{"id":"5493619e-ae0f-4148-83d2-53b0db76d4b9","arxiv_id":"2608.04946","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":15,"one_line_summary":"Off-stoichiometric CeCuBi2 is an antiferromagnet (TN ~ 14 K) with Kondo/heavy-fermion signatures, multiple metamagnetic phases, a field-induced glassy state, and anisotropic magnetoresistance up to about 10.9%.","lead":"This paper reports measurements on off-stoichiometric CeCuBi2 crystals, showing an antiferromagnet with Kondo-like heavy-electron behavior, several field-induced magnetic phases, and unusually large angle-dependent magnetoresistance. A general reader might care because the material combines magnetic order, strong anisotropy, and room-temperature magnetoresistance, which is interesting for correlated-matter and spintronics research.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The field-induced spin-glass assignment rests on frequency-shift fits at two fields in the metamagnetic region; without aging/memory or nonlinear-susceptibility checks, the glassy pillar of the central claim is not yet established.","rationale":"The reader's weakest_assumption targets the same load-bearing pillar I identify: the spin-glass-like state is inferred from AC susceptibility frequency shifts at only two fields, with no aging, memory, or nonlinear-susceptibility confirmation. I agree with that assessment. I add that tau0 at 51 kOe is at the cluster-glass end of the quoted range, and the paper's own limitation statement defers microscopic phase identification to future neutron or resonant X-ray scattering experiments. The non-glassy claims, such as TN ~ 14 K, enhanced gamma, metamagnetic transitions, and large anisotropic magnetoresistance, are directly supported by the presented data and internally consistent, so a REJECT verdict would be too strong. The existing CONDITIONAL verdict already captures the risk to the field-induced glassiness claim, and my proposed concrete test would settle the question directly without requiring new theory or a different dataset.","tokens_in":21237,"tokens_out":4537,"duration_ms":57801,"concrete_test":"At Hdc = 51 kOe, cool the crystal in the applied field to T approximately equal to Tf minus 1 K, wait for t_w = 1000 s, then measure the real part of the AC susceptibility on warming at 10 Hz and 1 kHz and compare with a reference warming curve obtained without the halt. A canonical spin glass shows a characteristic memory dip at the halt temperature and a frequency-dependent Tf that is reproducible for both ZFC and FC histories; if the memory dip is absent or the frequency shift disappears or changes sign for a field-cooled history, the response is dominated by first-order metamagnetic metastability or domain-wall blocking rather than equilibrium spin freezing. If available, also record the nonlinear susceptibility chi3(T) near Tf, which should show a divergent peak for canonical spin-glass freezing and a suppressed or broad response for superparamagnetic clusters or domain effects.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that off-stoichiometric CeCuBi2 hosts coexisting Kondo heavy-fermion behavior, large anisotropic magnetotransport, and field-induced glassiness stands or falls mainly on the last element. The glassy state is inferred solely from frequency-dependent AC susceptibility peaks at Hdc = 47 and 51 kOe (Figs. 8(e), 8(f), 9(a)-9(d)), with a Mydosh parameter K = 0.006-0.008 and critical-slowing-down fits giving tau0 ~ 7.2e-13 s and 7.3e-10 s. These fields lie inside the metamagnetic region IV, where DC magnetization already shows hysteresis, ZFC-FC bifurcation, and a ZFC/FC crossing (Fig. 3(e)-(f)). First-order-like metamagnetic transitions, domain-wall pinning, or thermal-history effects can produce frequency-dependent AC responses without canonical spin freezing. The paper reports no aging, memory, or nonlinear-susceptibility measurements, and no complete frequency list; the fits are performed at only two fields. Note also that tau0 = 7.3e-10 s sits at the cluster-glass/superparamagnetic end of the cited range rather than clearly in the canonical atomic spin-glass range. The authors themselves concede in Section IV that field-dependent neutron diffraction or resonant X-ray scattering is needed to determine the intermediate magnetic structures, so the microscopic nature of region IV remains unverified. The remaining claims, TN ~ 14 K, gamma ~ 102 mJ K-2 mol-1, multiple metamagnetic transitions, ~22% room-temperature MR, and ~10.9% AMR, are directly supported by the shown data and would survive even if the glassy assignment is weakened; but the 'coexistence' headline depends on the glassy pillar.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a comprehensive experimental study of single crystals identified by EDS as off-stoichiometric CeCuBi2 (Ce:Cu:Bi ≈ 1:0.8:2.2), covering structure, anisotropic magnetization, specific heat, resistivity, magnetoresistance, angle-dependent AMR, and AC susceptibility. The authors find antiferromagnetic ordering at TN ≈ 14 K with strong uniaxial anisotropy, a Sommerfeld coefficient γ ≈ 102 mJ K−2 mol−1, a Kondo-like resistivity hump near 47 K, multiple metamagnetic steps for H || c, frequency-dependent AC susceptibility peaks at 47 and 51 kOe interpreted as field-induced spin-glass-like behavior, and large anisotropic magnetotransport (MR ≈ 22% at 300 K and 90 kOe; AMR ≈ 10.9% at 2.5 K and 90 kOe). They conclude that off-stoichiometric CeCuBi2 is a platform where Kondo-driven heavy-fermion behavior, large anisotropic magnetotransport, and field-induced glassiness coexist.","tokens_in":21658,"tokens_out":7738,"duration_ms":88385,"significance":"If substantiated, this paper would provide a valuable experimental phase diagram for an anisotropic Kondo antiferromagnet with strong magnetotransport responses, and the reported 10.9% AMR is a useful addition to the antiferromagnetic AMR database. The core observations—TN ≈ 14 K, γ ≈ 102 mJ K−2 mol−1, metamagnetic step fields, MR values, and AMR values—are directly supported by the presented data; the angle-dependent measurements and polar plots are a particular strength. The two weakest points are the field-induced glassy-state interpretation, which rests on AC-susceptibility frequency shifts at only two DC fields without standard spin-glass diagnostics, and the sample composition, which rests on EDS alone. Neither point invalidates the transport and magnetization measurements, but the central claim of coexisting field-induced glassiness needs either new measurements or a substantial reinterpretation.","major_comments":[{"comment":"The central claim of field-induced glassiness is not established by the presented data. The frequency-dependent χ′(T) peaks are observed at only two DC fields, 47 and 51 kOe, both inside metamagnetic region IV, where DC magnetization already shows hysteresis, ZFC–FC bifurcation, and an unusual ZFC–FC crossing. No aging, memory, or nonlinear-susceptibility measurements are reported, and the frequency list used for the fits is not given. The critical-slowing-down fits at the two fields yield τ0 = 7.2×10−13 s and 7.3×10−10 s, a spread of about three orders of magnitude, and the latter value sits at the cluster-glass/superparamagnetic end of the canonical range the authors cite. The fits are also underreported: no zν values, fit ranges, or uncertainties are given. First-order metamagnetic transitions, domain-wall pinning, or thermal-history effects can produce frequency-dependent AC peaks without canonical spin freezing. To support the conclusion, the authors should provide a complete frequency series, aging or memory checks, nonlinear susceptibility data, or a clear exclusion of metamagnetic-transition artifacts; otherwise the conclusion should be weakened to 'slow spin dynamics' and the word 'glassiness' should be removed from the central claim.","section":"Section III, AC susceptibility measurements; Figs. 8(e), 8(f), 9(a)–9(d)"},{"comment":"The identification of the sample as off-stoichiometric 'CeCuBios2' with composition 1:0.8:2.2 rests entirely on EDS measurements on the crystal surface. The nine measured regions are mentioned, but no table or list of individual EDS values and no standard deviations are provided. Because the title, the lattice-contraction discussion, and the assignment of the Kondo hump and enhanced γ to Cu vacancies and Bi excess all depend on the composition being bulk and uniform, the authors need to either report the full EDS statistics or provide complementary bulk composition data (for example, wavelength-dispersive electron microprobe analysis or solution-based elemental analysis).","section":"Section II and Table I"},{"comment":"The H–T phase diagram labels five field-induced phases (I–V), but their boundaries are inferred from dM/dH and susceptibility anomalies, and the microscopic spin arrangements are not determined. The authors themselves state in Section IV that field-dependent neutron diffraction or resonant X-ray scattering is needed to determine the intermediate magnetic structures. Since the transport interpretation repeatedly invokes specific field-induced spin configurations (canting, spin-flop, spin-locked states), the macroscopic signatures alone cannot prove those configurations. The manuscript should clearly separate what is measured (metamagnetic steps, MR anomalies, AC peak shifts) from what is inferred (particular spin reorientation mechanisms and glassy freezing), and the concluding claim should be calibrated accordingly.","section":"Section IV and Fig. 8(a)"}],"minor_comments":[{"comment":"The notation CeCuBios2 is used before its definition is fully explained; please define it at first occurrence, including in the abstract if the symbol is used there.","section":"General"},{"comment":"Please include uncertainties for the new lattice parameters in Table I and a clear statement of how many spots were averaged for the EDS composition.","section":"Section II and Table I"},{"comment":"The text refers to 'at constant fields of 29.7 kOe and 30 kOe' (Fig. 8); please clarify whether these are two distinct measurements or whether one value is a typographical error.","section":"Section III, AC susceptibility"},{"comment":"The captions and text for the critical-slowing-down fits do not specify the excitation frequencies used or the number of points in each fit; please add this information so the fits can be evaluated.","section":"Fig. 9 and Section III"},{"comment":"There are minor typographical issues, including 'resistivty' in the magnetoresistance discussion and inconsistent hyphenation of 'spin-glass-like'; a careful proofreading pass is recommended.","section":"General"},{"comment":"No data availability statement is included; many journals now require one, so please add it.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper reports a solid set of experiments with directly supported values for TN, γ, metamagnetic fields, MR, and AMR. The main risk is overinterpretation: the field-induced spin-glass conclusion rests on AC susceptibility at two fields without aging/memory checks, and the sample composition rests on EDS alone. I recommend major revision rather than rejection because the requested changes are additions or reframing within the manuscript's scope. If the authors cannot add the spin-glass diagnostics, they should remove the glassiness claim from the title and conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a serious look. The core experimental results on off-stoichiometric CeCuBi2 look credible and are mostly well supported by the plotted data: TN ~14 K, gamma ~102 mJ K-2 mol-1, multiple metamagnetic steps for H||c, ~22% MR at 300 K, and ~10.9% AMR at 2.5 K and 90 kOe. The angle-dependent magnetotransport maps are genuinely new for this family, and the paper gives a useful comparison against stoichiometric CeCuBi2. The resistivity anisotropy, Kondo hump, and the H-T phase diagram are consistent with the susceptibility and specific-heat results. That part of the work is solid.\n\nThe weak link is the 'field-induced glassiness' headline. The spin-glass claim rests on frequency-dependent AC susceptibility peaks at two DC fields, 47 and 51 kOe, with Mydosh parameters 0.006-0.008 and tau0 values from critical-slowing-down fits. That is thin: no aging, memory, or nonlinear-susceptibility data, and fits at only two fields. Both fields sit inside metamagnetic region IV, where DC magnetization already shows hysteresis and ZFC-FC bifurcation. First-order-like metamagnetic transitions or domain-wall pinning can generate frequency-dependent AC responses without canonical freezing. Also tau0 = 7.3e-10 s at 51 kOe sits at the cluster-glass/superparamagnetic end of the cited range, not clearly in the canonical atomic spin-glass range. The authors themselves concede in Section IV that field-dependent neutron diffraction or resonant X-ray scattering is needed to identify the intermediate magnetic structures. That is the key missing confirmation.\n\nTwo smaller points. The off-stoichiometric composition rests on EDS alone, with no error bars or a second probe; acceptable as a first report but worth flagging. And the topological remarks cite a CeAuBi2 band-structure calculation (Ref. 11) as if it were about CeCuBi2; that citation mismatch should be fixed.\n\nOverall, the robust observations survive even if the glassy interpretation is dropped. I would send this to referees, but the referee should require either additional spin-glass evidence (aging/memory/nonlinear susceptibility) or a softened claim of 'glass-like dynamics' rather than canonical spin glass. I would cite it for the magnetotransport and phase-diagram data, and it is a reasonable reading-group candidate for the correlated-electron and antiferromagnetic-spintronics crowd.","headline":"Solid experimental characterization with credible magnetotransport and metamagnetic data; the field-induced spin-glass claim is under-supported.","tokens_in":22331,"tokens_out":3341,"would_cite":true,"duration_ms":33416,"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":"Off-stoichiometric CeCuBi2 single crystals are claimed to combine Kondo heavy-fermion behavior, field-induced spin-glass-like phases, and large anisotropic magnetoresistance.","keywords":["CeCuBi2","Kondo lattice","heavy fermion","metamagnetic transitions","spin glass","magnetoresistance","anisotropic magnetotransport"],"falsifier":"Cool the crystal at 47 and 51 kOe through the freezing peak with an aging stop and then with a memory protocol: a canonical spin glass shows a memory dip and time-dependent AC susceptibility, whereas a metamagnetic transition with domain-wall pinning does not. A parallel field-dependent neutron diffraction measurement would show whether the intermediate state is frozen disorder or a long-range canted magnetic structure.","tokens_in":20963,"feed_emoji":"🧲","tokens_out":14585,"duration_ms":144049,"temperature":0.7,"pith_summary":"The paper sets out to show that slightly copper-deficient CeCuBi2 crystals are one material in which Kondo-driven heavy-fermion behavior, several field-induced magnetic phases, and a spin-glass-like frozen state all appear. It reports antiferromagnetic order at $T_N\\sim14$ K, a Sommerfeld coefficient $\\gamma=102$ mJ K$^{-2}$ mol$^{-1}$, and at least five metamagnetic phases for fields along the easy $c$-axis, with frequency-dependent AC susceptibility peaks at 47 and 51 kOe whose Mydosh parameter and relaxation time fall in the canonical spin-glass range. The same crystals show large positive magnetoresistance, about 22 percent at 300 K and 90 kOe, and a butterfly-shaped anisotropic magnetoresistance reaching about 10.9 percent at 2.5 K and 90 kOe. If correct, the paper identifies off-stoichiometric CeCuBi2 as a tunable platform for correlated, field-tunable quantum phenomena in which the magnetism and the transport are two readings of one competition among Kondo hybridization, magnetic anisotropy, and Cu-vacancy disorder. The central novelty is that the glassy dynamics appear inside the metamagnetic region, not at zero field.","feed_headline":"CeCuBi2 ties Kondo physics, frozen spins, and 22% magnetoresistance","feed_subtitle":"Off-stoichiometric crystals show field-induced glassy phases and anisotropic magnetoresistance up to 10.9 percent.","key_machinery":"The load-bearing mechanism is the field-driven rearrangement of an Ising-like easy-axis antiferromagnet: for $\\mathbf{H}\\parallel c$, the competition among exchange, anisotropy, and Zeeman energy produces spin-flop and spin-flip transitions, with the spin-flop field $H_{SF}=\\sqrt{2H_AH_E-H_A^2}$, and the paper encodes the resulting states in a field–temperature phase diagram with regions I–V. The glassiness claim is carried by the Mydosh parameter $K=\\Delta T_f/(T_f\\,\\Delta\\log_{10}f)$ and by the critical slowing-down form $\\tau=\\tau_0(T_f/T_g-1)^{-z\\nu}$, which place the frequency shift of the AC susceptibility peaks at 47 and 51 kOe in the canonical spin-glass window. The magnetotransport claim is carried by angle-resolved magnetoresistance and AMR polar plots, whose evolution from two-lobed or four-lobed patterns at low field to eight-lobed butterfly-like patterns at 90 kOe links spin reorientation to anisotropic scattering.","core_discovery":"The authors claim that off-stoichiometric CeCuBi2, with elemental composition near Ce:Cu:Bi = 1:0.8:2.2, is an anisotropic Kondo antiferromagnet in which one set of crystals shows weak heavy-fermion behavior ($\\gamma=102$ mJ K$^{-2}$ mol$^{-1}$, $T_N\\sim14$ K), five field-induced metamagnetic phases for $\\mathbf{H}\\parallel c$, and a field-induced spin-glass-like state inside the metamagnetic region: AC susceptibility peaks at 47 and 51 kOe shift with frequency, with Mydosh parameter $K=0.006$–$0.008$ and relaxation time $\\tau_0\\sim7\\times10^{-13}$–$7\\times10^{-10}$ s. The same crystals show large anisotropic magnetotransport, with positive magnetoresistance of about 22 percent at 300 K and 90 kOe, roughly 32–35 percent at 2.5 K, and butterfly-like anisotropic magnetoresistance up to about 10.9 percent. The conclusion is that Kondo hybridization, magnetic anisotropy, and Cu-vacancy disorder compete to produce a single field–temperature phase diagram in which the magnetization and transport anomalies align, making this compound a possible platform for correlated and anisotropic quantum phenomena.","pith_inferences":["A decisive test not reported in the paper is to measure aging, memory, and the nonlinear susceptibility $\\chi_3$ in region IV; if those confirm canonical freezing, field-induced glassiness in a Kondo antiferromagnet would become a tunable model for disorder- and frustration-driven slow dynamics.","The butterfly-shaped AMR at 90 kOe resembles the angular magnetoresistance attributed in the isostructural nodal-line semimetal ZrSiS to Zeeman-tuned electron–hole compensation; angle-dependent Hall or quantum-oscillation measurements could test whether the same compensation contributes here.","The microscopic identity of the intermediate phases is still open: field-dependent neutron diffraction or resonant X-ray magnetic scattering would reveal whether the region IV state is a true frozen glass or a pinned first-order canted structure, and would refine the proposed H–T phase diagram."],"forward_implications":["Magnetization and resistivity track the same spin reconfiguration: the drop in magnetoresistance near 49–53 kOe coincides with the sharp rise in magnetization, so transport can be used as a probe of the metamagnetic phase boundaries.","The reported $\\gamma\\sim102$ mJ K$^{-2}$ mol$^{-1}$ and the broad resistivity hump near 47 K place off-stoichiometric CeCuBi2 in the weak heavy-fermion regime, where Kondo hybridization coexists with long-range antiferromagnetic order instead of destroying it.","At 47 and 51 kOe, the frequency-dependent AC susceptibility peaks, the Mydosh parameter $K=0.006$–$0.008$, and the relaxation time $\\tau_0\\sim10^{-13}$–$10^{-10}$ s indicate a field-induced spin-glass-like region that exists only in intermediate fields.","A magnetoresistance of about 22 percent at 300 K and 90 kOe, together with an anisotropic magnetoresistance up to about 10.9 percent at 2.5 K, makes the Néel-vector orientation a strong control knob for resistance in this antiferromagnet.","The angular resistivity patterns are strongly field-sensitive and develop higher-order lobes at 90 kOe, implying that the anisotropic Fermi-surface or scattering contributions change with the direction and strength of the field."],"supporting_citations":[{"why":"This provides the layered $P4/nmm$ crystal structure and the Cu-vacancy (CeCu$_{1-x}$Bi$_2$) off-stoichiometry that motivates the sample chemistry.","marker":"[1]"},{"why":"This supplies the neutron magnetic structure ($q=(1/2,1/2,0)$ with Ce moments along $c$) and the pristine-crystal resistivity baseline against which the off-stoichiometric enhancement is judged.","marker":"[2]"},{"why":"This gives the low-temperature magnetic properties and $\\gamma$ values of the Ce$T$Bi$_2$ family used to classify CeCuBi$_2$ as a weak heavy-fermion system.","marker":"[3]"},{"why":"This reports $T_N\\sim16$ K for stoichiometric CeCuBi$_2$, the baseline for the $T_N\\sim14$ K observed in the off-stoichiometric crystals.","marker":"[10]"},{"why":"This predicts field-induced nodal-line topological states in square-net pnictides, motivating the angle-dependent transport measurements and the interpretation of the polar plots.","marker":"[11]"},{"why":"This supplies the spin-flop and spin-flip phase-transition model for collinear easy-axis antiferromagnets used to assign the metamagnetic transitions.","marker":"[20]"},{"why":"This defines the Mydosh parameter and the canonical spin-glass criteria against which the AC susceptibility frequency shifts are compared.","marker":"[64]"}],"fun_headline_variants":["CeCuBi2: field-induced magnetic phases and 22% magnetoresistance","Anisotropic Kondo antiferromagnet with field-induced spin-glass and 22% MR","CeCuBi2 shows spin-glass-like states and 22% magnetoresistance","Field-induced magnetic phases in anisotropic Kondo metal CeCuBi2","CeCuBi2: 22% MR and butterfly AMR up to 10.9%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the frequency-dependent AC susceptibility peaks at 47 and 51 kOe are canonical spin-glass freezing; without aging, memory, or nonlinear-susceptibility measurements, those peaks could equally come from the first-order metamagnetic transitions, domain-wall pinning, or thermal-history effects that the same data show.","fun_headline_variants_meta":{"raw":{"variants":["CeCuBi2: field-induced magnetic phases and 22% magnetoresistance","Anisotropic Kondo antiferromagnet with field-induced spin-glass and 22% MR","CeCuBi2 shows spin-glass-like states and 22% magnetoresistance","Field-induced magnetic phases in anisotropic Kondo metal CeCuBi2","CeCuBi2: 22% MR and butterfly AMR up to 10.9%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000972,"raw_usage":{"total_tokens":4184,"prompt_tokens":1046,"completion_tokens":3138,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":662,"completion_tokens_details":{"reasoning_tokens":3028}},"tokens_in":662,"tokens_out":3138,"duration_ms":24150,"temperature":1.0,"reasoning_tokens":3028,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T13:09:29.327652+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Cool the crystal at 47 and 51 kOe through the freezing peak with an aging stop and then with a memory protocol: a canonical spin glass shows a memory dip and time-dependent AC susceptibility, whereas a metamagnetic transition with domain-wall pinning does not. A parallel field-dependent neutron diffraction measurement would show whether the intermediate state is frozen disorder or a long-range canted magnetic structure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"This provides the layered $P4/nmm$ crystal structure and the Cu-vacancy (CeCu$_{1-x}$Bi$_2$) off-stoichiometry that motivates the sample chemistry."},{"cited_title":"Adriano, P","cited_arxiv_id":null,"evidence_quote":"This supplies the neutron magnetic structure ($q=(1/2,1/2,0)$ with Ce moments along $c$) and the pristine-crystal resistivity baseline against which the off-stoichiometric enhancement is judged."},{"cited_title":"Thamizhavel, A","cited_arxiv_id":null,"evidence_quote":"This gives the low-temperature magnetic properties and $\\gamma$ values of the Ce$T$Bi$_2$ family used to classify CeCuBi$_2$ as a weak heavy-fermion system."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"This reports $T_N\\sim16$ K for stoichiometric CeCuBi$_2$, the baseline for the $T_N\\sim14$ K observed in the off-stoichiometric crystals."},{"cited_title":"Wang and X","cited_arxiv_id":null,"evidence_quote":"This predicts field-induced nodal-line topological states in square-net pnictides, motivating the angle-dependent transport measurements and the interpretation of the polar plots."},{"cited_title":"Li, Possible ground states and parallel magnetic- field-driven phase transitions of collinear antiferromag- nets, npj Comput","cited_arxiv_id":null,"evidence_quote":"This supplies the spin-flop and spin-flip phase-transition model for collinear easy-axis antiferromagnets used to assign the metamagnetic transitions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"This defines the Mydosh parameter and the canonical spin-glass criteria against which the AC susceptibility frequency shifts are compared."}],"review_version":1}