{"id":"318c6575-8552-457d-be77-3b30da0901dd","arxiv_id":"2511.03528","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"CST calculations show magnon-magnon interactions reduce the altermagnetic spin splitting by 14–20% and create a J2-dependent roton minimum in a 2D spin-1/2 altermagnet.","lead":"Continuous similarity transformations are used to compute how magnon-magnon interactions reshape the magnon spectrum of a minimal 2D altermagnet. The interactions reduce the altermagnetic spin splitting by about 14–20% and produce a roton minimum that grows with the frustrating next-nearest-neighbor coupling.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"qpc convergence criterion does not establish approximate magnon stability; the 14–20% splitting and roton depth may depend on an unresolved decay width.","rationale":"The reader's weakest-assumption identification is exactly the load-bearing point: qpc convergence within a finite energy resolution is being used to promote potentially decaying magnons to 'approximately stable' quasiparticles, and all quantitative results are extracted from that assumption. I agree with that assessment. The paper is careful and transparent about the lack of rigorous magnon stability, and the CST has prior validation on the square-lattice Heisenberg model, so this is not a claim of internal inconsistency or bad faith. However, the operational criterion — ROD < 10^-6 J1 and 1/ℓmax ≈ 0.004 J1 — is not tied to any computed decay rate. Eq. (B17) only justifies exponential suppression of off-diagonal elements when energy sectors are strictly ordered; in altermagnets, analytic results show decay channels exist, so the relevant question is quantitative: is the physical linewidth below or above 0.004 J1? The paper never answers this. This matters directly for the 14–20% spin-splitting reduction and for the roton-depth dependence on J2, both of which would be distorted if the true spectral peak has a width comparable to or larger than the stated resolution. A direct golden-rule or self-energy calculation on the same truncated model would settle the question without requiring a fundamentally new method. Until that check is performed, the appropriate verdict remains CONDITIONAL, i.e., unchanged from the reader's assessment.","tokens_in":26882,"tokens_out":6868,"duration_ms":68539,"concrete_test":"Using the same truncated Hamiltonian and identical momentum discretization, compute the one-magnon decay width from the residual 1↔3 couplings at the qpc stopping point — e.g., a golden-rule width Γ(k) = 2π Σ |V_{k,q}|² δ(ω_k^3 − ω_k^1) — for representative J2 = −0.16, 0, +0.16 at k = (π/2, π/2) and (π, 0). Compare Γ(k) with Δω ≈ 0.004 J1 from 1/ℓmax. If Γ(k) > Δω, the approximate-stability assumption fails and the CST spectral energies — hence the 14–20% reduction and the roton-depth dependence — are not reliable. If Γ(k) < Δω, the concern is resolved and the present verdict can stand.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims (Sec. 4.2.1/4.2.2) — a 14–20% CST reduction of the altermagnetic spin splitting and a J2-dependent roton minimum — hinge on interpreting a converging qpc flow as 'approximate magnon stability' within Δω = max(1/ℓmax, v_spin/L) (Sec. 1, App. B). This is not established. Eq. (B17) describes exponential suppression of off-diagonal elements only when one-magnon energies lie below three-magnon energies; for J2≠0 the altermagnetic upper branch is generically unstable (Refs. 28,30), so the exponent is positive for some states and the exact flow is divergent. The numerical procedure stops when the ROD falls below 10^-6 J1, before that divergence is visible, and then uses 1/ℓmax ≈ 0.004 J1 as the energy resolution. But the ROD is a residual off-diagonality in a truncated operator space; it does not measure the physical decay width Γ(k) of a magnon. If Γ(k) at the points used for ΔS and ΔR exceeds ~0.004 J1, the 'converged' qpc dispersion is an artifact of an incomplete generator flow: the extracted single-magnon energy is not the peak position of the true spectral function, and the reported renormalization percentages and roton depths would be wrong. No estimate of Γ(k) is provided anywhere in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a minimal spin-1/2 square-lattice Heisenberg model with inequivalent next-nearest-neighbor couplings, intended as a 2D altermagnet. The authors combine the Dyson-Maleev transformation, self-consistent spin-wave theory, and continuous similarity transformations (CST) to obtain a one-magnon effective Hamiltonian. They use the 0n and qpc generators to map the stability of the Néel phase and the approximate stability of magnons, respectively. The central quantitative claims are that magnon-magnon interactions reduce the altermagnetic spin splitting by 14--20% relative to LSWT and produce a roton minimum in the lower magnon branch whose depth depends non-monotonically on the altermagnetic NNN coupling J2. The paper also presents dynamical structure factors separated into one-, two-, and three-magnon sectors and emphasizes the possibility of spontaneous magnon decay for J2 ≠ 0, interpreting the qpc flow convergence as a criterion for approximate magnon stability.","tokens_in":27251,"tokens_out":6021,"duration_ms":59652,"significance":"If the central quantitative claims are correct, the paper provides a useful beyond-LSWT benchmark for a minimal 2D altermagnet and gives experimentally relevant predictions for inelastic neutron scattering. The CST framework is systematic, has no free parameters beyond the model couplings, and the authors are explicit about the limitations imposed by finite system size and finite flow cutoff. The separation of spectral contributions by magnon number and the mapping of qpc/0n convergence regions are valuable methodological contributions. However, the quantitative claims rest on interpreting a truncated, finite-resolution flow as evidence of approximate magnon stability; this interpretational step is not yet backed by a direct estimate of the physical single-magnon linewidth, which is essential for the 14--20% splitting reduction and the roton-depth results.","major_comments":[{"comment":"The extrapolated qpc convergence lower bound is J2,l^c ≈ (0.02±0.03)J1, i.e. essentially zero in the thermodynamic limit. This means the qpc flow is not expected to converge for any J2 < 0 as L→∞. Nevertheless, Figs. 3–5 present CST results for J2 = −0.16, −0.10, −0.05 and extrapolate them in 1/L. The finite-size convergence endpoints are not a surrogate for thermodynamic-limit convergence. The paper should either explain why the lower-bound divergence does not affect the single-magnon energies used for ΔS and ΔR, or restrict the quantitative claims to the J2 > 0 region. As written, the 14–20% reduction claim covers the negative-J2 part of the parameter range on a convergence basis that is absent in the L→∞ limit.","section":"Sec. 4.1, Fig. 2 and Sec. 4.2"},{"comment":"The ROD < 10−6 stopping criterion gives 1/ℓmax ≈ 0.004J1, but this is a measure of the residual off-diagonality in the truncated operator space, not a measurement of the physical single-magnon linewidth Γ(k). Equation (B17) shows exponential suppression of off-diagonal elements only when one-magnon energies lie below all three-magnon energies. For J2 ≠ 0 the upper magnon branch is generically unstable (Refs. [28,30]), so the exact qpc flow diverges; the numerical flow is stopped before that divergence becomes visible. The 'converged' single-magnon peak position used to define ΔS and ΔR is therefore the result of a truncated generator flow and need not coincide with the peak of the true spectral function. To support the central quantitative claims, the authors should estimate Γ(k) at the relevant momenta, e.g. from the imaginary part of the one-magnon self-energy or from an independent nu","section":"App. B, Eq. (B17); Sec. 4.2.1 and 4.2.2"},{"comment":"The text describes the CST as 'fully including magnon-magnon interactions' and reports the 14–20% reduction as the interaction-induced effect. However, the flow equations are truncated at scaling dimension d_sc ≤ 2, a truncation inherited from previous Heisenberg and XXZ studies. Appendix A argues that the altermagnetic NNN couplings introduce no new operator types, but it does not establish that omitted d_sc > 2 operators have negligible influence on ΔS and ΔR for the altermagnetic parameter range. Since both the roton minimum and the additional reduction beyond scSWT are genuine interaction effects, a truncation-order convergence check or a comparison with an independent method for this model is needed before the specific numbers in Figs. 4 and 5 can be taken as quantitative.","section":"Sec. 3.1 and Sec. 4.2.1"}],"minor_comments":[{"comment":"The single-magnon peaks in Figs. 6 and 8 are plotted with an artificial broadening of 0.1J1, which is much larger than 1/ℓmax and larger than typical decay widths the paper claims to ignore. The 'sharpness' of the peaks therefore does not provide independent evidence of approximate stability; this should be stated explicitly.","section":"Sec. 4.3"},{"comment":"Sec. 4.1 states that the qpc flow converges only if the energy of an n-magnon mode is always smaller than the energy of an m>n mode, while App. B notes that the qpc generator can in principle re-order eigenstates if the flow is not truncated. These two statements should be reconciled, since the convergence criterion in the main text is stated more restrictively than the appendix allows.","section":"Sec. 4.1 and App. B"},{"comment":"There is a typo 'glsnnn contributions' in the text near Eq. (A10). Also, the paragraph after Eq. (A8) notes corrections to earlier papers; these corrections are useful but should be presented consistently (the missing factor in E_HB^0 and the l^2 vs m^2 issue are mentioned only in passing).","section":"App. A"},{"comment":"The 14–20% interval is quoted without specifying which J2 values are used and how the boundary-condition error bars are included. Please state the precise procedure and the error estimate for the interval, especially given the large relative error bars near J2 = 0.","section":"Sec. 4.2.1, Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for cond-mat.str-el and presents a systematic method application. The central concern is not the method itself but the evidential weight placed on the qpc convergence criterion for the quantitative claims. The authors are transparent about the limitations, but the missing Γ(k) estimate is a genuine gap that should be addressed before publication. I do not recommend rejection, because the issue is identifiable and potentially fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a well-executed application of the established CST machinery to a minimal 2D altermagnet, and the new numbers — the 14–20% reduction of the spin splitting and the J2-dependent roton minimum — are plausible. But the strongest quantitative claims depend on an approximate-stability argument that the paper does not fully make rigorous.\n\nWhat is actually new: the authors apply the qpc generator to the J1–J2–J2′ model with J2′=0, map out 0n and qpc convergence boundaries, and extract renormalized single-magnon dispersions and sector-resolved spectral densities. The roton minimum's nonlinear and non-monotonic dependence on J2, which contradicts simple scaling, is a genuinely new result. The paper does this carefully: finite-size extrapolation with both boundary conditions, an explicit energy-resolution criterion Δω = max(1/ℓmax, v_spin/L), and honest acknowledgment that rigorous magnon stability is absent for J2≠0. That is credit where it is due.\n\nSoft spots, in proportion: (1) The central interpretation — that qpc convergence within 1/ℓmax ≈ 0.004 J1 implies approximate magnon stability — is not established. The ROD measures residual off-diagonality in a truncated operator space, not the physical decay width Γ(k). If Γ(k) at the momenta used for ΔS and ΔR exceeds ~0.004 J1, the 'converged' dispersion could be an artifact of stopping the flow before the divergence becomes visible. The authors argue qualitatively that phase space is small, but they never estimate Γ(k). This is the weakest link. (2) The flow equations themselves are not written out; they are referenced to prior work. For a model with two boson flavors, this makes independent verification harder. No code or data is shipped. (3) The truncation at d_sc≤2 is carried over from the square-lattice Heisenberg model without an independent check for this model. (4) The 14–20% range is quoted without full error propagation; the pbc/apbc difference is the main error estimate, and near J2=0 the relative normalization makes the error bars large.\n\nNone of these are fatal in my view. The qualitative message — spin splitting is robust, quantum effects renormalize it, and the roton minimum is enhanced by antiferromagnetic NNN couplings — is likely correct. But the precise percentages and roton depths should be treated as conditional on the decay-width assumption.\n\nWho benefits: anyone interpreting INS on altermagnets or assessing magnon stability for magnonics. It deserves a serious referee; the right referee will ask for a decay-width estimate, ideally from a complementary method (tDMRG or NLSWT) in the overlapping parameter range.","headline":"A careful and plausible CST study of a minimal 2D altermagnet, but the headline renormalization numbers rest on an approximate-stability assumption that needs a decay-width estimate before they can be taken at face value.","tokens_in":27724,"tokens_out":3683,"would_cite":true,"duration_ms":35581,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Quantum interactions between magnons reduce the altermagnetic spin splitting by 14–20% relative to linear spin-wave theory and produce a roton minimum in the lower branch that deepens when the next-nearest-neighbor coupling is frustrating.","keywords":["altermagnet","magnon spectrum","continuous similarity transformation","spin splitting","roton minimum","magnon-magnon interaction","square-lattice Heisenberg model","spin-1/2"],"falsifier":"Computing the single-magnon spectral function with a higher flow cutoff or on larger lattices so that decay linewidths below 0.004 J1 are resolved; if the one-magnon peak acquires a Lorentzian width exceeding the current resolution, or if the qpc flow diverges for values of J2 where it is reported to converge, the roton minimum and the 14–20% splitting reduction are artifacts of the truncation. A complementary falsifier is an inelastic neutron scattering experiment on a quasi-2D altermagnet that resolves a magnon linewidth comparable to 1/ℓmax.","tokens_in":26831,"feed_emoji":"🧲","tokens_out":8647,"duration_ms":66622,"temperature":0.7,"pith_summary":"The paper asks whether the spin splitting of magnon bands in a two-dimensional altermagnet survives quantum fluctuations in a spin-1/2 Heisenberg model, and how those fluctuations reshape the spectrum. Using a continuous similarity transformation that decouples sectors with different magnon numbers, it computes the one-magnon dispersion beyond linear spin-wave theory. The central quantitative claims are that magnon-magnon interactions reduce the altermagnetic spin splitting by 14% to 20% relative to linear spin-wave theory, and that a roton minimum appears in the lower magnon branch whose depth grows nonlinearly when the next-nearest-neighbor coupling is frustrating (J2 > 0). The paper also maps the stability region: the Néel order holds up to J2 ≈ 0.66 J1, while magnons are only approximately stable within a narrower window, bounded by the convergence of the transformation. A sympathetic reader would care because these are concrete, measurable predictions for inelastic neutron scattering on quasi-2D altermagnets.","feed_headline":"Magnon interactions cut altermagnet spin splitting by 14–20%","feed_subtitle":"Interactions between spin waves also create a roton dip that deepens with frustration.","key_machinery":"The central mechanism is a momentum-space continuous similarity transformation (CST) applied to a bosonized Heisenberg Hamiltonian after a self-consistent mean-field decoupling and a canonical rotation to gapless magnons. Two generators are used: a 0n generator that removes vacuum fluctuations and whose convergence marks the Néel-phase boundary, and a quasi-particle-conserving (qpc) generator that block-diagonalizes the Hamiltonian into fixed magnon-number sectors; the qpc flow converges only when one-magnon and three-magnon sectors do not overlap in energy, so its convergence (with residual off-diagonality below 10^-6 J1 and energy resolution 1/ℓmax ≈ 0.004 J1) is treated as a proxy for app","core_discovery":"The discovery is that magnon-magnon interactions, treated fully through the continuous similarity transformation, renormalize the altermagnetic magnon spectrum in a quantitative and partly counterintuitive way. At the momentum (π/2, π/2), the spin splitting ΔS = ω↓ − ω↑, which is 2J2 in linear spin-wave theory, is reduced by 14–20% for |J2| ≤ 0.16 J1, with the largest reduction for antiferromagnetic (frustrating) next-nearest-neighbor coupling. The same interactions create a roton minimum at (π,0) in the ω↓ branch: ΔR = ω↓(π/2,π/2) − ω↓(π,0) is exactly zero in LSWT and in self-consistent spin-wave theory, but finite in CST, and its depth increases strongly and nonlinearly with J2 > 0 while d","pith_inferences":["If the resolution-limited criterion holds, the quantitative predictions — 14–20% splitting reduction and a J2-dependent roton depth — are sharp enough to be tested by inelastic neutron scattering on quasi-2D altermagnets with weak interlayer coupling; a measurement resolving the roton would discriminate between mean-field and full interaction theories.","The near-zero lower bound of qpc convergence for J2<0 suggests that even weak ferromagnetic NNN coupling may open decay channels or level crossings that LSWT misses; testing whether this is a physical instability or a truncation artifact via exact diagonalization on larger clusters would clarify the true stability window.","The non-monotonic dependence of the roton depth on J2 implies that strain or pressure tuning of the NNN coupling in candidate altermagnets could be a control knob for quantum fluctuations, potentially enhancing or suppressing magnon decay."],"forward_implications":["The altermagnetic spin splitting of the magnon bands is a robust signature: even with full magnon-magnon interactions it survives with 80–86% of its linear-spin-wave value for |J2| ≤ 0.16 J1.","The roton minimum, a purely quantum effect absent in LSWT and scSWT, is present in the lower magnon branch for J2 = 0 and grows nonlinearly with frustrating NNN coupling, so its depth can serve as a measure of quantum fluctuations.","The stability map is asymmetric: the 0n flow converges for all J2 < 0 and up to J2,c ≈ 0.66 J1 for J2'=0, while the qpc flow converges only in a window roughly 0 ≲ J2/J1 ≲ 0.2, delimiting where single-magnon quasiparticles are approximately valid.","Spectral densities show that interactions shift weight from the one-magnon peak into the three-magnon continuum, and the two-magnon peak at (π/2,π/2) splits only when interactions are included, giving distinctive signatures for neutron scattering."],"fun_headline_variants":["Magnon interactions shrink altermagnet spin splitting by up to 20%","Spin-wave interactions create roton dip in altermagnets","Altermagnet magnon spectrum renormalized: splitting reduced 14-20%","Quantum effects cut altermagnet band splitting, add roton minimum","Magnon-magnon interactions alter altermagnet spin splitting and roton"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that a converging quasi-particle-conserving flow, stopped at a finite cutoff with energy resolution 1/ℓmax ≈ 0.004 J1, reliably indicates that magnons are approximately stable, so that the extracted dispersion, roton minimum, and splitting renormalization are physical rather than truncation artifacts — even though exact magnon stability is known to be absent for altermagnets.","fun_headline_variants_meta":{"raw":{"variants":["Magnon interactions shrink altermagnet spin splitting by up to 20%","Spin-wave interactions create roton dip in altermagnets","Altermagnet magnon spectrum renormalized: splitting reduced 14-20%","Quantum effects cut altermagnet band splitting, add roton minimum","Magnon-magnon interactions alter altermagnet spin splitting and roton"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000222,"raw_usage":{"total_tokens":1275,"prompt_tokens":716,"completion_tokens":559,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":460,"completion_tokens_details":{"reasoning_tokens":461}},"tokens_in":460,"tokens_out":559,"duration_ms":4625,"temperature":1.0,"reasoning_tokens":461,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T23:53:43.193601+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Computing the single-magnon spectral function with a higher flow cutoff or on larger lattices so that decay linewidths below 0.004 J1 are resolved; if the one-magnon peak acquires a Lorentzian width exceeding the current resolution, or if the qpc flow diverges for values of J2 where it is reported to converge, the roton minimum and the 14–20% splitting reduction are artifacts of the truncation. A complementary falsifier is an inelastic neutron scattering experiment on a quasi-2D altermagnet that resolves a magnon linewidth comparable to 1/ℓmax.","supporting_citations":[],"review_version":1}