{"id":"57ae6828-d61b-463a-8309-1a81f789e873","arxiv_id":"2411.13469","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In a Floquet time crystal, out-of-time-ordered correlators show frozen then logarithmically growing scrambling along the protected magnetization direction, and all spin distances eventually merge into one growth curve.","lead":"The paper simulates a small disordered kicked spin chain that hosts a Floquet time crystal and finds that quantum information scrambles much more slowly along the spin direction protected by the time-crystalline order than along the other directions. Because time crystals are promising platforms for quantum sensing and simulation, understanding how they store or leak information matters for those applications.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central OTOC envelope claim rests on L=8 alone; no finite-size scaling shown for the distance-independent growth.","rationale":"The reader's conditional verdict is appropriate. The central claim—distance-independent late-time scrambling of Czz—rests on a single L=8 simulation (Fig. 1) and on a two-point exponential fit for T2(L). The time-scale hierarchy at L=8 is compressed: T2≈31, T3≈2000, leaving only ~1.5 decades for the purported logarithmic envelope. In such a small system, the OTOC for all distances is expected to approach its thermal maximum 1 on timescales that may be only weakly ℓ-dependent, so the 'envelope' could be a finite-size artifact rather than a genuine collective decoherence mode. A concrete way to distinguish these is to enlarge L to 12/14 and test whether the incremental growth is ℓ-independent and whether the extracted time scales grow exponentially with stable exponents. If the envelope persists and β,β′ stabilize, the paper's interpretation is supported; otherwise the headline phenomenon is not established. The reader's weakest assumption overlaps with this concern: the small L limits the reliability of both the ℓ-bit description and the time-scale scaling. I also note the appendix contains algebraic errors (e.g., the commutator [U_F,σ^z_l]=0 used in Eq. A1 is not true for finite kick, and the step from Eq. A3 to the identity for Cxx is not justified), but these are supporting calculations and do not directly determine the numerical envelope claim. For these reasons the verdict should remain conditional.","tokens_in":16729,"tokens_out":13793,"duration_ms":140681,"concrete_test":"Exactly diagonalize L=12 and L=14 (if feasible) with the same parameters (J=1, hz=1, hx=0.2, ϕ=π), ≥1000 disorder samples, and report standard errors. Extract T2(L), T3(L) from magnetization, and the incremental OTOC ΔCzz(ℓ,t)=Czz(ℓ,t)−Czz(ℓ,T2) over T2<t<T3. Check: (i) is ΔCzz(ℓ,t)/ΔCzz(1,t) ℓ-independent at late times? (ii) do T2,T3 grow exponentially with L with stable β≈0.43, β′≈0.95? A 20% shift in β or loss of the ℓ-independent increment at L=12 would falsify the envelope claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that in the decoherence window T2<t<T3 the z-OTOC increments become distance-independent and merge into an envelope—is demonstrated only for L=8 (Fig. 1). The supporting time-scale hierarchy relies on fitting T2∼e^{βL} with β≈0.43 using only L=8 and L=10, giving T2≈31 and ≈74; the decoherence window spans roughly 1.5–2 decades at these sizes. At such small L, all distances are comparable to the system size and the wavefront has already crossed the chain, so the apparent collapse of Czz(ℓ,t) increments may simply reflect trivial convergence to the common thermal maximum rather than a genuine distance-independent scrambling rate. No error bars or L=10 OTOC curves are provided, and the extracted exponents β, β′ and the envelope exponent d are therefore unvalidated. If this is a finite-size artifact, the abstract's headline phenomenon would not survive in larger systems.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript studies information scrambling and entanglement growth in a disordered Floquet spin-1/2 chain in an MBL-stabilized discrete time crystal (FTC) phase. Using numerical simulations for L=8 and L=10 spin chains with 1000 disorder realizations, the authors report a hierarchy of stroboscopic time scales in the magnetization (initial decay, period-doubling plateau, decoherence, thermalization) and direction-dependent OTOC dynamics: Cxx shows MBL-like logarithmic growth while Czz remains quasi-frozen and later grows logarithmically, with all Czz(l,t) curves merging into a distance-independent envelope in the decoherence regime. The entanglement entropy is reported to grow logarithmically until thermal volume-law saturation. The paper also contains an analytic treatment of a fine-tuned no-scrambling limit (Appendix A) and a comparison with the ergodic phase (Appendix B).","tokens_in":16948,"tokens_out":11784,"duration_ms":125488,"significance":"If the reported phenomenology is robust, it would extend the MBL scrambling picture to Floquet time crystals and identify a genuinely new signature: a preferred quasi-protected direction with frozen-then-logarithmic scrambling and distance-independent late-time growth, distinct from both conventional MBL and thermal behavior. The paper's strengths are that it studies a concrete, experimentally relevant model, averages over 1000 disorder realizations, and connects the results to the established l-bit/MBL phenomenology. The main limitations are the small system sizes (L=8 and L=10), the fitted rather than derived growth laws, and the absence of finite-size scaling for the central OTOC envelope claim, so the conclusions currently rest on evidence that is suggestive but not yet demonstrative.","major_comments":[{"comment":"The central claim that in the decoherence window all Czz(l,t) curves merge into a distance-independent envelope is demonstrated only for L=8; no L=10 OTOC curves or finite-size collapse are shown. At L=8, all distances are comparable to the system size, and with the quoted exponents the decoherence window T2<t<T3 spans only about 1.8 decades (T2~31, T3~2000 for L=8; T2~74, T3~13000 for L=10). The apparent merging may therefore reflect trivial saturation to a common thermal maximum after the wavefront has crossed the chain, rather than a genuine distance-independent scrambling rate. Please provide Czz and Cxx data for L=10 and, if possible, L=12, with error bars, and show that the envelope structure persists away from the trivial saturation regime.","section":"III.B, Fig. 1"},{"comment":"The exponential time scales T2~exp(beta L) with beta~0.43 and T3~exp(beta' L) with beta'~0.95 are obtained from only L=8 and L=10. Two points cannot validate an exponential law, and the manuscript does not report the fitting procedure, residuals, or statistical uncertainties. This exponential hierarchy is load-bearing for calling the z-direction quasi-protected. Please compute T2(L) and T3(L) for at least three system sizes, or provide a scaling collapse, and report uncertainties on beta and beta'.","section":"III.A"},{"comment":"The exact OTOC calculation is internally inconsistent. It begins with hx=0 and phi=pi/2, but then uses U_K^dagger sigma^z U_K = -sigma^z, which holds only for phi=pi (for phi=pi/2, U_K^dagger sigma^z U_K = -sigma^y). It also defines the one-period evolution as (U_F U_K)^dagger sigma^z (U_F U_K), which does not match the Heisenberg evolution generated by U_F in Eq. (2). In addition, the disorder average in Eq. (A7) does not follow from Cxx(1,t)=1-cos(4 J0 t): the correct average is 1 - cos(4 J t) sin(2 J t)/(2 J t), not 1 - cos(6 J t) sin(2 J t)/(2 t). This appendix should be corrected or removed; as written, it does not support the statement that the perfect-FTC limit has no scrambling.","section":"Appendix A"},{"comment":"The l-bit interpretation is partly circular. The protected z-direction and the small overlap of sigma^z_i onto orthogonal tau^x_j and tau^y_j operators are inferred from the smallness of Czz itself, which is exactly the quantity the l-bit picture is invoked to explain. The coefficients c^{[i,alpha]}_{j,beta} in Eq. (7) are not computed independently. To make this explanation load-bearing, extract the l-bit overlaps from the model (for example, by constructing approximate l-bits or analyzing the diagonal ensemble) and show that they quantitatively reproduce the relative magnitudes of Cxx and Czz and the distance dependence of A_zz(l).","section":"III.B, Eq. (7)"},{"comment":"The claim that the entanglement entropy grows logarithmically over all time until thermal volume-law saturation is based on small system sizes and fitted scalings. In Fig. 4, the saturation time is reported to grow exponentially with L, but only a few system sizes are shown, no error bars are given, and the logarithmic-growth claim is not supported by a collapse or by an independent check of the time-window where the growth is purely logarithmic. Please provide error bars and a finite-size scaling analysis for S(t) and tsat(L), analogous to the analysis presented for the ergodic phase in Appendix B.","section":"III.C, Figs. 3-4"},{"comment":"The statement 'our results are independent of the choice of initial state' is asserted without systematic evidence. All reported scalings are obtained for a single initial product state with theta=pi/8. Since OTOC growth rates and entanglement growth in disordered systems can depend on the initial state's overlap with the l-bit structure, please demonstrate the key claims (the Czz envelope, the log-growth exponents, and the T2/T3 scaling) for at least one additional initial state, or explicitly restrict the claims to the class of initial states simulated.","section":"III, initial-state independence"}],"minor_comments":[{"comment":"The phrase 'entanglement of entropy' should be 'entanglement entropy'.","section":"Abstract"},{"comment":"The exact value of the kicking phase phi used in Figs. 1 and 4 is not stated (the text only says phi~pi); please give it explicitly, especially because Appendix A uses phi=pi/2.","section":"II and III"},{"comment":"The time scales T1, T2, and T3 are defined only qualitatively by visual inspection of Fig. 1; please provide operational definitions (for example, threshold crossings of Z(t)) so that the fits are reproducible.","section":"III.A"},{"comment":"Please reconcile the notation A_alpha beta(l) ~ exp(-v_alpha beta l) in Eq. (6) with Czz(l,t) ~ exp(-b_l l) + c log(t)^d; the subscript l in b_l is easily confused with the distance variable l.","section":"III.B"},{"comment":"Because S is defined with an explicit division by L_A, it is an entanglement entropy density rather than the total entanglement entropy; please state this explicitly, since the 'volume-law saturation' and comparisons with thermal values depend on the normalization.","section":"III.C, Eq. (8)"},{"comment":"The vertical dotted lines labeled T1, T2, and T3 should be defined in the caption; currently they are mentioned only in the text and their numerical values are not given.","section":"Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"The main risk is that the headline envelope phenomenon is a finite-size artifact: it is shown only for L=8, and the supporting time-scale hierarchy is fitted from L=8 and L=10 alone. I would use the revision to require explicit finite-size evidence, error bars, and operational definitions of the time scales. The Appendix A inconsistencies are also severe enough that the authors should rewrite that section carefully rather than leaving it as is. If the finite-size evidence cannot be provided, the contribution would be reduced to a qualitative observation on very small chains."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe new content here is the direction-dependent OTOC phenomenology in an MBL-stabilized Floquet time crystal: along the z/ℓ-bit direction the OTOC freezes during the period-doubling window, then grows logarithmically as the magnetization decoheres, and in the late regime all distances merge into a single distance-independent envelope. That envelope is the paper's most interesting claim and also its most fragile one.\n\nThe paper does several things well. The model is a standard kicked disordered Ising chain, the ED setup is clear, and the qualitative distinction between Cxx and Czz is convincing in the figures. The single-realization panels showing plateaus with quadratic growth are a nice touch. The ℓ-bit interpretation is reasonable and consistent with the prior MBL literature; the analytic appendix for the fine-tuned case is a useful sanity check.\n\nThe soft spots are real but mostly fixable. First, the envelope claim is demonstrated only for L=8; there are no L=10 OTOC curves for Czz, even though the time-scale hierarchy is fitted using L=8 and L=10. At L=8 all distances are comparable to the system size, so the collapse of the increments could partly reflect trivial convergence to the common thermal maximum. This needs a finite-size check before the abstract's headline survives. Second, the appendix has internal inconsistencies: it sets ϕ=π/2 but uses U_K^† σ^z U_K = −σ^z, which holds for ϕ=π, and the disorder average in Eq. (A7) does not follow from averaging cos(4J0 t) over J0 ∈ [J/2, 3J/2] — I get 1 − [sin(6Jt) − sin(2Jt)]/(4Jt). These are errors, not just typos in notation. Third, the claim that results are independent of the initial state is asserted without systematic checks. Fourth, the fitted growth exponents have no error bars; for a numerical phenomenology paper that is acceptable, but the precision should be stated.\n\nThe circularity concern is milder than the reader's note suggests. The ℓ-bit picture is a standard framework, and the OTOC is a different observable from the magnetization; using the magnetization to identify the protected direction and then interpreting the OTOC in that basis is a reasonable physics step. Still, the paper could acknowledge the dependence more explicitly.\n\nWho is this for: people working on scrambling, MBL, and driven systems. It does not open a new branch, but it gives a concrete characterization of information flow in FTCs that is absent from the cited literature. I would send this to a serious referee. The main revisions: add L=10 OTOC data for the envelope, correct the appendix, and report statistical precision. A desk reject would be wrong; the qualitative claims are plausible and the topic is active.","headline":"Plausible numerics for a new direction-dependent scrambling signature in Floquet time crystals, but the headline envelope rests on L=8 alone and the appendix has fixable errors.","tokens_in":17455,"tokens_out":3321,"would_cite":true,"duration_ms":34016,"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":"In Floquet time crystals, information scrambling is frozen first, then grows logarithmically along a protected direction.","keywords":["out-of-time-ordered correlators","Floquet time crystal","many-body localization","ℓ-bit","entanglement entropy","information scrambling","disordered spin chain","period doubling"],"falsifier":"Measure the OTOC curves $C_{zz}(\\ell,t)$ at system sizes beyond $L=10$: if the late-time growth rates for different distances do not collapse onto one envelope, or if the period-doubling window $T_2$ does not grow exponentially with $L$, the central claim collapses.","tokens_in":16502,"feed_emoji":"⚛️","tokens_out":5448,"duration_ms":54257,"temperature":0.7,"pith_summary":"This paper asks how quantum information spreads in a disordered spin chain that forms a Floquet time crystal, a phase that oscillates with twice the period of its driving. The authors compute out-of-time-ordered correlators and entanglement entropy and find that the time crystal has a preferred 'quasi-protected' direction along which scrambling is unusually slow. Along that direction, the correlator stays frozen while the period-doubling order is stable, then grows logarithmically as the order melts, and finally all spin distances collapse onto a single distance-independent growth curve. Along the orthogonal direction, scrambling looks like that of a usual many-body localized system. If true, the results show that time-crystalline order leaves a fingerprint on how quantum information gets scrambled, and that the melting of the time crystal can be read off from the merging of OTOC curves.","feed_headline":"In time crystals, information scrambling is frozen, then logarithmic","feed_subtitle":"A protected spin direction keeps correlations pinned, then all distances merge into one growth curve.","key_machinery":"The argument is carried by the $\\ell$-bit phenomenological picture of many-body localization, applied to a Floquet system. The local spin operators are expanded in a basis of quasi-local integrals of motion $\\hat\\tau_j^\\alpha$ with exponentially decaying coefficients; the preferred z-direction is the one along which the period-doubling magnetization is protected. The OTOC, defined as the squared commutator $C_{\\alpha\\beta}(\\ell,t)=\\frac12\\langle[\\hat\\sigma^\\alpha_\\ell(t),\\hat\\sigma^\\beta_1]^2\\rangle$, measures how much of the operator leaks into orthogonal directions. The timescales $T_1$, $T_2$, $T_3$ — set by the exponential stability of the period-doubling order — organize the different behaviors. In the exactly solvable limit of perfect spin flips and no transverse field, the OTOCs vanish identically, which serves as the reference point from which the slow scrambling develops.","core_discovery":"The paper's central claim is that in a Floquet time crystal stabilized by many-body localization, the out-of-time-ordered correlator $C_{zz}(\\ell,t)$ along the quasi-protected $\\ell$-bit direction evolves through three distinct regimes. For times $t<T_1$ it undergoes the same initial relaxation seen in all phases. During the stable period-doubling window $T_1<t<T_2$, the correlator is essentially frozen at small, distance-dependent plateaus, because a conserved piece of the $\\hat\\sigma^z$ operator overlaps with the local integrals of motion. In the decoherence regime $T_2<t<T_3$, the correlator grows logarithmically, $C_{zz}(\\ell,t)\\sim e^{-b_\\ell \\ell} + c \\log(t)^d$, with the logarithmic term independent of distance once the correlation wavefront has traversed the chain. In the late thermal regime, all distances merge into a single envelope and saturate. The entanglement entropy grows logarithmically throughout and saturates to a thermal volume law at exponentially long times.","pith_inferences":["The collapse of all $C_{zz}(\\ell,t)$ curves onto one envelope could serve as a practical diagnostic for the onset of time-crystal melting, since the merging time should correlate with $T_2$.","The anisotropy in scrambling suggests that other quantities, such as operator spreading widths or Krylov complexity, will also be direction-dependent inside the FTC; the paper does not compute those.","A natural testable extension is to drive the system closer to the ergodic side of the transition, where the prediction would be that the envelope merging disappears and is replaced by the usual ballistic light cone.","The initial-state-independence claim could be checked against tilted or N\\'eel initial states; if logarithmic growth rates vary with the tilt angle, the $\\ell$-bit description may need refinement."],"forward_implications":["The protected $\\ell$-bit direction makes scrambling strongly anisotropic: $C_{xx}$ can be orders of magnitude larger than $C_{zz}$ while the magnetization still exhibits period doubling.","The logarithmic growth of entanglement entropy persists over the entire prethermal window, so the time crystal is 'slow' not only in its magnetization but also in how it builds up entanglement.","The late-time distance independence of $C_{zz}$ means that once the wavefront has crossed the chain, the scrambling rate is set by the decoherence process itself, not by the distance between probes.","OTOC measurements on existing superconducting-qubit platforms that realize Floquet time crystals could observe the frozen plateau and the later envelope merge directly."],"supporting_citations":[{"why":"Establishes the Floquet time crystal phase and its period-doubling order, which this work studies.","marker":"[2]"},{"why":"Supplies the stability result that MBL prevents heating in Floquet systems, justifying the exponentially long period-doubling time.","marker":"[90]"},{"why":"Provides the $\\ell$-bit phenomenological description of MBL that the paper uses to explain the quasi-protected direction and frozen OTOC.","marker":"[94]"},{"why":"Shows the typical logarithmic growth of the out-of-time-ordered commutator in MBL systems, the comparison point for the slow scrambling regime.","marker":"[92]"},{"why":"Demonstrates the logarithmic growth of entanglement entropy in MBL, which the paper finds persists through the FTC decoherence regime.","marker":"[87]"},{"why":"Supplies the scrambling wavefront conjecture used to characterize the ergodic-phase OTOC and the late-time distance-independent growth.","marker":"[72]"},{"why":"Gives the interferometric protocol for measuring OTOCs on superconducting processors, the platform the paper proposes for experimental observation.","marker":"[66]"}],"fun_headline_variants":["Time crystals freeze information, then scramble it slowly","Information spreading in Floquet time crystals: frozen, then logarithmic","Floquet time crystals: scrambling freezes, then grows logarithmically","In time crystals, correlations stay frozen, then all distances merge","Scrambling in Floquet time crystals: a frozen start, then slow growth"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the assumption that the Floquet time crystal at the simulated parameters is properly described by a complete set of local integrals of motion — the $\\ell$-bits — with a preferred z-direction, and that the period-doubling time grows exponentially with system size; the numerics only reach $L=8$ and $L=10$, so this exponential scaling and the $\\ell$-bit structure are inherited from prior MBL phenomenology rather than demonstrated here.","fun_headline_variants_meta":{"raw":{"variants":["Time crystals freeze information, then scramble it slowly","Information spreading in Floquet time crystals: frozen, then logarithmic","Floquet time crystals: scrambling freezes, then grows logarithmically","In time crystals, correlations stay frozen, then all distances merge","Scrambling in Floquet time crystals: a frozen start, then slow growth"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000208,"raw_usage":{"total_tokens":1428,"prompt_tokens":996,"completion_tokens":432,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":341}},"tokens_in":612,"tokens_out":432,"duration_ms":4501,"temperature":1.0,"reasoning_tokens":341,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:22:32.671860+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the OTOC curves $C_{zz}(\\ell,t)$ at system sizes beyond $L=10$: if the late-time growth rates for different distances do not collapse onto one envelope, or if the period-doubling window $T_2$ does not grow exponentially with $L$, the central claim collapses.","supporting_citations":[{"cited_title":"Sierant, M","cited_arxiv_id":null,"evidence_quote":"Supplies the stability result that MBL prevents heating in Floquet systems, justifying the exponentially long period-doubling time."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the $\\ell$-bit phenomenological description of MBL that the paper uses to explain the quasi-protected direction and frozen OTOC."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows the typical logarithmic growth of the out-of-time-ordered commutator in MBL systems, the comparison point for the slow scrambling regime."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates the logarithmic growth of entanglement entropy in MBL, which the paper finds persists through the FTC decoherence regime."},{"cited_title":"Xu and B","cited_arxiv_id":null,"evidence_quote":"Supplies the scrambling wavefront conjecture used to characterize the ergodic-phase OTOC and the late-time distance-independent growth."}],"review_version":1}