{"id":"ffe81f58-ef3a-48fb-8c8d-998e4d887944","arxiv_id":"2608.11316","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"In small lattice lambda phi^4 theory, a highly occupied coherent field produces particles and relaxes to the diagonal-ensemble values for a finite time before quantum recurrences appear, and the dynamics maps efficiently onto quantum circuits.","lead":"Simulations of a small lattice scalar field theory show that energy initially stored in a classical field is converted into particles, after which the system appears equilibrated for a while before oscillating again. The paper offers a proof of principle that quantum-computing methods could one day simulate such nonequilibrium quantum field dynamics from first principles.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equilibration intervals and recurrence times are computed at one local truncation (n_Q=6 / fixed N_max) and one bond dimension (chi_max=25); without convergence checks the headline plateau may be a finite-truncation effect, especially for N=7-11.","rationale":"Read in good faith: the paper is a careful proof-of-principle numerical study; the N=2 exact/ensemble agreement and the N=5 TN benchmark are genuine internal consistency checks. For the central claim to be a statement about lattice lambda-phi^4 QFT rather than about one finite matrix model, two conditions must hold: (i) the local Hilbert-space truncation must not change the dephasing/recurrence phenomenology, and (ii) the tensor-network bond-dimension truncation must not determine the N-scaling trend. Neither is demonstrated. The reader already flagged (i); the present pass adds that (ii) is the sharper, testable weak point because Fig. 3 compares different N at a fixed chi_max without a convergence series. I would keep the CONDITIONAL verdict: the existence proof at N=2 is internally sound, but the scope claim is conditional on cutoff stability. This concern is not an accusation of error; it is a request for a convergence check that the paper's own open questions acknowledge.","tokens_in":17433,"tokens_out":10564,"duration_ms":101527,"concrete_test":"Run the N=11 tensor-network simulation in Fig. 3 with chi_max=50 and 100 (same parameters, same delta_t) and overlay phi_cl, P, and n; if the residual oscillation amplitude or the plateau endpoints change beyond line width, the fixed-chi_max=25 trend is not converged. Separately, rerun the N=2 exact calculation with n_Q=8 (or larger N_max) and re-extract the equilibration intervals and the phi=2 recurrence time; if these shift by more than the quoted ranges, the local truncation controls the headline numbers.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical claim - that a well-defined equilibration plateau lasting several times the initial relaxation time precedes quantum recurrences - is established only within a fixed, finite model. In Fig. 1 and Fig. 2, the 'exact' calculations use a truncated oscillator/qubit encoding (n_Q=6 per site for N=2), and the diagonal ensemble in Eq. (9) is computed in that same truncated space. Any finite closed system will dephase to its diagonal ensemble on a timescale set by its discrete spectrum; the nontrivial physical statement is that this truncated spectrum represents lattice lambda-phi^4. No convergence with n_Q or N_max is shown, so the plateau lengths quoted (e.g., t=11.8-18.5 for phi=5; 14.2-63.2 for phi=10) and the recurrence at t about 411.6 for phi=2 could be peculiar to the cutoff. For N=5,7,9,11 the situation is sharper: all larger-N results rely on TDVP with fixed chi_max=25 (Fig. 3 caption). The N=5 validation against direct exponentiation (Supplemental Fig. 7) checks only that one point; it cannot certify chi_max=25 for N=11, where entanglement is larger. Bond-dimension truncation removes Schmidt weight and can mimic dissipation, damping residual oscillations and extending the apparent plateau - exactly the qualitative trend reported as system-size dependence. The Conclusion itself lists larger volumes and the continuum limit as open, so the present evidence supports existence of dephasing and recurrences in a small truncated model, not a lattice-QFT property that is stable under cutoff variation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies particle production and equilibration starting from highly occupied coherent states in lattice λφ^4 theory in 1+1 dimensions. Numerically, it uses exact diagonalization for two- and five-site lattices, ideal quantum-circuit simulation (Qiskit) for two sites, and tensor-network (TDVP) evolution for up to eleven sites. The central observation is that the field expectation value, pressure, and occupation numbers relax to the diagonal-ensemble values on a timescale of order tens of m^{-1}, remain close to these values for an interval several times longer, and then display revivals interpreted as quantum recurrences; with increasing system size the residual oscillations become weaker. The paper also provides a resource estimate for a fault-tolerant quantum algorithm for the same dynamics.","tokens_in":17828,"tokens_out":6298,"duration_ms":108019,"significance":"Taken as statements about the finite truncated models actually simulated, the paper is a clean proof of principle: it benchmarks a quantum-circuit construction against exact evolution, validates TDVP against exact diagonalization at N=5, and makes a parameter-free comparison with the diagonal ensemble. The qualitatively new observation—dephasing to the diagonal ensemble followed by recurrences in a relativistic scalar lattice theory with high occupancy—is potentially interesting for the discussion of thermalization in small systems relevant to heavy-ion physics. However, the extrapolation from these small, fixed-cutoff lattices to lattice λφ^4 quantum field theory proper and to quark-gluon droplets is not yet supported; the missing convergence tests are the main gap.","major_comments":[{"comment":"All quantitative results in Fig. 2—the 1% equilibration windows (t=11.8–18.5 for φ=5, t=14.2–63.2 for φ=10) and the recurrence at t≈411.6 for φ=2—are computed at a single local truncation (n_Q=6 qubits per site for the exact/qubit calculations, and an unspecified fixed N_max in the HO-basis runs), with no convergence study in the truncation level. Any finite closed quantum system dephases to its diagonal ensemble on a timescale set by its discrete spectrum, so the physical statement that this is the dynamics of lattice λφ^4 rather than an artifact of the cutoff requires showing that the plateau boundaries and recurrence time are stable as n_Q (or N_max) is increased. Without such a check, the quoted timescales cannot be claimed as properties of the lattice model.","section":"Equilibration and quantum recurrences in small systems (Fig. 2)"},{"comment":"The N=7, 9, and 11 results rely entirely on TDVP with fixed maximal bond dimension χ_max=25, and no bond-dimension convergence test or discarded-weight/truncation-error estimate is reported. Since bond truncation removes Schmidt weight and can mimic dissipation by damping residual oscillations, the reported trend of progressively weaker oscillations and longer-lived equilibration with increasing N could be an artifact of fixed χ_max. The N=5 benchmark in Supplemental Fig. 7 validates only one system size and one parameter set; it does not certify χ_max=25 for N=11, where entanglement is larger. The authors should provide χ_max scans (e.g., 25, 50, 100) and report quantitative equilibration intervals for the larger systems rather than relying on visual inspection.","section":"Quantum equilibration in larger systems (Fig. 3)"},{"comment":"The displacement-operator circuit uses the Baker–Campbell–Hausdorff approximation e^{A+B} ≈ e^A e^B, and the text states that the omitted commutator terms are 'numerically negligible for the parameters considered' without showing the magnitude. Because the truncated finite-dimensional operators do not satisfy canonical commutation relations exactly, this step should be quantified (e.g., by reporting the fidelity between the approximate and exact displacement operators, or the size of the leading commutator correction as a function of n_Q and α). The exact quantum-circuit benchmark at N=2, n_Q=6 provides some validation, but it does not establish that the approximation remains valid for the larger n_Q values envisioned in the resource estimate.","section":"Supplemental Material, Eq. (S16)"}],"minor_comments":[{"comment":"The phrase 'values predicted by the diagonal ensemble' could be misread as an independent prediction; since ρ_eq is constructed from the same Hamiltonian and initial state used for the evolution, the late-time agreement is a diagnostic of dephasing in the truncated spectrum rather than a test of a thermal ensemble. Please rephrase to make this distinction explicit and to connect with the Gibbs-ensemble question already listed as open in the Conclusion.","section":"After Eq. (9)"},{"comment":"The statements that residual oscillations 'become progressively weaker' and that the equilibration stage 'persists for longer times' as N increases are made without quantitative definitions. Please provide the same kind of numerical equilibration windows used for the N=2 case, or clearly state the criterion applied to the tensor-network data.","section":"Fig. 3 and text below it"},{"comment":"The local Hilbert-space dimensions used for the N=5 exact diagonalization and for the HO-basis tensor-network runs are not specified; please state the N_max values used per momentum mode (and whether they are mode-dependent) so that the truncation is reproducible.","section":"Setup and Supplemental Material"},{"comment":"What is identified at t≈411.6 for φ=2 is a revival of a single observable (φ_cl) within a 10% window, not a true Poincaré recurrence of the quantum state. Calling it 'the first recurrence' is acceptable if clearly defined, but the text should distinguish this partial observable revival from a full state recurrence to avoid overstatement.","section":"Recurrence discussion, paragraph after Fig. 2"},{"comment":"The abstract's phrase 'first-principles framework for nonequilibrium quantum-field dynamics' is stronger than what is demonstrated, given the small lattices and fixed truncations; the Conclusion already appropriately lists larger volumes and the continuum limit as open. Qualifying the claim with 'on small lattices' would better match the evidence.","section":"Abstract and Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The paper is a proof-of-principle with sound exact benchmarks at small system sizes, but the central quantitative claims (plateau durations, recurrence time, and the trend with system size) currently rest on single fixed truncation and bond-dimension values. I would encourage the editor to request a revision that includes truncation-convergence checks and quantitative reporting for the tensor-network data; without those, the hep-ph readership may overinterpret the results as a lattice-QFT statement rather than a model study."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know up front. First, the paper is a genuine proof-of-principle: coherent initial states in lattice lambda phi^4, evolved with exact diagonalization, an ideal quantum circuit simulator, and tensor networks, and the observables do relax to the diagonal ensemble before recurrences appear. The N=2 circuit matches exact results, and the N=5 tensor network matches exact results. That part holds up. Second, the main physical claim—that an equilibration plateau lasting several times the initial relaxation time is a property of the lattice theory—is not yet supported, because every result sits at one truncation level: n_Q=6 qubits per site for the circuit, fixed N_max in the HO basis for the tensor network, and chi_max=25 for all systems above N=5. There is no convergence check in any of these cutoffs, so the plateau lengths and the recurrence time could be artifacts of the truncated spectrum.\n\nWhat the paper does well: the diagonal ensemble comparison has no free parameters; they check that the long-time average agrees; they cross-benchmark TN against exact at N=5; and the conclusion is honest about the open continuum and volume questions. The heavy-ion and reheating framing is clearly speculative, and they do not oversell it in the conclusion.\n\nThe soft spots are real but proportionate. The stress-test concern about bond dimension mimicking dissipation is plausible, especially for N=7–11; a truncated MPS can remove Schmidt weight and damp oscillations exactly in the direction of a longer plateau. The diagonal ensemble is built from the same Hamiltonian and initial state, so the late-time agreement is an identity of dephasing, not an independent prediction. The 1% criterion is defined on a finite spectrum, so the quoted intervals (t=11.8–18.5, 14.2–63.2) are numbers attached to that cutoff. None of this is hidden; the authors list larger volumes and the continuum limit as open in the conclusion.\n\nMy take: it deserves a serious referee. The proof-of-principle is useful, and the algorithmic work is careful. The referee should ask for at least one truncation scan—for example, n_Q=4,6,8 on N=2, or a brief chi_max dependence for N=7—so the plateau is shown to be stable, not an artifact. As written, I would not cite the plateau lengths as physics, but I would cite the paper as a well-benchmarked numerical demonstration. I'd bring it to a reading group; the truncation issue is a good discussion.","headline":"Solid proof-of-principle on small lattices; the missing truncation-dependence check keeps the plateau claim from being physics yet.","tokens_in":18329,"tokens_out":5393,"would_cite":true,"duration_ms":53143,"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 shows that particle production from highly occupied classical fields in lattice $\\lambda\\phi^4$ theory can be simulated directly on a quantum computer, and that the produced excitations equilibrate to the diagonal ensemble…","keywords":["lattice scalar field theory","coherent states","particle production","equilibration","diagonal ensemble","quantum Poincaré recurrences","quantum simulation","nonequilibrium quantum field theory"],"falsifier":"Re-run the same protocol at fixed $N=5$ while increasing $n_Q$ from 6 to 8 to 10 (and the harmonic-oscillator cutoff accordingly): if the diagonal-ensemble plateau values shift, shorten, or disappear, the claimed equilibration is an artifact of the truncated spectrum rather than a property of the lattice field theory. The companion check is the $\\lambda=0$ control, where the paper itself predicts no relaxation; any observed damping there would signal dephasing from numerical error rather than physical particle production.","tokens_in":17210,"feed_emoji":"⚛️","tokens_out":12800,"duration_ms":113135,"temperature":0.7,"pith_summary":"The paper argues that quantum computers can simulate, from first principles, particle production from highly occupied classical fields—the regime relevant to the earliest moments of heavy-ion collisions and to reheating after inflation. Using lattice $\\lambda\\phi^4$ theory (a scalar field on a spatial grid with a quartic self-interaction) with small lattices and truncated bosonic spaces, it shows that a coherent state initially matching a classical field configuration relaxes: the field expectation value decays, the pressure and occupation numbers rise, and all observables settle at the values predicted by the diagonal ensemble. The equilibrated stage lasts several times the initial relaxation time before late-time oscillations set in, which the paper identifies with quantum Poincaré recurrences. A sympathetic reader would take the claim to be that equilibration is an intrinsic quantum, interaction-driven dephasing phenomenon, not a classical or long-time-average effect, and that this can be studied with polynomial quantum resources.","feed_headline":"Fields equilibrate, then recur in quantum simulation","feed_subtitle":"Coherent phi^4 states relax to diagonal-ensemble values before late-time revivals.","key_machinery":"The central object is the coherent initial state $|\\psi(0)\\rangle=\\bigotimes_p |\\alpha_p\\rangle$, with Glauber coherent-state parameters fixed so that $\\langle\\hat\\phi\\rangle$ and $\\langle\\hat\\pi\\rangle$ equal the classical field and momentum configurations; the zero-momentum mode carries all the initial occupation. The mechanism is quantum dephasing: under unitary time evolution, relative phases among energy eigenstates erase off-diagonal contributions to expectation values, so observables settle to the diagonal ensemble $\\rho_{\\rm eq}=\\sum_n |c_n|^2 |E_n\\rangle\\langle E_n|$ even though the true state stays pure. Numerically, the analysis uses exact Hamiltonian exponentiation in a truncated harmonic-oscillator basis, digital quantum-circuit simulation with $n_Q$ qubits per lattice site and second-order Trotter steps, and, for larger systems, time-dependent variational-principle evolution of matrix-product states with a maximum bond dimension $\\chi_{\\rm max}=25$. The algorithm's resource count, $\\mathrm{CNOT}$ gates $O(N\\,n_Q^4\\,t^{1+1/p}/\\varepsilon^{1/p})$, is what makes the authors' first-principles programme scalable in principle.","core_discovery":"Starting from a spatially homogeneous coherent state with $\\phi_{\\rm cl}(0)=5$ and $\\pi_{\\rm cl}(0)=0$ on a two-site lattice with $a=0.25$, $m=1$, and $\\lambda=1$, the field expectation value relaxes to zero, the pressure rises from $-37.78$ to $19.20$, and the occupation numbers grow from $f_{p_0}=6.5$ to $17.03$ and from $f_{p_1}=0$ to $3.3\\times10^{-3}$; by $t\\approx10\\,m^{-1}$ all match the diagonal ensemble $\\rho_{\\rm eq}=\\sum_n |c_n|^2 |E_n\\rangle\\langle E_n|$, although the underlying many-body state remains pure. Relaxation is sequential in momentum space—the zero mode equilibrates before the higher-momentum mode—and for $\\phi_{\\rm cl}(0)=2,5,10$ the same pattern holds, with residual oscillations shrinking at larger amplitude. The equilibration interval runs from $t=11.8$ to $18.5\\,m^{-1}$ for $\\phi_{\\rm cl}(0)=5$ and from $14.2$ to $63.2\\,m^{-1}$ for $\\phi_{\\rm cl}(0)=10$; for $\\phi_{\\rm cl}(0)=2$ a first recurrence is identified at $t\\simeq411.6\\,m^{-1}$. In larger one-dimensional lattices ($N=5,7,9,11$), the equilibration stage lengthens and recurrence effects move to later times as the Hilbert-space dimension grows.","pith_inferences":["If the plateau persists with increasing truncation and volume, the phenomenon is dephasing in a finite spectrum rather than thermalisation in the thermodynamic sense; the paper's own open questions about the Gibbs ensemble and eigenstate thermalisation point exactly there.","Because equilibration proceeds mode by mode from low to high momentum, a testable extension is to measure the droplet-size dependence of momentum-mode equilibration in small collision systems predicted by kinetic theory.","Starting from multi-mode or squeezed coherent states would test whether the plateau length and recurrence times depend on the number of initially populated modes, which the single-mode homogeneous initial states do not probe.","The pressure plateau suggests a practical observable: if such systems occur in heavy-ion collisions, the early equation of state may be measurable from hadron momentum anisotropies in small systems."],"forward_implications":["Equilibration in this setting is intrinsically quantum and interaction-driven: at $\\lambda=0$ a coherent state stays coherent and follows the classical periodic solution, so the observed relaxation and particle production require the quartic interaction and quantum dephasing.","Observables relax to diagonal-ensemble values within a finite time window without explicit long-time averaging, and the equilibration plateau lasts several times the initial relaxation time.","For more highly occupied initial states the plateau is longer and residuals smaller; for $\\phi_{\\rm cl}(0)=10$ the 1% equilibration interval is about 4.5 times longer than for $\\phi_{\\rm cl}(0)=5$.","Larger lattices display longer-lived equilibration stages and later recurrences, consistent with recurrence times growing with Hilbert-space dimension, and the plateau values converge as $N$ increases.","An effective equation of state can emerge from only a few particles, which is directly relevant to searches for the smallest locally equilibrated quark-gluon droplets at hadron colliders."],"supporting_citations":[{"why":"supplies the quantum algorithms for lattice scalar field theory, including product-formula evolution and the qubit encoding used for real-time dynamics.","marker":"[30–32, 35]"},{"why":"defines the Glauber coherent states whose expectation values match the classical field configuration.","marker":"[44]"},{"why":"provides the Kitaev–Webb algorithm for preparing the free vacuum with polynomial complexity.","marker":"[45]"},{"why":"establishes the diagonal-ensemble dephasing picture of equilibration that gives the benchmark values for all relaxed observables.","marker":"[53–56]"},{"why":"fixes the notion of quantum Poincaré recurrences used to interpret the late-time oscillations.","marker":"[58]"},{"why":"provides the tensor-network implementation used to simulate the larger lattices N=7, 9, and 11.","marker":"[60]"},{"why":"supplies the kinetic-theory observation of momentum-dependent thermalisation that the sequential equilibration is compared with.","marker":"[51, 52]"}],"fun_headline_variants":["Equilibrating fields and Poincaré recurrences in simulation","From classical fields to equilibrium and quantum revivals","Simulation: particles, equilibration, and recurrences","Quantum field relaxation then recurrence on a lattice","Pure-state equilibration and late-time recurrences"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a small, hard-truncated system—up to eleven lattice sites with six qubits per site or a truncated harmonic-oscillator basis—faithfully reproduces the quantum field theory, and the paper does not demonstrate convergence in truncation, continuum, or volume.","fun_headline_variants_meta":{"raw":{"variants":["Equilibrating fields and Poincaré recurrences in simulation","From classical fields to equilibrium and quantum revivals","Simulation: particles, equilibration, and recurrences","Quantum field relaxation then recurrence on a lattice","Pure-state equilibration and late-time recurrences"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000959,"raw_usage":{"total_tokens":4150,"prompt_tokens":1072,"completion_tokens":3078,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":688,"completion_tokens_details":{"reasoning_tokens":3002}},"tokens_in":688,"tokens_out":3078,"duration_ms":23936,"temperature":1.0,"reasoning_tokens":3002,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:12:57.538517+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the same protocol at fixed $N=5$ while increasing $n_Q$ from 6 to 8 to 10 (and the harmonic-oscillator cutoff accordingly): if the diagonal-ensemble plateau values shift, shorten, or disappear, the claimed equilibration is an artifact of the truncated spectrum rather than a property of the lattice field theory. The companion check is the $\\lambda=0$ control, where the paper itself predicts no relaxation; any observed damping there would signal dephasing from numerical error rather than physical particle production.","supporting_citations":[{"cited_title":"Bocchieri and A","cited_arxiv_id":null,"evidence_quote":"fixes the notion of quantum Poincaré recurrences used to interpret the late-time oscillations."}],"review_version":1}