z2 SSB"]},"model":"grok-4.5","effort":"low","cost_usd":0.004268,"raw_usage":{"total_tokens":1204,"prompt_tokens":739,"num_sources_used":0,"completion_tokens":93,"cost_in_usd_ticks":42680000,"prompt_tokens_details":{"text_tokens":739,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":372,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":739,"tokens_out":93,"duration_ms":3840,"temperature":1.0,"reasoning_tokens":372,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-11T22:51:03.383981+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Show that some other two-site Clifford gate (or a non-sequential arrangement of Clifford gates) yields strictly lower entanglement entropy than U_KW for the exact AKLT MPS under the same left-to-right sweep, or that the long-distance correlator of S^z after the KW map fails to approach 1/4.","supporting_citations":[],"review_version":1}