{"id":"f85c9279-ebc6-4349-82bb-de3db170d3be","arxiv_id":"2412.04382","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Adding XX interactions to a non-Hermitian Floquet Ising chain shifts its time-crystal phase boundaries and, above K_c ≈ 0.085, stabilizes a new x-ferromagnetic phase at R=1.","lead":"This paper studies what happens when an interaction between neighboring spins is added to a non-Hermitian, periodically driven Ising chain that can host time-crystalline order. The interaction shifts the phase boundaries and, beyond a critical strength, creates a new symmetry-breaking transition in the steady state.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Leading-order average Hamiltonian truncation, not the full Floquet dynamics, is what produces Kc≈0.085; higher-order terms break the artificial U(1)/integrability and may destroy the predicted transition and NFM1 survival.","rationale":"I read the paper in good faith. Its goal is to show that a non-integrable XX interaction in a non-Hermitian Floquet Ising model shifts phase boundaries and, at sufficiently large K, produces a new symmetry-broken x-ferromagnetic transition, while the NFM1 time crystal survives for K>Kc. The evidence is a combination of TEBD, mean-field theory, and a leading-order average Hamiltonian. The TEBD and mean-field results independently corroborate the phase-boundary shift, and the TEBD jump at K=0.1 near R=1 is a genuine numerical observation. I did not find an algebraic inconsistency in the leading-order derivation of Eqs. A28-A29. The concern is that the paper's most novel quantitative claims—the value Kc≈0.085 and the stability of NFM1 for K>Kc—depend entirely on the truncation of the BCH series at first order, and this truncation is not justified by an error estimate. At Kc the expansion parameters are not asymptotically small, and the truncated Hamiltonian has emergent symmetry and integrability properties that the exact Floquet operator does not share. Higher-order terms can therefore be not merely small quantitative corrections but relevant perturbations that shift or destroy the predicted phase. The reader's weakest assumption already identified the truncation issue; my stress test sharpens it by noting the U(1)/integrability artifacts and by pointing out that Fig. 7 is a property of Heff alone. Because the qualitative picture is plausible and partially supported by TEBD, the correct verdict remains conditional rather than accept or reject. The proposed numerical test would directly determine whether the truncation, not the full dynamics, is responsible for the predicted transition.","tokens_in":15191,"tokens_out":12249,"duration_ms":208299,"concrete_test":"Compute the second-order Magnus/BCH correction to U^6 in the representation of Appendix A (e.g., by symbolic or numeric exponentiation for L=6-10), and compare the norm of the U(1)-breaking part of Heff^(2) with the leading coefficients at K=0.1, R≈1. Then exactly diagonalize the full U^6 and e^{-6iHeff} for L=8-12 and compare ⟨X⟩_ss for K=0.05, 0.085, 0.15 and Re(Δγ)=±0.0001. If the mismatch is comparable to the jump size, or if the U(1)-breaking correction is a relevant operator at the XXZ fixed point, then the predicted Kc and the NFM1 survival for K>Kc are not established by the leading-order truncation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central new claim is that the non-integrable perturbation K∑XX induces a ferromagnetic transition at Kc≈0.085 that is absent in mean-field theory and that the NFM1 time-crystalline phase survives for K>Kc. The load-bearing step is the identification of this transition with the steady state of the leading-order average Hamiltonian in Appendix A (Eqs. A28-A29): Heff = i∑[Δγ X + γK XX + (J/2)(YY+ZZ)]. Every subsequent prediction—Kc, the XFM phase, and its stability—is a consequence of this operator. The derivation assumes |Δγ|,|K|,|J| much less than π and drops all Baker-Campbell-Hausdorff terms beyond first order, but no error bound or explicit higher-order estimate is supplied. This is not a purely cosmetic omission. At Kc≈0.085, the leading coefficients γK and J/2 are comparable, so the words 'much less' are not satisfied in the regime where the transition is predicted. Moreover, the truncated Heff has an emergent continuous U(1) symmetry (rotation about the x-axis) and is integrable (XXZ). Neither property belongs to the original Floquet circuit; both are artifacts of keeping only the first Magnus order. The exact U has only the discrete symmetries of the period-6 drive, so higher-order corrections generically contain U(1)-breaking terms. In one dimension, such terms are often relevant at the Luttinger-liquid fixed point and can open a gap, which is exactly the mechanism that would destroy the NFM1 phase the paper predicts to survive for K>Kc. Figure 7, which sets Kc≈0.085, is computed only from Heff, not from the full Floquet evolution; the TEBD data show a jump for K=0.1 but do not determine Kc or demonstrate survival of NFM1 away from the R=1 line. Thus the existence and location of the new transition are supported by the numerics only qualitatively, while the quantitative phase diagram and the stability claim rest on the unvalidated first-order average Hamiltonian.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a non-Hermitian Floquet transverse-field Ising model with an added non-integrable interaction K∑XX. Using TEBD, mean-field theory, and average Hamiltonian theory, it claims two effects: a shift of the PM/NFM1 phase boundaries, and a new symmetry-breaking transition near R=1 that is absent in mean-field theory. The new transition is attributed to an x-ferromagnetic phase of a leading-order non-Hermitian XXZ effective Hamiltonian, with a critical point Kc≈0.085, and the NFM1 time-crystalline phase is predicted to survive for K>Kc.","tokens_in":15470,"tokens_out":5862,"duration_ms":61367,"significance":"If established, these results would show that integrability-breaking interactions can qualitatively change non-Hermitian Floquet phase diagrams and connect a new time-crystal-related transition to well-studied non-Hermitian XXZ physics. The paper is clearly written and combines complementary methods: TEBD data, a mean-field analysis that correctly captures the boundary shift, and an effective-Hamiltonian derivation with no fitted parameters. The main limitation is that the central new transition is inferred from a leading-order truncation of the Floquet expansion without controlling higher-order terms, and the TEBD evidence for the transition is not backed by finite-size scaling or error estimates.","major_comments":[{"comment":"The identification Kc≈0.085 is made from the leading-order average Hamiltonian, but the validity of this truncation at the predicted transition is not controlled. At K=0.1 the expansion parameters are |J|≈0.096 and |γK|≈0.052, which are not small compared with the terms whose competition sets Kc, while the detuning used in Fig. 7 is |Re Δγ|=10^-4. Second-order BCH terms are of order JγK≈0.005, which is fifty times larger than this detuning. The paper needs an explicit estimate or computation of the leading higher-order corrections to Eq. (14), or a benchmark of Kc by exact numerics on the full circuit, before the transition can be attributed to the full Floquet dynamics.","section":"§III B and Appendix A, Eq. (14), Fig. 7"},{"comment":"The new phase transition is inferred from a jump in the steady-state expectation value ⟨X⟩ at a single system size L=50, without finite-size scaling or error bars. A first-order transition in one dimension should sharpen with system size, and a jump at L=50 could in principle be a finite-size crossover. The authors should show ⟨X⟩ as a function of R for several system sizes (and confirm bond-dimension convergence at each size), or explicitly state the finite-size limitations of the evidence for the transition.","section":"§III B, Figs. 4 and 6"},{"comment":"The prediction that NFM1 survives for K>Kc is based on the stability of the Luttinger liquid against weak integrability-breaking perturbations. However, the leading-order Heff in Eq. (14) has a continuous U(1) symmetry about the x-axis and is integrable, while the exact Floquet circuit has neither property; the higher-order terms that break these properties are exactly the terms whose magnitude is not controlled. To support the survival claim, the paper should provide numerical evidence for power-law temporal correlations C(N) in the full circuit for K>Kc, or a controlled argument that the non-Hermitian steady state is insensitive to the U(1)-breaking perturbations.","section":"§III B, final paragraph and Fig. 8"}],"minor_comments":[{"comment":"There appears to be a factor-of-two inconsistency between the first line of Eq. (10), which has 2Kγ in the exponent, and the mean-field expression that follows, which has γK without the factor 2; please clarify the convention.","section":"§III A, Eq. (10)"},{"comment":"The caption says the time evolution of ⟨X⟩ is shown for a few values of R but does not state which R values are plotted; please include them in the caption or legend.","section":"Fig. 4 caption"},{"comment":"The caption 'Expectation value from Heff and H' is ambiguous; H should be identified as the full Floquet circuit or the TEBD result, and the curves should be clearly labeled.","section":"Fig. 6 caption"},{"comment":"There are several typos, including 'considitions' in the introduction, 'differant' in §III, and 'demostrate' in §III B; these should be corrected.","section":"Introduction and §III B"},{"comment":"The sentence 'This confirms that the phase structure of Heff is imprinted on that of the full interacting Floquet circuit' is too strong, because Fig. 7 shows only Heff; the comparison to the full circuit is made in Fig. 6 and is qualitative.","section":"§III B"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper extends the non-Hermitian Floquet Ising model from Basu et al. by adding an XX interaction. The two main results are a shift of the PM-NFM1 boundaries and a new symmetry-breaking jump near R=1. The shift is convincingly established by TEBD and mean-field. The jump is qualitatively reproduced by a leading-order average Hamiltonian that maps to a non-Hermitian XXZ chain, and the paper connects the time-crystal phase to the easy-plane Luttinger liquid of that XXZ model. That mapping is the most valuable piece here.\n\nWhat is genuinely new: the interaction term and the resulting transition are absent from the prior work; the effective-Hamiltonian derivation is new for this model and is done cleanly in Appendix A. The paper is honest about the limitations of TEBD for power-law correlations and does not oversell the numerics.\n\nThe soft spots are real but not fatal. The critical point Kc≈0.085 is computed from Heff alone (Figure 7); the TEBD data show a jump at K=0.1 but with no finite-size scaling or error bars, so the location of the transition is not directly established. More importantly, the prediction that the NFM1 phase survives for K>Kc is an extrapolation based on the stability of the Luttinger liquid in the effective model. The leading-order Heff has an artificial U(1) symmetry and is integrable; higher-order BCH terms break both. The paper gives no estimate of these terms. In one dimension such perturbations can be relevant, so the survival claim is not safe without further analysis. That said, the neglected terms are small in the parameter region considered, and the qualitative picture is plausible. This is an addressable gap, not a demonstrated error.\n\nMissing data/code is a minor annoyance; providing it would help referees check the TEBD convergence claims.\n\nVerdict: deserve a serious referee. I'd send it out and ask for (i) a higher-order estimate or a full-Floquet check of Kc, (ii) finite-size scaling for the jump, and (iii) a more careful statement about the stability of NFM1. This is a solid contribution within an established program, not a breakthrough, and the referee time is justified.","headline":"A solid extension of the non-Hermitian Floquet time-crystal program; the phase-boundary shift is well supported, but the new transition's location and the stability claim for K>Kc rest on a leading-order effective Hamiltonian that needs higher-order checks.","tokens_in":16166,"tokens_out":3029,"would_cite":true,"duration_ms":29365,"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":"Adding a non-integrable transverse interaction to a non-Hermitian Floquet Ising model shifts its phase boundaries and, above a critical strength $K_c\\approx 0.085$, drives the steady state into a new x-ferromagnetic phase that mean-field…","keywords":["non-Hermitian time crystal","Floquet systems","integrability breaking","non-Hermitian XXZ model","symmetry-breaking transition","average Hamiltonian theory","quasi-long-range order","transverse-field Ising model"],"falsifier":"A decisive check is to compute the first neglected Baker--Campbell--Hausdorff correction to $H_{\\text{eff}}$ at $K=0.1$, $J\\approx 0.1$, and $\\Delta\\gamma=10^{-4}$ and compare its operator norm with the norms of $\\Delta\\gamma X_j$ and $\\gamma K X_jX_{j+1}$: if the correction is not at least an order of magnitude smaller, the leading-order average Hamiltonian is not a faithful description and the predicted transition is unproven.","tokens_in":14948,"feed_emoji":"🕐","tokens_out":15988,"duration_ms":127668,"temperature":0.7,"pith_summary":"This paper asks whether the non-Hermitian Floquet time crystal survives the addition of an integrability-breaking interaction, and what new physics that interaction brings. Working with a non-Hermitian Floquet transverse-field Ising chain plus an $X$-$X$ coupling of strength $K$, the authors find two effects: the phase boundaries between paramagnet and time-crystalline phases shift with $K$, and above $K_c\\approx 0.085$ a new symmetry-breaking transition appears at the special point $R=1$, visible as a jump in the steady-state transverse magnetization. Mean-field theory captures the boundary shift but completely misses the jump. Using a rotating-frame average-Hamiltonian expansion, the authors show that the full Floquet circuit reduces at leading order to a non-Hermitian XXZ model, and that the new transition is the easy-axis ferromagnetic transition of that model. They predict that the quasi-long-range time-crystalline (NFM1) phase survives finite $K$, even above $K_c$.","feed_headline":"Interactions create a new phase in a non-Hermitian time crystal","feed_subtitle":"Past a critical coupling, the driven spin chain develops a new magnetically ordered phase that mean-field theory misses.","key_machinery":"The rotating-frame average-Hamiltonian (Floquet high-frequency) expansion around the fine-tuned point $\\tanh\\beta=e^{i\\pi/3}$ with $J=K=0$. Because $U_0^6\\propto 1$, one considers the super-cycle $U^6$, moves to the rotating frame, and keeps only the leading Baker--Campbell--Hausdorff term, obtaining the non-Hermitian XXZ Hamiltonian of Eq. (14). This effective Hamiltonian carries the argument: its easy-plane versus easy-axis phase structure explains the NFM1 time-crystal phase and the new x-ferromagnetic transition, and its steady-state magnetization reproduces the TEBD jump.","core_discovery":"At the special point $R=1$, where the transverse field is purely imaginary with $\\tanh\\beta=e^{i\\pi/3}$, the six-cycle Floquet operator $U_0^6$ is proportional to the identity, so $J$, $K$, and $\\Delta\\gamma=\\gamma-\\gamma_0$ can be treated as weak perturbations. Time-averaging the rotating-frame circuit at leading order yields an effective Hamiltonian $H_{\\text{eff}}=i\\sum_j[\\Delta\\gamma\\,X_j+\\gamma K\\,X_jX_{j+1}+\\tfrac12 J(Z_jZ_{j+1}+Y_jY_{j+1})]$, which is an anisotropic non-Hermitian XXZ chain with quantization axis $x$. In this description, the non-Hermitian time-crystal phase (NFM1) corresponds to the easy-plane Luttinger-liquid regime, while the integrability-breaking $X$-$X$ term drives an easy-axis ferromagnetic transition. TEBD data for the steady-state magnetization $\\langle X\\rangle_{\\text{ss}}$ show a jump near $R=1$ for $K\\gtrsim K_c\\approx 0.085$, and the same jump appears in the steady state of $H_{\\text{eff}}$. The paper concludes that this x-ferromagnetic phase is present in the full Floquet circuit, that mean-field theory misses it because it neglects the relevant fluctuations, and that the NFM1 time crystal remains stable for finite $K$, even above $K_c$.","pith_inferences":["If $H_{\\text{eff}}$ is quantitatively reliable at $K\\gtrsim K_c$, the same jump should appear in other steady-state observables, for example the $X$-$X$ structure factor or the overlap with a symmetry-broken product state; these provide a direct TEBD falsifiability check.","The super-cycle construction uses $\\theta=\\pi/3$ so that six Floquet steps return to the identity; for irrational $\\theta$ no such finite super-cycle exists, and the paper leaves open whether the transition survives as a smooth boundary or fragments into a fractal one -- a numerical scan over nearby rational angles could settle this.","Because the mean-field decoupling misses the transition entirely, a symmetry-broken mean-field ansatz that allows spontaneous $x$-magnetization might capture the jump while keeping the free-fermion sector exact, isolating which fluctuations are essential.","The predicted survival of NFM1 above $K_c$ implies the temporal correlation exponent should remain power-law; measuring $C(N)$ at larger system sizes and bond dimensions for $K>K_c$ could confirm or refute that stability."],"forward_implications":["The PM--NFM1 phase boundaries move systematically with $K$, and a self-consistent mean-field decoupling that renormalizes the complex transverse field captures this shift at small and moderate $K$.","For $K>K_c\\approx 0.085$, the steady state at $R=1$ develops a jump in $\\langle X\\rangle_{\\text{ss}}$, a first-order-like symmetry-breaking transition into an x-ferromagnetic phase that is absent at $K=0$ and invisible to mean-field theory.","The NFM1 time-crystal phase, with quasi-long-range oscillatory temporal correlations, is predicted to survive finite $K$, including $K>K_c$, because it maps to the stable Luttinger-liquid regime of the effective XXZ chain.","Near $K_c$, the transition should show conventional XXZ criticality, with correlation functions possibly acquiring damping factors; weak symmetry-respecting perturbations, including higher-order Floquet terms, are expected not to destroy the phase structure.","The new x-ferromagnet is distinct from the $z$-ferromagnet present at $K=0$ and requires the integrability-breaking $X$-$X$ interaction to be realized."],"supporting_citations":[{"why":"Defines the non-Hermitian Floquet Ising model, its free-fermion integrability, and the PM/FM/NFM1/NFM2 phase diagram that this paper extends.","marker":"[18]"},{"why":"Establishes Luttinger-liquid behavior in the non-Hermitian XXZ chain, the basis for identifying the NFM1 time crystal with the easy-plane regime of the effective Hamiltonian.","marker":"[37]"},{"why":"Provides the tensor-network time-evolution implementation used for all TEBD correlation and steady-state data.","marker":"[33]"},{"why":"Characterizes nonequilibrium steady states of the quantum XXZ chain, supporting the steady-state analysis of the effective model.","marker":"[35]"},{"why":"Supplies the XXZ spin-chain representation used to interpret the half-filled easy-axis regime.","marker":"[36]"},{"why":"Gives the Jordan-Wigner mapping that underlies the free-fermion integrability of the $K=0$ model.","marker":"[32]"}],"fun_headline_variants":["Unexpected ferromagnetism in non-Hermitian time crystal","Interactions shift phases and unlock order in time crystal","Mean-field misses symmetry breaking in non-Hermitian time crystal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole analysis rests on the assumption that the leading-order rotating-frame average Hamiltonian of Appendix A (Eqs. A28--A29) faithfully describes the full Floquet circuit at the parameters used ($K=0.1$, $J\\sim 0.1$, $|\\Delta\\gamma|\\le 10^{-4}$ near $R=1$), with all higher-order correction terms in the Floquet expansion negligible; no error bound is given for that truncation.","fun_headline_variants_meta":{"raw":{"variants":["Unexpected ferromagnetism in non-Hermitian time crystal","Interactions shift phases and unlock order in time crystal","Mean-field misses symmetry breaking in non-Hermitian time crystal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000716,"raw_usage":{"total_tokens":3256,"prompt_tokens":1018,"completion_tokens":2238,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":634,"completion_tokens_details":{"reasoning_tokens":2184}},"tokens_in":634,"tokens_out":2238,"duration_ms":19475,"temperature":1.0,"reasoning_tokens":2184,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:24:33.361858+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check is to compute the first neglected Baker--Campbell--Hausdorff correction to $H_{\\text{eff}}$ at $K=0.1$, $J\\approx 0.1$, and $\\Delta\\gamma=10^{-4}$ and compare its operator norm with the norms of $\\Delta\\gamma X_j$ and $\\gamma K X_jX_{j+1}$: if the correction is not at least an order of magnitude smaller, the leading-order average Hamiltonian is not a faithful description and the predicted transition is unproven.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the non-Hermitian Floquet Ising model, its free-fermion integrability, and the PM/FM/NFM1/NFM2 phase diagram that this paper extends."},{"cited_title":"Non-Hermitian Discrete Time Crystals","cited_arxiv_id":"2410.22713","evidence_quote":"Establishes Luttinger-liquid behavior in the non-Hermitian XXZ chain, the basis for identifying the NFM1 time crystal with the easy-plane regime of the effective Hamiltonian."},{"cited_title":"Autti, V","cited_arxiv_id":null,"evidence_quote":"Provides the tensor-network time-evolution implementation used for all TEBD correlation and steady-state data."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Characterizes nonequilibrium steady states of the quantum XXZ chain, supporting the steady-state analysis of the effective model."}],"review_version":1}