{"id":"3069d3b6-fbd7-4c00-a2dd-64ec0e130601","arxiv_id":"2507.09143","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Field-dependent analytical escape times and octupole relaxation times for strained Mn3Sn, validated by stochastic LLG simulations in the low-barrier regime.","lead":"Thermal noise makes the magnetic octupole in strained Mn3Sn flip between two states, and this paper derives formulas for how fast those flips happen under an external magnetic field. The formulas match computer simulations and could underpin faster random number generators and probabilistic computers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Numerical validation of Eqs. (7a)/(7b) depends on a calibrated escape-time threshold ξ=0.5, so the claimed strong LLG agreement is not a fully independent test of the analytic formulas.","rationale":"I read the paper in good faith: it derives closed-form HTST escape times for field-driven octupole dynamics in strained Mn3Sn, checks the depopulation factor numerically, validates the one-Kagome-plane approximation against a six-spin model, and honestly documents deviations at high fields. These are real strengths. The central claim, however, is an agreement claim: the analytic formulas are said to match comprehensive LLG simulations. For that claim to be load-bearing, the simulation observable must be an unambiguous realization of the analytic rate. The supplement's threshold-based extraction (ξ=0.5, calibrated to Ref. [46]) introduces a free parameter into the comparison. This is not fatal by itself: Fig. 8(c) shows the extracted escape time is only mildly ξ-dependent for the claimed barrier range, and the two plane/six-spin benchmark is a thoughtful control. But because the calibration target comes from a prior model with the same perturbative assumptions, the comparison is not fully independent. The reader's weakest assumption concerned the perturbative reduction of the Hamiltonian; that concern is partially mitigated by the fact that the numerical LLG uses the full three-sublattice Hamiltonian, and the authors explicitly discuss where the perturbative expression fails. The threshold calibration is therefore the more immediate soft spot in the evidence for the quantitative claim. A threshold-free relaxation-time comparison would settle it cleanly. I therefore keep the reader's CONDITIONAL verdict (no change) and recommend requesting that additional analysis, along with the code/data artifacts noted by the reader.","tokens_in":21568,"tokens_out":21277,"duration_ms":225682,"concrete_test":"Re-analyze the existing full-LLG simulation data without any ξ threshold: for each field/barrier condition in Fig. 3, compute the ensemble-averaged octupole moment moct,y(t) and fit it to an exponential decay to extract τ_relax,fit. Compare these threshold-free relaxation times with the analytic prediction τ_relax = τesc,↑τesc,↓/(τesc,↑+τesc,↓) obtained from Eqs. (7)-(8). If the relaxation-based comparison matches as well as the ξ=0.5 escape-time comparison, the calibration concern is resolved. If the match degrades markedly at high fields (≳75 mT) or at the lowest barriers (ΔE0 = 3 kBT), the apparent agreement in Fig. 3 is partly an artifact of the ξ choice.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the HTST escape-time formulas, Eqs. (7a) and (7b), quantitatively predict the inter-well escape time and octupole relaxation time in strained Mn3Sn. The supporting evidence is the agreement with full three-sublattice stochastic LLG simulations in Fig. 3. However, the escape time is not extracted from the simulations as the mean first-passage time to the saddle point; it is defined by the condition that moct,y crosses ±ξ, with ξ a free parameter. The supplement (Sec. I B 2) states that ξ=0.5 is selected because it 'reduces dependence on ξ' and 'matches well with reported relaxation time' from Ref. [46]. This calibration is load-bearing: the analytic HTST rate describes escape over the barrier, while the simulation observable is first crossing of an arbitrary interior contour. Any systematic discrepancy between the HTST prefactor and the true LLG rate can be partially absorbed by the choice of ξ, so the agreement in Fig. 3 is not a completely independent confirmation of Eq. (7). The ξ-sensitivity analysis in Fig. 8(c) shows moderate sensitivity for ΔE0 ≥ 3 kBT, but the calibration target itself derives from a model sharing the same assumptions, so it does not break the loop. The perturbative-reduction concern raised by the reader is real but is directly tested by the full-LLG simulations; the threshold calibration is the remaining unvalidated element in the evidence chain.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes thermally activated dynamics of the magnetic octupole moment in strained, epitaxial Mn3Sn under an external magnetic field. Starting from the three-sublattice Hamiltonian and a perturbatively reduced octupole Hamiltonian, the authors apply harmonic transition-state theory (HTST) to derive closed-form expressions for the inter-well escape time (Eqs. 7a and 7b) and for the relaxation time of the octupole moment (Eq. 10). They benchmark these expressions against numerical solutions of coupled stochastic Landau-Lifshitz-Gilbert (LLG) equations for energy barriers of 3-6 kBT and fields up to 0.1 T, and they discuss implications for random number generation and probabilistic computing. The supplement contains the full HTST derivation, LLG solver benchmarking, sensitivity analyses, and an explicit statement that the perturbative Hamiltonian degrades at higher fields.","tokens_in":21856,"tokens_out":12998,"duration_ms":137038,"significance":"If the analytical formulas are valid, the paper provides design-oriented closed-form expressions for the field- and temperature-dependent escape and relaxation times in low-barrier strained Mn3Sn, which is a useful step for probabilistic computing proposals based on this material. The derivation is transparent and mostly self-contained in the supplement, material parameters are taken from the published literature rather than fit to the numerical data, and the authors honestly restrict the analysis to the claimed validity regime and identify the high-field breakdown. The main weaknesses lie in the numerical validation protocol: the LLG escape-time observable is defined by a threshold that is calibrated against the same type of theoretical framework used to derive the formulas, and the supplement states a damping-range validity that appears inconsistent with the damping used in the main-text simulations.","major_comments":[{"comment":"The numerical escape time is extracted as the time when moct,y crosses ±ξ, and ξ = 0.5 is selected because, as stated in Sec. I B 2, it \"matches well with reported relaxation time\" from Ref. [46]. Since Ref. [46] is itself a theoretical octupole-fluctuation model of essentially the same type as the one used here, the LLG curves in Fig. 3 are not a fully independent test of the absolute scale of Eqs. (7a)-(7b); a constant prefactor error in the HTST rate could be partially absorbed by the threshold choice. I ask the authors to provide a validation that does not depend on this calibration, for example mean first-passage times to the saddle point, or a multi-ξ study demonstrating that the field dependence of τesc in Fig. 3 is unchanged (within stated tolerances) when ξ is varied without re-tuning.","section":"Supplement Sec. I B 2, Fig. 8"},{"comment":"The main-text validation uses α = 0.003 (Table I), but Sec. I C 3 states that \"the theory fails for Hay when α < 5 × 10−3.\" Since 0.003 < 0.005, the Hay-assisted results in Fig. 3(b) appear to lie in the regime where the theory is said to fail. Please clarify whether the intended threshold is actually 5 × 10−4 (as suggested by the reported comparison between α = 10−4 and α = 5 × 10−4), or whether the Hay panel of Fig. 3 must be re-evaluated; in the latter case, the central validation for the y-field direction is affected.","section":"Supplement Sec. I C 3, Fig. 14 vs. Table I"}],"minor_comments":[{"comment":"The text states that numerical simulations cover ΔE0 from kBT to 5kBT, while the analysis and conclusion use (3-6)kBT; these ranges should be harmonized.","section":"Main text, simulation range"},{"comment":"The symbol δE is used both for the strain parameter in Eq. (1) and for the dissipated energy in Eq. (5); a distinct symbol for the dissipated energy would avoid confusion.","section":"Eqs. (1), (5), (6)"},{"comment":"The caption \"The schematic of escape time\" is not descriptive; it should state that the escape time from the up state to the down state is plotted as a function of applied field for different ΔE0.","section":"Fig. 3 caption"},{"comment":"The text describes the plotted quantity as a steady-state probability while the horizontal axis is pulse width; please clarify whether the probability is evaluated at the end of a pulse of given duration.","section":"Fig. 4(b) and surrounding text"},{"comment":"For clarity, define τescy,↑ and τescy,↓ explicitly rather than relying on the sentence after Eq. (7) to specify which barrier height enters the exponent for the down-state case.","section":"Eq. (7b)"}],"recommendation":"major_revision","confidential_remarks":"The derivation itself appears sound and the high-field limitations are handled honestly. The main risk is the ξ-calibrated numerical validation and the apparent contradiction between the damping used in the main text and the damping-range validity stated in the supplement; both should be resolved before the paper can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Alex -- quick read of arXiv:2507.09143. The paper does something useful: it takes Konakanchi's field-free octupole relaxation framework for strained Mn3Sn, adds an external magnetic field, and derives closed-form HTST escape-time expressions (Eqs. 7a/7b) with the full derivation in the supplement. The formulas are parameter-free in the sense that material constants come from Yoon et al. and the only adjustable piece is the numerical escape-time extraction threshold xi. The LLG comparison is honest: they show agreement for Delta E0 between 3 and 6 kBT and fields up to ~75 mT, and they explicitly attribute the high-field deviation to breakdown of the perturbative Hamiltonian, including a supplementary section on that breakdown. That is the right way to handle a model's edge.\n\nThe new physics is modest but real: for x-axis fields the two barriers split differently than for y-axis fields, giving two distinct relaxation modes--one symmetric (RNG) and one asymmetric (probabilistic control)--and they derive the corresponding tau_relax formulas. The RNG/probabilistic computing discussion is reasonable, not over-sold.\n\nSoft spots, in proportion. The stress-test note about xi is fair. Escape time in the simulations is not first-passage to the saddle; it is first crossing of moct,y = +/-0.5, and xi=0.5 was partly chosen because it matches the field-free relaxation time from Konakanchi et al. So the agreement in Fig. 3 is not a fully independent test of the HTST prefactor--the threshold can absorb some systematic prefactor error. I don't think it is load-bearing: the xi-sensitivity plot shows the extracted tau_esc flattens for Delta E0 >= 3 kBT, and the barrier dependence is dominated by exp(Delta E/kBT), which xi does not affect much. But the authors should state this more openly and ideally show one comparison with mean-first-passage-time extraction.\n\nOther nits: no code/data release ('available on request' is weak for a computational paper of this type), and the numerical points have no error bars (they say 1000 runs, but no confidence intervals). The two-Kagome-plane check is good, but only shown for two barriers and both field directions.\n\nBottom line: the central derivation is sound in the stated regime, the limits are stated, and the paper is a useful engineering-oriented extension. I would send it to review: the referee should push on the xi definition and ask for error bars, but the paper deserves referee time. I would cite it if I work on AFM thermal stability.","headline":"Solid HTST extension of octupole relaxation theory to applied fields, with honest scope limits; the xi-extraction threshold is a real but non-fatal soft spot.","tokens_in":22369,"tokens_out":2021,"would_cite":true,"duration_ms":23626,"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":"The paper claims that thermal escape of the octupole moment in a strained Mn3Sn film under an applied magnetic field is captured exactly by two closed-form transition-state formulas, matched by coupled-LLG simulations for barriers of 3 to…","keywords":["Mn3Sn","non-collinear antiferromagnet","thermal noise","harmonic transition-state theory","octupole moment","escape time","random number generation","probabilistic computing"],"falsifier":"Simulate or measure the up-to-down escape time of a strained Mn3Sn bit with barrier 4 kBT while sweeping Hay from 0 to 80 mT and compare with Eq. (7b); the formula is falsified if the full-Hamiltonian LLG result or a nanodot experiment departs by more than the few-percent agreement shown in the paper's Fig. 3(b) at the higher fields.","tokens_in":21358,"feed_emoji":"🧲","tokens_out":5778,"duration_ms":66449,"temperature":0.7,"pith_summary":"The paper tries to establish that thermal fluctuations in strained Mn3Sn under an applied field are quantitatively understood by a two-state rate theory. It derives analytic formulas for the inter-well escape time and the octupole relaxation time that match full numerical LLG simulations for low barriers (3 to 6 kBT) and fields up to 0.1 T. This matters because Mn3Sn is a non-collinear antiferromagnet with GHz-scale dynamics, and the formulas convert material parameters and field settings directly into switching timescales. The results point toward practical random-number generators and probabilistic computing elements built from a single Mn3Sn bit.","feed_headline":"Mn3Sn escape times match closed-form formula from 3 to 6 kBT","feed_subtitle":"Field and barrier determine octupole switching speed in a strained antiferromagnet, opening the path to GHz random-number generators.","key_machinery":"The engine is harmonic transition-state theory in the form of Eq. (5), which expresses the escape time as a depopulation factor times 2π/λ+, times the square-root of the ratio of Hessian eigenvalues at the saddle point versus the minimum, times exp(ΔE/kBT). Each term is evaluated using the reduced octupole Hamiltonian H(θoct, φoct) of Eq. (2), in which strain reduces the six-fold energy landscape to two-fold and the magnetic field further lifts degeneracy; λ+ is the positive eigenvalue of the linearized LLG equations at the saddle point, and the Hessian ratio accounts for the fluctuation modes. The specific closed forms in Eqs. (7a) and (7b), together with the barrier formulas of Eqs. (3) and (4), are what make the theory directly usable as design equations for device operation.","core_discovery":"The central claim is that the thermally activated escape of the octupole moment in a strained Mn3Sn film under an external magnetic field can be described by closed-form analytical formulas, Eqs. (7a) and (7b), obtained from harmonic transition-state theory applied to the effective octupole Hamiltonian. For energy barriers between 3 and 6 kBT and fields up to 0.1 T, these formulas predict the escape time and the octupole relaxation time in agreement with full stochastic LLG simulations. The formulas distinguish the field direction: a field along the -x axis creates two non-equivalent saddle points while keeping the two equilibrium states equally populated, whereas a field along the -y axis keeps the saddles degenerate but biases the equilibrium populations. This puts thermal switching of Mn3Sn into a tractable two-state rate model, giving field-tunable random-bit generation and probabilistic control in a single material.","pith_inferences":["A testable extension that the paper leaves implicit is that the same rate-theory construction should transfer to other Kagome antiferromagnets such as Mn3Ge or Mn3Ir by rescaling exchange, DMI, and anisotropy constants, since Eqs. (2)–(7) contain only those parameters.","The paper's damping sweep shows the theory breaks down for a y-axis field when α < 5×10^-3, hinting at a Kramers-turnover regime where energy exchange with the thermal bath becomes the bottleneck; quantifying that crossover would complete the low-damping picture.","Because the barrier formulas depend on strain through δE and the anisotropy constants, a precise escape-time measurement as a function of field orientation could serve as a non-invasive probe of the epitaxial strain in Mn3Sn films."],"forward_implications":["A field along the -x axis leaves the two octupole states equally populated, so a single strained Mn3Sn junction can act as a random bit source with sampling intervals near 2 ns and rates above 1 GHz.","A field along the -y axis breaks the symmetry between the two states, so the switching probability can be tuned continuously by field magnitude and pulse width, which is the operating principle of a probabilistic bit.","The escape-time formulas give closed-form design equations connecting energy barrier, film volume, field strength, and damping to switching speed, removing the need for repeated LLG simulation in parameter sweeps.","The two-state rate model predicts exponential relaxation of the octupole moment with time constant τrelax = τesc,↑τesc,↓/(τesc,↑+τesc,↓) and a field-dependent steady-state polarization given by (τesc,↑−τesc,↓)/(τesc,↑+τesc,↓).","The demonstrated validity range, low barriers (3–6 kBT) and low fields (≤0.1 T), is the operating window for thermal random number generators rather than for high-barrier memory bits."],"supporting_citations":[{"why":"Supplies the strained-film Hamiltonian, the two-fold energy landscape, and the material parameters used in Table I.","marker":"[18]"},{"why":"Provides the perturbative octupole Hamiltonian of Eq. (2) on which the escape-time derivation is built.","marker":"[49]"},{"why":"Supplies the harmonic transition-state theory framework for thermally activated escape in antiferromagnets used in Eq. (5).","marker":"[51]"},{"why":"Provides the HTST depopulation factor A and the intermediate-to-high-damping analysis entering Eq. (6).","marker":"[39]"},{"why":"Prior theoretical analysis of octupole relaxation time, used as the zero-field benchmark for the numerical LLG solver.","marker":"[46]"},{"why":"Earlier SOT-driven dynamics results used to benchmark the LLG solver in the noiseless limit.","marker":"[37]"},{"why":"Provides the stochastic thermal-field form of the coupled LLG equations used in the simulations.","marker":"[53]"}],"fun_headline_variants":["Closed-form escape times for Mn3Sn across 3–6 kBT","Mn3Sn thermal escape solved analytically under field","Field-tunable random bits from Mn3Sn escape rates","Analytic rate law for Mn3Sn octupole switching"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole calculation reduces the three manganese spins to a single octupole variable by assuming the three spins keep a rigid 120-degree arrangement; if an applied field or a thermal kick appreciably distorts that arrangement, the derived barriers and prefactors lose accuracy.","fun_headline_variants_meta":{"raw":{"variants":["Closed-form escape times for Mn3Sn across 3–6 kBT","Mn3Sn thermal escape solved analytically under field","Field-tunable random bits from Mn3Sn escape rates","Analytic rate law for Mn3Sn octupole switching"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000254,"raw_usage":{"total_tokens":1562,"prompt_tokens":932,"completion_tokens":630,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":559}},"tokens_in":548,"tokens_out":630,"duration_ms":6978,"temperature":1.0,"reasoning_tokens":559,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:02:50.526179+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate or measure the up-to-down escape time of a strained Mn3Sn bit with barrier 4 kBT while sweeping Hay from 0 to 80 mT and compare with Eq. (7b); the formula is falsified if the full-Hamiltonian LLG result or a nanodot experiment departs by more than the few-percent agreement shown in the paper's Fig. 3(b) at the higher fields.","supporting_citations":[{"cited_title":"\\ Yoon , author P","cited_arxiv_id":null,"evidence_quote":"Supplies the strained-film Hamiltonian, the two-fold energy landscape, and the material parameters used in Table I."},{"cited_title":"He \\ and\\ author L","cited_arxiv_id":null,"evidence_quote":"Provides the perturbative octupole Hamiltonian of Eq. (2) on which the escape-time derivation is built."},{"cited_title":"R \\'o zsa , author S","cited_arxiv_id":null,"evidence_quote":"Supplies the harmonic transition-state theory framework for thermally activated escape in antiferromagnets used in Eq. (5)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the HTST depopulation factor A and the intermediate-to-high-damping analysis entering Eq. (6)."},{"cited_title":"Shukla , author S","cited_arxiv_id":null,"evidence_quote":"Earlier SOT-driven dynamics results used to benchmark the LLG solver in the noiseless limit."},{"cited_title":"Go , author M","cited_arxiv_id":null,"evidence_quote":"Provides the stochastic thermal-field form of the coupled LLG equations used in the simulations."}],"review_version":1}