{"id":"da11321a-611c-4a62-a6c1-6f627bc3bda5","arxiv_id":"2608.09689","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The σt-J model is found to keep its spin pseudogap and spin-sector BKT transition almost doping-independent, which a slave-fermion mean-field theory with a PSG-selected ansatz explains as spin-charge separation.","lead":"An altered version of the t-J model of doped magnets, the σt-J model, shows a hole-induced spin pseudogap and a Berezinskii-Kosterlitz-Thouless magnetic transition that barely move as holes are added. A tensor-network study plus a slave-fermion mean-field theory suggests this modified model may be a clean theoretical testing ground for spin-charge separation in doped Mott insulators.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Loop-TNR BKT identification rests on unquantified bond-dimension truncation; without a higher-D or stiffness check, T_BKT≈0.33 and its doping robustness are not yet established.","rationale":"The reader's weakest_assumption and my read converge on the same load-bearing concern: the spin-sector BKT claim is the least secure pillar of the central claim. The paper is transparent about bond dimensions and explicitly states that the data 'should not be regarded as a high-precision determination' (Sec. III), which is creditable. The T* pseudogap part is supported by iPEPS susceptibility and specific-heat curves, and the mean-field theory is a parameter-free, self-consistent construction that independently reproduces the enhanced xy antiferromagnetism and the T* scale; those parts deserve credit. However, the BKT pillar relies on heavily truncated partition-function tensors and CFT data read at RG step 10, with no resolved ξ divergence and no measurement of the Nelson-Kosterlitz universal jump. This is a correctable numerical limitation rather than an internal inconsistency, so CONDITIONAL remains the appropriate verdict; the proposed check of doubling D_TNR directly tests whether the c≈1 plateau is a truncation artifact.","tokens_in":33239,"tokens_out":8289,"duration_ms":80122,"concrete_test":"At δ=0.12, rerun the loop-TNR central-charge analysis with D_TNR=48 (and D_pre=96), extracting c and the compactified-boson spectrum over RG steps 5–15; if T_BKT, defined by the low-T c≈1 boundary, shifts by more than 0.05 from the reported 0.33, or if no stable plateau survives, the claimed doping-independent spin BKT transition rests on the truncation and is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline claim that T_BKT≈0.33 is stable for 0.05≲δ≲0.15 depends on a loop-TNR central-charge analysis whose numerical input is heavily truncated: the iPEPO is first produced by NTU with D=12 (Sec. III), then its virtual bond dimension is reduced from D²=144 to D_pre=48 via the projectors in Eq. (16), and the loop-TNR flow is run at D_TNR=24 with conformal data read at RG step 10 (Figs. 6–7). No truncation errors are reported, and the Appendix A benchmark is the pure XXZ model, which does not test the projector step in a doped fermionic setting. The only independent supporting evidence is the algebraic-exponent fit in Eq. (23) over r=2–7, with no resolved correlation-length divergence near T_BKT (Fig. 8). A U(1)-symmetric finite-temperature tensor network can exhibit an extended c≈1 quasi-long-range window even when the actual transition temperature is shifted or absent; without a stiffness measurement or a converged D_TNR check, T_BKT≈0.33 and its weak doping dependence are not pinned down. If this BKT identification fails, the 'robust BKT upon doping' part of the central claim collapses, even though the T* pseudogap result may survive.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the finite-temperature properties of the σt-J model, a t-J variant with spin-dependent hopping designed to eliminate the phase-string sign frustration, using finite-temperature iPEPS/iPEPO methods (NTU) and loop-TNR. It reports a spin susceptibility maximum at T*≈0.9J and a broad specific-heat peak, and interprets a low-temperature c≈1 compactified-boson regime found by loop-TNR as a spin-sector BKT transition at T_BKT≈0.33J. A central claim is that both scales remain nearly doping independent for 0.05≲δ≲0.15, in contrast to the t-J model and its easy-plane variants, and that doped holes enhance xy-plane antiferromagnetic correlations. The paper then constructs a slave-fermion mean-field theory, with the PSG class selected from zero-temperature correlation data of Ref. [22], and shows that self-consistent solutions yield doping-enhanced in-plane order, a U(1) order-parameter manifold consistent with BKT physics, and a pseudogap-like magnetic response governed by the RVB pairing scale.","tokens_in":33428,"tokens_out":5692,"duration_ms":55684,"significance":"If the numerical identification holds, the σt-J model is a valuable controlled companion to the t-J model: it removes the phase-string interference while retaining pseudogap-like thermodynamics, and it admits a concrete slave-fermion mean-field theory whose parameters are solved self-consistently rather than fitted. The paper is commendably explicit about its algorithmic setup, including fermionic tensor conventions, the NTU environment, bond dimensions, and the limitation that the data are not high-precision. The Appendix A benchmarks against the XXZ model and the Heisenberg transient flow are useful checks on the loop-TNR machinery. However, the headline BKT claim currently rests on heavily truncated tensor-network data without convergence checks or error bars, so the significance of the paper is conditional on strengthening this evidence.","major_comments":[{"comment":"The BKT identification is the load-bearing step for the claim that T_BKT≈0.33 is robust upon doping, but it currently rests on truncated tensor-network data without convergence checks. The input iPEPO is obtained with D=12 and then projected from D²=144 to D_pre=48 via Eq. (16); loop-TNR is run at D_TNR=24, and the c≈1 plateau and compactified-boson spectrum are read at RG step 10 (Figs. 6–7), i.e., over only about five RG steps. No D_TNR or D_pre dependence is reported, and no error bars are given for c, R, or the scaling dimensions. The Appendix A benchmark is the pure XXZ model, which validates the loop-TNR machinery for a bosonic model but does not test the projector truncation in a doped fermionic model. The independent check in Eq. (23) fits η over r=2–7 and does not resolve the expected ξ divergence near T_BKT (Fig. 8). Because a truncated U(1)-symmetric tensor network can show an extended c≈1 window even when the true transition temperature is shifted, I ask for a helicity-modulus/stiffness measurement or an equivalent universal-jump diagnostic, plus at least one convergence test (e.g., D_TNR=32 and D_pre=64 at δ=0.12), before T_BKT≈0.33 and its doping robustness can be considered established.","section":"Section III C, Figs. 6–8, Eq. (23), Appendix A"},{"comment":"The weak-doping-dependence claim for T* is quantitative, but the susceptibility data are taken at a single bond dimension D=12 with probe field h=0.1, and T* is extracted from fourth-order polynomial fits and numerical derivatives without uncertainty estimates. The paper itself states in Section III that the data 'should not be regarded as a high-precision determination.' Because T*≈0.9 with only a mild decrease at δ=0.20 is one of the two headline scales, I request at least one D-convergence check (e.g., D=16 at δ=0.10 and 0.15) and an h→0 check at one doping to confirm that the maximum in χ is not a truncation artifact.","section":"Section III B, Fig. 4"},{"comment":"The mean-field ansatz is selected using zero-temperature hopping and pairing correlations from Ref. [22], whose author list overlaps with that of this manuscript, and the same numerical family is then presented as evidence for the resulting mean-field theory. This is not a fitting circularity—κ, χ, and Δ are solved self-consistently, and no finite-T data are used as input—but it does weaken the claim of independent confirmation. Please state this dependence explicitly and, where feasible, test the selected Z2(0,0) ansatz against an independent ground-state method (e.g., DMRG on cylinders) for the σt-J model.","section":"Section IV A and IV B"}],"minor_comments":[{"comment":"The phrase 'provides strong evidences' should be 'provides strong evidence.'","section":"Abstract"},{"comment":"The sentence introducing Fig. 7 contains a typesetting artifact ('theσt-JRG flow'); similar missing spaces appear around 'σt-Jmodel' in several places and should be corrected.","section":"Section III C"},{"comment":"The statement that 'R is the compactification radius obtained from the lowest nonzero Δ' should specify which lowest nonzero scaling dimension is used and how the modular parameter τ is fixed in the fits.","section":"Figure 7 caption"},{"comment":"The mean-field calculation sets t=1 as the energy unit while the numerical simulations use t=2, J=1; please state the conversion explicitly so the mean-field T*/J can be compared directly with Fig. 4.","section":"Section IV B"},{"comment":"The fourth-order polynomial fitting used to convert observables from fixed chemical potential to fixed doping is described, but no fit-quality measure (e.g., residuals or number of μ points) is reported; adding this would help assess the T* extraction.","section":"Section II C"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for cond-mat.str-el and addresses a timely question about doping robustness in a phase-string-free t-J variant. The main risk is the numerical identification of the BKT transition: the evidence is suggestive but not yet convergent by the paper's own standards. If the authors can supply a stiffness measurement or a D_TNR/D_pre convergence check, and preferably a D-convergence check for T*, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Finite-temperature tensor network study of the σt-J model, with a slave-fermion mean-field theory on top. The genuinely new results are the doping-robust spin susceptibility maximum T* and specific-heat peak, the loop-TNR evidence for a spin-sector BKT transition at T_BKT ≈ 0.33 J, and the PSG-constrained mean-field argument tying both to the suppressed phase string. The ground state and the phase-string cancellation already came out of Refs. [18,22]; the finite-T view is what this paper adds.\n\nWhat is done well: the mean-field is solved self-consistently, not fitted to the target curves; the PSG classification is careful and the selection of the Z2(0,0) ansatz is based on independent numerical correlations (even though that prior paper shares authors, using its data to constrain the ansatz is legitimate). The comparison with t-XX and Zeeman-field t-J is exactly the right control to separate easy-plane symmetry from hole-spin interference. The topological argument — doping collapses the spinon order-parameter manifold from S2 to U(1), which is what makes a finite-T BKT possible — is crisp and matches the numerics qualitatively. The benchmarks on XXZ and Heisenberg give me confidence the TNR pipeline is not obviously broken.\n\nThe soft spot is the BKT claim itself. The loop-TNR input is heavily truncated: iPEPS at D=12, then the iPEPO bond projected from 144 to 48, then loop-TNR at D_TNR=24, with conformal data read at RG step 10. No truncation errors are reported anywhere. The eta fit runs over r=2..7 and the expected BKT correlation-length divergence is not resolved, and the Appendix A benchmark is the pure XXZ model, which does not exercise the projector step in a doped fermionic setting. So T_BKT ≈ 0.33 and its weak doping dependence are plausible but not established. If the c≈1 plateau is an artifact of the truncation, the 'robust BKT upon doping' headline collapses; the T* story might still survive. The authors themselves say the data should not be regarded as high-precision, which is honest, but that caveat is buried in the methods rather than attached to the phase diagram.\n\nThere is also a mild circularity worth noting: the PSG ansatz is selected using correlation data from a paper by overlapping authors [22], and the same numerical family is then explained by that ansatz. I don't think it is disqualifying — using prior numerics to pick an ansatz is fine — but the independence is weaker than it appears.\n\nWho gains from this: people working on phase-string physics, doped Mott insulators, and finite-T tensor network methods. It deserves a serious referee, and a good referee will push for higher bond dimensions, a stiffness check, or a clear error estimate before T_BKT is accepted. I would send it out.","headline":"The finite-T results and PSG mean field are solid and worth refereeing; the robust-BKT-upon-doping claim needs more numerical control before it is believable.","tokens_in":34056,"tokens_out":3354,"would_cite":false,"duration_ms":31102,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A sign flip in the t-J model makes both its spin pseudogap and spin-ordering transition survive doping up to 20 percent, and a symmetry-guided mean-field theory explains why.","keywords":["sigma t-J model","spin pseudogap","Berezinskii-Kosterlitz-Thouless transition","tensor network renormalization","projective symmetry group","slave-fermion mean-field theory","spin-charge separation","phase string"],"falsifier":"Re-run the loop-TNR analysis with substantially larger bond dimensions or a truncation-free thermal TNR scheme and check whether the $c\\approx 1$ plateau persists and whether the spin correlation length $\\xi(T)$ shows the expected BKT divergence as $T$ approaches $T_{\\mathrm{BKT}}$; if the plateau disappears or $\\xi$ saturates instead of diverging, the BKT claim collapses. A separate decisive check is to measure the universal BKT jump in the spin stiffness (helicity modulus) of the $\\sigma t$-$J$ model in the doping range $0.05\\lesssim\\delta\\lesssim 0.15$.","tokens_in":32958,"feed_emoji":"🧲","tokens_out":12011,"duration_ms":105868,"temperature":0.7,"pith_summary":"The paper argues that the $\\sigma t$-$J$ model—an ordinary $t$-$J$ model with the sign of the spin-down hopping flipped, which removes the phase-string interference between doped holes and the spin background—has a spin pseudogap that survives hole doping in a way the $t$-$J$ model's does not. The uniform spin susceptibility develops a maximum at $T^*$ near $0.9J$ and the specific heat a broad peak, and both remain nearly unchanged for dopings from about $0.05$ to $0.20$. Tensor network renormalization gives evidence that the spin sector undergoes a Berezinskii-Kosterlitz-Thouless transition at $T_{\\mathrm{BKT}}$ near $0.33J$ over the same doping range, while the doped holes enhance rather than suppress the antiferromagnetic correlations in the $xy$ spin plane. A slave-fermion mean-field theory, whose ansatz is selected from projective symmetry group classes using the numerically observed hopping and pairing correlations, explains all three features: $T^*$ is the temperature at which short-range singlet pairing dissolves, and doping collapses the magnetic order-parameter manifold from a sphere to a circle, making the BKT transition possible. The work matters because it isolates which hole-spin interference effects are essential to pseudogap physics.","feed_headline":"Sign-flipped t-J model keeps spin pseudogap under doping","feed_subtitle":"The spin-pseudogap scale stays near 0.9 J, and magnetic ordering stays near 0.33 J, as holes are added up to 20 percent.","key_machinery":"The load-bearing object is the sign-flipped hopping term $H_{\\sigma t}$ together with the slave-fermion ansatz it dictates. The $\\sigma t$ term flips the spin-down hopping sign; combined with a staggered sublattice factor it forms an exact composite $Z_2$ symmetry that enforces opposite-sign nearest-neighbor spinon hopping for the two spin species. The projective symmetry group analysis leaves 32 algebraic classes, and the coexistence of nearest-neighbor hopping and pairing with next-nearest-neighbor hopping, as found numerically, selects the zero-flux $Z_2(0,0)$ class, whose mean-field pattern is uniform purely imaginary hopping and pairing on all nearest-neighbor bonds. In the self-consistent solution the spinons condense at momenta $\\pm Q$ with locked amplitudes, producing staggered magnetization in the $xy$ plane whose phase parametrizes a circle, so the doped system has the $U(1)$ order-parameter manifold needed for a BKT transition, while at zero doping both Nambu sectors condense and the manifold becomes a sphere with no finite-temperature transition. The same theory sets $T^*$ as the temperature at which the RVB pairing amplitude vanishes, which is why that scale is controlled by $J$.","core_discovery":"The paper's central claim is that the finite-temperature spin physics of the $\\sigma t$-$J$ model is robust against hole doping, in contrast to the $t$-$J$ model and its easy-plane variants. Using finite-temperature iPEPS and loop-TNR, it finds a spin-susceptibility maximum at $T^*$ close to $0.9J$ and a broad specific-heat peak that shift only weakly for $0.05\\lesssim\\delta\\lesssim 0.20$, alongside a stable $c\\approx 1$ compactified-boson regime in the renormalization-group flow that it interprets as a BKT transition at $T_{\\mathrm{BKT}}\\approx 0.33J$ in the $U(1)$-symmetric spin sector. Three-point conditional spin correlations show that holes suppress $z$-axis correlations but strengthen $xy$-plane antiferromagnetic correlations, so the spin background is preserved and even reinforced. The authors' slave-fermion mean-field theory, with fermionic holons and Schwinger-boson spinons, uses the projective symmetry group to fix the allowed ansatz and the numerical correlations to select the zero-flux $Z_2(0,0)$ class; its self-consistent solution gives a holon Fermi pocket of area $\\delta$ times the Brillouin zone, spinon condensation at momenta $\\pm Q$ producing staggered $xy$ magnetization whose moment grows under dilute doping, a $U(1)$ order-parameter manifold, and an RVB pairing amplitude that vanishes at $T^*$.","pith_inferences":["The paper does not compute the electron spectral function; if its spin-charge-separation picture holds, above $T_{\\mathrm{BKT}}$ the spinon condensate loses phase coherence and the spectral weight of the composite electron should be suppressed, producing a spectroscopic pseudogap that could be searched for in future tensor-network calculations.","The order-parameter manifold argument predicts a universal BKT jump in the spin-sector helicity modulus (spin stiffness) at $T_{\\mathrm{BKT}}$; measuring this jump at $\\delta\\approx 0.12$ would give an independent test that the paper does not perform.","The PSG-selection procedure—using numerically determined bond correlations to narrow the algebraic PSG classes to a single ansatz—is a general recipe that could be applied to other strongly correlated models where parton mean-field theories are underdetermined by symmetry alone.","Because $T^*$ is controlled purely by $J$ in the mean-field theory, varying the exchange coupling while keeping $t/J$ fixed should move $T^*$ proportionally; a quantitative comparison of susceptibility curves across $J$ values would directly test the mechanism."],"forward_implications":["The spin pseudogap of a doped Mott antiferromagnet can be a two-scale phenomenon: a crossover at $T^*\\approx J$ where short-range singlet correlations develop, and a genuine BKT transition at $T_{\\mathrm{BKT}}\\approx 0.33J$ below which only quasi-long-range $xy$-plane antiferromagnetism survives.","Doping does not inevitably destroy the spin background: in the $\\sigma t$-$J$ model, holes increase the $xy$ component of the spin correlations, so the susceptibility peak and the BKT temperature remain near $0.9J$ and $0.33J$ up to $\\delta\\approx 0.20$.","The BKT transition is a spin-sector transition, not a superconducting one, because the fixed-point tensor network becomes purely bosonic and the staggered spin correlation exponent tends toward $\\eta=1/4$ rather than toward an Ornstein-Zernike value.","The slave-fermion mean-field theory predicts a Fermi-liquid-like electron state: hole pockets of area $\\delta\\times A_{\\mathrm{BZ}}$ shifted by the spinon condensation momenta, coexisting with in-plane antiferromagnetism.","Easy-plane anisotropy alone cannot explain the robustness, since the $t$-$J$ model under a Zeeman field and the $t$-XX model still show strongly doping-dependent $T^*$ and $T_{\\mathrm{BKT}}$; suppressing the phase-string interference is the decisive ingredient."],"supporting_citations":[{"why":"Introduces the $\\sigma t$-$J$ model as a companion to the $t$-$J$ model and gives its DMRG ground-state structure, which the paper uses as the zero-temperature anchor.","marker":"[18]"},{"why":"Previous iPEPS study showing the $\\sigma t$-$J$ ground state has $xy$-plane antiferromagnetism plus a Fermi-liquid-like hole sector; the numerically observed hopping and pairing correlations select the PSG class.","marker":"[22]"},{"why":"Finite-temperature iPEPS study of the $t$-$J$ model whose purification method and pseudogap phenomenology provide the methodological baseline and the comparison point.","marker":"[32]"},{"why":"Defines the BKT transition in two-dimensional systems with continuous symmetry; the spin-sector transition in the $\\sigma t$-$J$ model is interpreted as this mechanism.","marker":"[35]"},{"why":"Provides the Schwinger-boson functional-integral mean-field formalism used for the spinon sector and the half-filled benchmark moment.","marker":"[38]"},{"why":"Establishes the projective symmetry group framework for classifying symmetric spin-liquid ansätze, used to constrain the mean fields.","marker":"[40]"},{"why":"Supplies the Schwinger-boson PSG classification on the square lattice that the paper reduces to 32 gauge-inequivalent classes.","marker":"[43]"},{"why":"Loop-TNR algorithm that removes short-range entanglement during coarse graining; the stable $c\\approx 1$ flow under this algorithm is the main evidence for the BKT regime.","marker":"[50]"},{"why":"BKT scaling forms for the $XY$/XXZ model, including $\\eta_{\\mathrm{BKT}}=1/4$, used to interpret the staggered correlation fit and identify the spin-sector transition.","marker":"[60]"}],"fun_headline_variants":["Spin pseudogap survives hole doping in σt-J model","Doping strengthens xy-spin order in sign-flipped t-J model","σt-J model: spin pseudogap stays put under doping","Spin-charge separation in σt-J model withstands hole doping"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the loop-TNR flow, run with the tensor bond dimension truncated from 144 to 48 and with conformal data read off after ten coarse-graining steps, faithfully reveals the true phase transition, so the stable scale-invariant regime is a genuine Berezinskii-Kosterlitz-Thouless transition rather than an artifact of the truncation.","fun_headline_variants_meta":{"raw":{"variants":["Spin pseudogap survives hole doping in σt-J model","Doping strengthens xy-spin order in sign-flipped t-J model","σt-J model: spin pseudogap stays put under doping","Spin-charge separation in σt-J model withstands hole doping"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00081,"raw_usage":{"total_tokens":3685,"prompt_tokens":1208,"completion_tokens":2477,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":824,"completion_tokens_details":{"reasoning_tokens":2402}},"tokens_in":824,"tokens_out":2477,"duration_ms":18750,"temperature":1.0,"reasoning_tokens":2402,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:41:10.712411+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the loop-TNR analysis with substantially larger bond dimensions or a truncation-free thermal TNR scheme and check whether the $c\\approx 1$ plateau persists and whether the spin correlation length $\\xi(T)$ shows the expected BKT divergence as $T$ approaches $T_{\\mathrm{BKT}}$; if the plateau disappears or $\\xi$ saturates instead of diverging, the BKT claim collapses. A separate decisive check is to measure the universal BKT jump in the spin stiffness (helicity modulus) of the $\\sigma t$-$J$ model in the doping range $0.05\\lesssim\\delta\\lesssim 0.15$.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Previous iPEPS study showing the $\\sigma t$-$J$ ground state has $xy$-plane antiferromagnetism plus a Fermi-liquid-like hole sector; the numerically observed hopping and pairing correlations select the PSG class."},{"cited_title":"Czarnik, L","cited_arxiv_id":null,"evidence_quote":"Finite-temperature iPEPS study of the $t$-$J$ model whose purification method and pseudogap phenomenology provide the methodological baseline and the comparison point."},{"cited_title":"Zhang, A","cited_arxiv_id":null,"evidence_quote":"Defines the BKT transition in two-dimensional systems with continuous symmetry; the spin-sector transition in the $\\sigma t$-$J$ model is interpreted as this mechanism."},{"cited_title":"Yoshioka, Slave-fermion mean field theory of the Hubbard model, J","cited_arxiv_id":null,"evidence_quote":"Establishes the projective symmetry group framework for classifying symmetric spin-liquid ansätze, used to constrain the mean fields."}],"review_version":1}