{"id":"d9a36884-a1cd-4c57-8100-913feb84f470","arxiv_id":"2501.04915","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In a 2D Hubbard model with spin-dependent nearest-neighbor interactions favoring equal-spin alignment, fourth-order perturbation theory finds ferromagnetic plaquette fluctuations that reshape the Fermi surface into quasi-1D bands and make equal-spin triplet p-wave pairing the leading pairing…","lead":"This paper studies a 2D Hubbard model with nearest-neighbor interactions that attract equal spins and repel opposite spins. It finds that these interactions create short-lived ferromagnetic plaquettes, which flatten the electronic bands, trigger a Lifshitz transition, and favor equal-spin triplet p-wave pairing.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fourth-order truncation at U/t up to 4.75, βt=5 lacks convergence evidence; the Lifshitz transition and quasi-1D bands may be truncation artifacts.","rationale":"The reader's weakest assumption is correct: the fourth-order truncation is the most load-bearing condition for the paper's central claim. The single-particle spectral functions and the phase boundary in Fig. 6 are extracted from ε_k+ReΣ(k,0) at fourth order. At half-filling, the bare density of states is logarithmically enhanced by the van Hove singularity, and at βt=5 the effective coupling U·N(0) reaches order one for U/t≈3–5, so a truncated bare series is not obviously controlled. The paper's statement (Sec. IV) that higher-order corrections are minimal is an assertion without supporting evidence. The Fermi-surface reconstruction is the paper's main novel result, so if the 5th-order contribution materially changes ReΣ(k,0), the quasi-1D bands and the Lifshitz transition could disappear. I agree with the reader's assessment that CONDITIONAL is the appropriate verdict, pending a convergence check.","tokens_in":18885,"tokens_out":6392,"duration_ms":64299,"concrete_test":"Using the same AMI machinery, compute the full fifth-order self-energy coefficients for the (U, Vuu, Vud) parameters of Figs. 4–6 at βt=5. Evaluate the zero-frequency self-energy Σ(5) at U/t=3.75 and 4.75 at the X and M points, and compare to the fourth-order contribution. If |Σ(5)/Σ(4)| exceeds ~0.3, or if adding Σ(5) shifts the zeros of ε_k+ReΣ(k,0) or changes their number along X→M, the claimed Lifshitz transition is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of a Lifshitz transition at U/t=3.75–4.75 rests on the zero-frequency renormalized dispersion ε_k+ReΣ(k,0) computed from a bare fourth-order perturbation series. At half-filling the non-interacting DOS is enhanced by the van Hove singularity (cut off only by T=0.2t), giving U·N(0) ~ 1.4 at U=4.75t; the expansion has no small parameter. The Discussion (Sec. IV) asserts that 'the fourth-order expansion has minimal higher-order corrections,' but no 5th-order estimate, resummation, or non-perturbative comparison is provided. If higher-order terms are significant, the 'additional poles' along X–M (Fig. 5) that define the Lifshitz transition and the quasi-1D bands (Fig. 4) could be truncation artifacts. This is the most load-bearing uncertainty: the paper's headline physics is directly read off from the sign changes of a fourth-order quantity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a two-dimensional single-band extended Hubbard model with nearest-neighbor interactions that are spin-differentiated: the equal-spin coupling Vuu and the opposite-spin coupling Vud are treated as independent parameters. Using Algorithmic Matsubara Integration, the authors compute the self-energy, static spin susceptibility, and pairing susceptibilities to fourth order in U and Vuu at βt=5 in the thermodynamic limit. For Vuu=-0.2U and Vud=0.2U they identify competing q=(0,0), q=(π,0), and q=(π,π) spin fluctuations, interpret the real-space susceptibility as 2×2 ferromagnetic plaquettes arranged in plaquette-antiferromagnetic or plaquette-stripe patterns, and argue that kinetic frustration produces a bandwidth reduction and a Lifshitz transition at U≈3.75t. They also report equal-spin triplet p-wave pairing as the leading pairing symmetry in the weak-coupling limit. The paper's central assertion is that finite-range magnetic fluctuations can reconstruct the Fermi surface and stabilize triplet pairing without breaking translational symmetry.","tokens_in":19118,"tokens_out":9911,"duration_ms":92958,"significance":"If quantitatively reliable, the paper would provide a new route to Fermi-surface reconstruction and equal-spin triplet pairing from spin-dependent nonlocal interactions, and it would demonstrate the usefulness of AMI for high-order diagrammatic computations in the thermodynamic limit. The symbolic disentangling of powers of U and Vuu, the full momentum resolution, and the parameter-space scan leading to the phase boundary in Fig. 6 are valuable technical contributions, and the phase boundary is a concrete falsifiable prediction for non-perturbative methods. However, the central claims are not yet established: the fourth-order truncation is uncontrolled at the parameter values used, the Lifshitz criterion is qualitative, the pairing susceptibility omits self-energy diagrams, and the equal-spin triplet pairing is to a large extent a direct consequence of the attractive Vuu input. The paper is therefore a promising but incomplete contribution.","major_comments":[{"comment":"The Lifshitz transition and the quasi-1D band structure are read off from the fourth-order renormalized dispersion ε_k + ReΣ(k,0) at βt=5. At half-filling the noninteracting density of states is enhanced by the van Hove singularity, and at U/t=3.75-4.75 the expansion has no small parameter. The Discussion states that 'the fourth-order expansion has minimal higher-order corrections,' but no fifth-order estimate, resummation, or non-perturbative comparison is provided. Because the additional poles along X-M that define the transition are sign changes of a fourth-order quantity, the phase boundary in Fig. 6 and the quasi-1D bands in Fig. 4 may be truncation artifacts. The authors should add a convergence test or restrict the central claim to parameter regions where the perturbation series is controlled.","section":"Section III.C and IV, Figs. 4-6"},{"comment":"The phase boundary is defined by 'the formation of at least two additional poles next to k_an along the X→M line' in ε_k+ReΣ(k,0). This criterion is not quantitative, and no error bars are shown for the stochastic momentum integrations or for the regularization parameter γ used in analytic continuation. A different choice of γ or pole-counting rule could shift the boundary substantially. The paper should specify a reproducible algorithmic criterion for the zero-frequency spectral function (for example, a topological index or a threshold on spectral weight) and report error estimates for the boundary.","section":"Section III.C, Fig. 6"},{"comment":"The claim that equal-spin triplet p-wave pairing 'emerges' from the induced ferromagnetic plaquette is undermined by the model construction. The Hamiltonian in Eq. (3) contains an attractive nearest-neighbor equal-spin interaction Vuu<0, and the momentum-space coupling Wuu(q)=Vuu[2cos(qx)+2cos(qy)] in Eq. (4) projects directly onto the odd-parity p-wave channel at tree level. The attractive Sz=±1 p-wave response in the U→0+ limit shown in Fig. 9(a) is therefore a direct consequence of the bare Vuu, not of plaquette fluctuations. To support the emergence claim, the authors should separate the bare-Vuu contribution from the fluctuation-mediated part, or compare with a model in which the equal-spin attraction is generated dynamically rather than put in by hand.","section":"Section III.D, Eq. (11) and Fig. 9"},{"comment":"The pairing susceptibility is computed from vertex diagrams only, without self-energy insertions or quasiparticle-weight corrections. Given that the same paper finds strong momentum-dependent ReΣ and ImΣ at the parameters of interest, the vertex-only pairing response may not reliably determine the leading symmetry, especially near the Lifshitz transition where spectral weight is strongly redistributed. The authors should include the full set of fourth-order diagrams or provide a quantitative justification for omitting self-energy corrections.","section":"Section III.D, text after Eq. (11)"}],"minor_comments":[{"comment":"The name 'Lifshitz' is repeatedly misspelled as 'Lifshiftz' (for example, in Figs. 4-6 and Section III.C).","section":"Throughout"},{"comment":"The displayed series writes the second interaction power as V^i_uu, but the text describes coefficients a[i,j] in powers of U^i and V^j_uu; the notation should be made consistent.","section":"Eq. (6)"},{"comment":"The sentence 'This behavior is consistent with is expected in FM and AFM-like interactions' is ungrammatical and should be rewritten.","section":"Section III.A, text near Fig. 1(b)"},{"comment":"The caption contains 'upper brach' and should read 'upper branch', and the fitting procedure for extracting A and B should be stated explicitly in the main text.","section":"Section III.C, Fig. 7 caption"},{"comment":"The symbols \\bar{\\epsilon} for the plaquette-model dispersion and \\tilde{\\epsilon} for the renormalized dispersion are visually similar; using distinct notation would improve readability.","section":"Section III.B, Eqs. (8)-(9)"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically interesting and the AMI-based computations are a strength, but the headline claims outrun the evidence. The triplet pairing is essentially built into the attractive Vuu input, and the Lifshitz transition currently rests on an uncontrolled fourth-order truncation with a qualitative pole-counting criterion. I would be willing to consider a revised version that adds a convergence analysis, a quantitative Lifshitz criterion, and a reframed pairing discussion. I do not see a misconduct concern; the limitation in Sec. IV is merely asserted rather than demonstrated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a fourth-order diagrammatic study of the 2D Hubbard model with spin-dependent nearest-neighbor interactions (Vuu and Vud). The technical core—AMI-based enumeration of 571 self-energy and 4537 susceptibility diagrams—is solid, and the paper is clearly written. What is new: the specific phase diagram for Vuu<0, the interpretation in terms of 2x2 ferromagnetic plaquettes, and the report of a Lifshitz transition to quasi-1D bands at U/t ~ 3.75–4.75. The connection between the susceptibility peaks and the real-space spin textures is nicely done.\n\nBut the headline results have a load-bearing weakness. The Lifshitz transition is identified from zeros (additional poles) of epsilon_k + ReSigma(k,0), computed at fourth order with U/t up to 4.75 at beta t=5. That is not a controlled expansion: at half-filling the van Hove enhancement puts U*N(0) ~ 1.4, and there is no small parameter. The authors assert 'minimal higher-order corrections' in Sec. IV but give no 5th-order estimate, no resummation, and no comparison to a non-perturbative method. If higher orders are significant, the extra poles that define the transition could be artifacts. This is exactly the stress-test concern, and it holds up on reading.\n\nThe triplet p-wave result is also less 'emergent' than the abstract suggests. With Vuu<0, equal-spin NN attraction is put into the Hamiltonian by hand; that immediately projects onto the odd-parity triplet channel. Fig. 9a shows the p-wave channel attractive at U→0+, which is tree-level input, not a fluctuation-driven effect. The authors do note the effective attraction, but they frame it as a consequence of plaquettes, which is at least partly backwards.\n\nMinor but real: the pairing susceptibility includes only vertex diagrams, and the Lifshitz criterion is based on hand-counted poles at gamma=0.2, which is not small relative to the spectral width at beta t=5.\n\nWhat the paper does well: the diagrammatic machinery is a legitimate advance for this class of model, and the qualitative picture—kinetic frustration from short-range FM fluctuations—is plausible and testable. The authors are honest about wanting non-perturbative confirmation.\n\nWho should read it: people working on extended Hubbard models, interaction-driven band reconstruction, and equal-spin pairing. It deserves a serious referee; the results would be much stronger if the authors added even a single convergence check (5th-order diagram or a comparison to DCA/QMC at moderate U). I'd treat the Lifshitz transition as suggestive, not established.","headline":"Careful fourth-order diagrammatics, but the headline Lifshitz transition and triplet p-wave rest on an uncontrolled truncation and a partially built-in interaction.","tokens_in":19641,"tokens_out":3370,"would_cite":false,"duration_ms":31793,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B20","82D55","81V70"],"pacs":["71.10.Fd","74.20.Mn","75.30.Fv"],"model":"deepseek-v4-flash","headline":"A spin-differentiated nearest-neighbor interaction that favors parallel spins creates short-ranged ferromagnetic plaquettes; these reconstruct the Fermi surface into quasi-one-dimensional bands and make equal-spin triplet p-wave pairing…","keywords":["Hubbard model","extended Hubbard model","spin-dependent interaction","Lifshitz transition","ferromagnetic plaquette","triplet p-wave pairing","spin susceptibility","diagrammatic perturbation theory"],"falsifier":"Evaluate the zero-frequency spectral function at $U/t = 4.0$, $V_{uu} = -0.2U$, $V_{ud} = 0.2U$, and $\\beta t = 5$ with a numerically exact many-body method in the thermodynamic limit or on large lattices; if the additional poles along the $X$--$M$ line and the quasi-1D band structure do not appear, the predicted Lifshitz transition is an artifact of the fourth-order expansion.","tokens_in":18694,"feed_emoji":"🧲","tokens_out":10175,"duration_ms":83009,"temperature":0.7,"pith_summary":"The paper aims to show that a nearest-neighbor interaction which treats parallel and antiparallel spins differently can, by itself, create short-ranged ferromagnetic plaquettes in the square-lattice Hubbard model. These plaquettes frustrate electron hopping and progressively flatten the single-particle band, eventually reconstructing the Fermi surface into quasi-one-dimensional bands at moderate interaction strength, a Lifshitz transition. Inside the same plaquettes, the interaction supplies an effective attraction between equal spins, making equal-spin triplet p-wave pairing the leading superconducting channel in the weak-coupling limit. A sympathetic reader would care because it offers a route to Fermi-surface reconstruction and spin-triplet superconductivity that is driven by interactions alone, without geometric frustration or broken translational symmetry.","feed_headline":"Spin-split interaction drives Lifshitz transition and p-wave pairing","feed_subtitle":"Nearest-neighbor spin-polarized interaction triggers a Lifshitz transition and equal-spin triplet p-wave pairing.","key_machinery":"The object that carries the argument is the spin-differentiated nearest-neighbor interaction, written in momentum space as $W_{uu}(q) = V_{uu}[2\\cos q_x + 2\\cos q_y]$ and $W_{ud}(q) = U[1 + (V_{ud}/U)(2\\cos q_x + 2\\cos q_y)]$. The computational engine is fourth-order direct diagrammatic perturbation theory evaluated by Algorithmic Matsubara Integration, which performs Matsubara sums symbolically and leaves internal momenta continuous, so results are in the thermodynamic limit. The physical mechanism is the $2\\times2$ ferromagnetic plaquette: Pauli exclusion blocks hopping between aligned spins, and the alternating (PAFM) or stripe (PS) arrangement of plaquettes yields effective dispersions $\\bar{\\epsilon} = -2[\\cos(k_x)+\\cos(k_y)]/D$ and $\\bar{\\epsilon} = \\pm 2\\cos(k)/D$, respectively. The quasi-1D band observed at $U/t \\ge 3.75$ is exactly the PS dispersion with $D = 2$, and the effective attraction between equal spins inside the plaquette channels the pairing response into equal-spin triplet p-wave.","core_discovery":"The central discovery is that with $V_{uu} < V_{ud}$, specifically $V_{uu} = -0.2U$ and $V_{ud} = 0.2U$ at $\\beta t = 5$, the static spin susceptibility develops a strongly enhanced $\\mathbf{q} = (0,0)$ ferromagnetic mode along with collinear $\\mathbf{q} = (\\pi,0)$ and staggered $\\mathbf{q} = (\\pi,\\pi)$ modes, whose real-space form is a two-by-two ferromagnetic plaquette that decays into stripe and checkerboard patterns. The self-energy renormalization flattens the zero-frequency dispersion; at $U/t = 3.75$ additional poles form along the $X$--$M$ line, and the Fermi surface becomes two quasi-one-dimensional bands whose fitted weights $A \\approx 1.0$, $B = 0$ match the effective dispersion of plaquette-stripe order with plaquette size $D = 2$. The uniform pairing susceptibility computed to fourth order shows that equal-spin $S_z = \\pm 1$ $p_x + ip_y$ p-wave pairing is attractive already in the $U \\to 0^+$ limit and remains leading until singlet $d$-wave pairing turns attractive near $U/t = 3.5$. The paper concludes that finite-range magnetic fluctuations, acting through kinetic frustration, can restructure both single-particle and pairing properties while translational symmetry remains intact.","pith_inferences":["I infer that the same plaquette-frustration mechanism should survive at finite doping, but the Fermi-surface nesting will change, shifting both the Lifshitz threshold and the momentum structure of the quasi-1D bands; a doping scan would map this out.","Extending the ferromagnetic bias to next-nearest neighbors or beyond should grow the plaquettes; by the paper's own $D$-scaling, larger plaquettes flatten the bands further and may turn the equal-spin pairing into higher angular-momentum channels such as $f$-wave.","A direct testable extension is the dynamical spin susceptibility: the $\\mathbf{q} = (0,0)$ and $\\mathbf{q} = (\\pi,0)$ modes should appear at distinct energies, giving an experimental fingerprint for the plaquette fluctuations in spectroscopic probes."],"forward_implications":["At $V_{uu} = -0.2U$ the Lifshitz transition occurs at $U/t = 3.75$; strengthening the ferromagnetic bias to $V_{uu} = -0.4U$ lowers the threshold to $U/t = 2.70$, so the transition is tunable by the interaction ratio.","Above the transition the hopping along one axis is effectively frozen, with fitted weights $A \\approx 1.0$ and $B = 0$, exactly the $D = 2$ plaquette-stripe dispersion, so the quasi-1D bands are a quantitative signature of the plaquette fluctuation.","Equal-spin triplet p-wave pairing is attractive in the weak-coupling limit and competes with singlet $d$-wave at intermediate coupling, with $d$-wave becoming the leading channel near $U/t = 3.5$.","Reversing the spin polarization ($V_{uu} > V_{ud}$) removes the Lifshitz transition and instead enhances $(\\pi,\\pi)$ magnetic fluctuations and $d$-wave pairing, so the sign of $V_{uu} - V_{ud}$ controls which physics emerges.","All results are obtained at finite temperature in the thermodynamic limit without spontaneously broken translational symmetry, consistent with the Mermin-Wagner theorem."],"supporting_citations":[{"why":"Supplies the Algorithmic Matsubara Integration method used to resolve all Matsubara-frequency integrals in the diagrammatic expansions.","marker":"[44]"},{"why":"Provides the symbolic multi-indexed diagrammatic expansion that lets the paper factor powers of U and Vuu and scan interaction space.","marker":"[46]"},{"why":"Shows that renormalized local interactions generate spin-dependent nonlocal interactions, motivating the Vuu < Vud parameter regime.","marker":"[43]"},{"why":"Gives the baseline pairing susceptibility of the plain 2D Hubbard model, against which the new leading p-wave channel is identified.","marker":"[60]"},{"why":"Supplies the Mermin-Wagner theorem, used to frame why the plaquette textures are finite-temperature fluctuations rather than broken-symmetry order.","marker":"[47]"},{"why":"Provides the plaquette antiferromagnetic and plaquette-stripe spin configurations used to classify the susceptibility patterns and effective dispersions.","marker":"[39]"},{"why":"Defines the singlet/triplet pairing symmetry classification used for the pairing susceptibility channels.","marker":"[59]"}],"fun_headline_variants":["Spin-polarized interactions drive Lifshitz transition and p-wave pairing","Ferromagnetic plaquettes induce p-wave pairing and Fermi surface reshaping","Spin-split Hubbard model shows triplet p-wave pairing without frustration","Magnetic fluctuations produce p-wave pairing and Lifshitz transition","Fourth-order theory reveals p-wave pairing from spin-dependent interactions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that fourth-order truncated perturbation theory is quantitatively reliable at $\\beta t = 5$ up to $U/t = 4.75$; the paper asserts that higher-order corrections are minimal but supplies no convergence test, so the additional poles that define the Lifshitz transition could be truncation artifacts if higher-order diagrams shift the zero-frequency self-energy.","fun_headline_variants_meta":{"raw":{"variants":["Spin-polarized interactions drive Lifshitz transition and p-wave pairing","Ferromagnetic plaquettes induce p-wave pairing and Fermi surface reshaping","Spin-split Hubbard model shows triplet p-wave pairing without frustration","Magnetic fluctuations produce p-wave pairing and Lifshitz transition","Fourth-order theory reveals p-wave pairing from spin-dependent interactions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000243,"raw_usage":{"total_tokens":1619,"prompt_tokens":1129,"completion_tokens":490,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":745,"completion_tokens_details":{"reasoning_tokens":403}},"tokens_in":745,"tokens_out":490,"duration_ms":4723,"temperature":1.0,"reasoning_tokens":403,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:22:29.476398+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the zero-frequency spectral function at $U/t = 4.0$, $V_{uu} = -0.2U$, $V_{ud} = 0.2U$, and $\\beta t = 5$ with a numerically exact many-body method in the thermodynamic limit or on large lattices; if the additional poles along the $X$--$M$ line and the quasi-1D band structure do not appear, the predicted Lifshitz transition is an artifact of the fourth-order expansion.","supporting_citations":[{"cited_title":"Gazizova and J","cited_arxiv_id":null,"evidence_quote":"Supplies the Algorithmic Matsubara Integration method used to resolve all Matsubara-frequency integrals in the diagrammatic expansions."},{"cited_title":"Taheridehkordi, S","cited_arxiv_id":null,"evidence_quote":"Provides the symbolic multi-indexed diagrammatic expansion that lets the paper factor powers of U and Vuu and scan interaction space."},{"cited_title":"Terletska, S","cited_arxiv_id":null,"evidence_quote":"Shows that renormalized local interactions generate spin-dependent nonlocal interactions, motivating the Vuu < Vud parameter regime."},{"cited_title":"Farid and J","cited_arxiv_id":null,"evidence_quote":"Supplies the Mermin-Wagner theorem, used to frame why the plaquette textures are finite-temperature fluctuations rather than broken-symmetry order."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the plaquette antiferromagnetic and plaquette-stripe spin configurations used to classify the susceptibility patterns and effective dispersions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the singlet/triplet pairing symmetry classification used for the pairing susceptibility channels."}],"review_version":1}