{"id":"cfe1b123-a3d9-4da7-a3a3-19cbedcf8305","arxiv_id":"2607.28426","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Shifting the Vlasov–Poisson distribution by a Maxwellian and minimizing over Lyapunov matrices extends Carleman embedding convergence to physically allowed collision frequencies for Landau-damping-type initial data.","lead":"A quantum algorithm for nonlinear plasma kinetics can now be proved to converge at realistic collision rates by shifting the distribution around equilibrium and optimizing a Lyapunov matrix. This matters because earlier Carleman embeddings needed unphysically strong damping, blocking useful kinetic closures for fusion fluid models.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"Sufficiency gap in the R_P<1 criterion, plus incomplete multi-start Lyapunov minimization, is still the load-bearing soft spot for the physical-ν claim.","rationale":"The reader correctly isolates the load-bearing limitation: the paper improves the earlier unphysical-ν result of Vaszary et al. by the Maxwellian shift plus Lyapunov optimization, and for the Landau family the P=I analytic bound already sits inside \nu≲1. Everything else (basis-agnostic growth, averaged-versus-resolved complexity split) is solid supporting material. The only material uncertainty is precisely the one the reader flags—the gap between a sufficient criterion and actual convergence, compounded by the acknowledged incompleteness of the multi-start minimization. No internal contradiction or algebraic error undermines the analytic P=I bound for f_pert, so the verdict remains CONDITIONAL rather than REJECT or ACCEPT; a single classical truncation study would decide whether the gap is empty or not. Minor appendix typos in the prefactor of \nu_I (F3 versus the derivation) do not affect the order-of-magnitude claim.","tokens_in":15387,"tokens_out":657,"duration_ms":80373,"concrete_test":"Fix the Landau initial condition f_pert (α=0.01), N_x=8, N_v=16 and \nu=0.8 (safely above the P=I analytic bound but below 1). Integrate the Carleman-truncated ODE system for increasing N_c=2…10 and measure the ℓ2 truncation error against a high-accuracy classical nonlinear reference at T=10. If the error fails to decay exponentially in N_c, the sufficient-condition guarantee does not materialize at physical \nu.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central guarantee (convergence at ν≲1 for shifted Fourier–Hermite Landau data) rests entirely on the sufficient condition R_P<1 of Jennings et al. (inequality (10)/(14)). For P=I the analytic bound already lies inside the physical window for f_pert up to N_v~10^3 (Eqs. (16)–(18) and App. F), which is the strongest part of the claim. The numerical Riemannian multi-start (N_start=100) is only used to push the bound lower; the authors themselves report that it fails to beat P=I for N_x≥14 on f_pert (Fig. 1c) and leaves two-stream \nu_P slightly above 1 as N_v grows (Fig. 1b). Because the criterion is only sufficient, and because the optimization is local and incomplete, it remains possible that the true radius of convergence is smaller than the reported \nu_P (or that better P exist that would bring two-stream inside the window). No classical truncation-error experiment is supplied to close the sufficiency gap.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript extends Carleman embedding for the nonlinear 1+1 Vlasov–Poisson system by shifting the unknown to f̂ = f − f_M, expanding in a Fourier–Hermite basis, and deriving lower bounds on the collision frequency ν required for convergence of the truncated embedding. Using the sufficient criterion R_P < 1 of Jennings et al., they obtain an analytical P = I bound and numerically minimized Lyapunov bounds ν_P. For perturbed-Maxwellian (Landau) initial data the bounds lie inside the physical window ν ≲ 1 for large mode counts; for two-stream data the minimized bounds sit near or slightly above that window. They further prove a basis-agnostic lower bound showing that required ν must grow with velocity resolution for a broad class of expansions, and they compare quantum query complexity for time-averaged versus time-resolved outputs.","tokens_in":15645,"tokens_out":1328,"duration_ms":35489,"significance":"If the claims hold, the work meaningfully enlarges the regime in which Carleman-based quantum algorithms can be applied to kinetic plasma equations relative to Vaszary et al., who required unphysically large dissipation. The analytical P = I bound for Landau data (Eqs. (16)–(18), App. F) already places N_v up to ~10^3 inside ν ≲ 1, which is a concrete and useful guarantee. Explicit constructions of F_1/F_2, closed-form log- and spectral-norm evaluations, the basis-agnostic growth result (App. G), and a clear complexity distinction between averaged and resolved observables are genuine strengths. The paper is therefore a solid incremental contribution at the interface of quantum algorithms and kinetic plasma physics.","major_comments":[{"comment":"The central physical-ν claim rests on the sufficient (not necessary) criterion R_P < 1 (inequalities (10) and (14), imported from Jennings et al.). For P = I the analytic bound already lies inside ν ≲ 1 for f_pert up to large N_v, which is the strongest part of the result. However, no classical Carleman-truncation experiment is supplied to quantify the sufficiency gap. Without such a check it remains possible that the true radius of convergence is smaller than the reported ν_P (or that better P exist). A short numerical demonstration of truncation error versus N_c for a few (N_x, N_v, ν) points inside the claimed window would substantially strengthen the guarantee.","section":"Convergence of the Carleman Embedding; Eqs. (10), (14)"},{"comment":"Figure 1(b) shows that even after Riemannian multi-start minimization the two-stream lower bound remains slightly above ν = 1 as N_v grows, and Figure 1(c) reports that the optimizer fails to beat P = I for N_x ≥ 14 on f_pert. The abstract and introduction nevertheless state that convergence is established “for physically reasonable collision frequencies.” The claim should be qualified more carefully by initial condition: it is solid for Landau/perturbed-Maxwellian data via the analytic P = I bound, but only marginal or conditional for two-stream data. The text already notes the sufficient nature of the criterion; the abstract and conclusions should mirror that nuance.","section":"Fig. 1; Abstract; Conclusion"},{"comment":"The complexity discussion asserts that extracting a time-resolved final state incurs a factor g = max_t ∥u(t)∥/∥u(T)∥ that is “likely” exponential in T because of the Maxwellian shift and the physical upper bound on ν (paragraph after Eq. (24)). This is plausible but not derived. Either a rigorous lower bound on g under the shifted dynamics, or an explicit statement that the exponential cost is conjectural, is needed before the strong contrast with the time-averaged case (Eq. (23)) can be treated as established.","section":"Complexity; Eqs. (23)–(24)"}],"minor_comments":[{"comment":"Eq. (4) writes the collision term as = ν f̂ on the right-hand side, while the surrounding text and the unshifted Eq. (1) use −ν(f − f_M). The sign convention after the shift should be stated once for clarity (the subsequent ODE (7) correctly has −ν C_{p,r}).","section":"Shifted Vlasov-Poisson System; Eq. (4)"},{"comment":"Appendix E (unshifted minimization) is useful but the main-text reference to it is brief. A one-sentence pointer in the caption of Fig. 1 or in the Convergence section would help readers locate the comparison.","section":"Fig. 1 caption; Appendix E"},{"comment":"Typographical issues: “errof” → “error of” (Introduction); “theerof” → “thereof” (Complexity); “F ourier” spacing artifacts in several appendix headings; “desciptions” → “descriptions” (App. D).","section":"Throughout"},{"comment":"The related work [34] is noted only in a final Note. A short comparison in the Introduction or Conclusion (weak nonlinearity vs. the present shifted approach) would better situate the contribution.","section":"Note; Introduction"},{"comment":"Figure 1 panels would be clearer with explicit legends distinguishing “minimized ν_P”, “P = I”, and “ν = 1” rather than relying solely on colour and the caption.","section":"Fig. 1"}],"recommendation":"minor_revision","confidential_remarks":"The analytic P = I Landau bound is the load-bearing positive result and is already publishable; the numerical Lyapunov optimization and the two-stream claims are weaker and should not be oversold. Fit for a quant-ph / quantum-algorithms venue is good; plasma-physics novelty is incremental but the quantum-complexity discussion is the distinctive part. No integrity or citation concerns."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The real news is that shifting by the Maxwellian turns the collision term into −νf̂ and, with the Jennings Lyapunov criterion, puts Carleman convergence inside the physical window ν≲1 for perturbed-Maxwellian (Landau) data. That removes the main obstacle left by Vaszary et al.\n\nWhat they do well is concrete and checkable. Fourier–Hermite maps cleanly onto F1/F2 (Apps. B–C). The P=I analytic bound is explicit (App. F): ν_I = 1/2 + O(α√N_v), so for α=0.01 you get N_v up to ~10^3 before hitting ν=1—plenty for the resolutions people actually use. Numerical multi-start over P (Pymanopt, 100 starts) improves the two-stream bounds and keeps Landau comfortably below 1 (Fig. 1c–d). The basis-agnostic argument (App. G) that required ν must grow with velocity resolution for a large class of bases is clean and useful. Complexity split is honest: history-state averages look cheap (Õ(N_x√N_v)); time-resolved extraction likely reintroduces an exponential-in-T factor precisely because physical ν forced the shift.\n\nSoft spots are real but proportionate. The criterion is only sufficient; they never close the gap with a classical truncation-error check. Multi-start fails to beat P=I for N_x≥14 on f_pert and leaves two-stream slightly above 1 as N_v grows—so the “physically reasonable for all cases” claim is stronger for Landau than for two-stream. No code ships. None of this breaks the Landau result or the analytic N_v bound.\n\nMath and citations look solid; they build directly on Vaszary and Jennings without circular fitting. This is for people already working quantum algorithms for kinetic plasma or Carleman embeddings. Worth a serious referee. I’d bring it to reading group and would cite the Landau bound and the basis-growth lemma.","headline":"Solid incremental advance: Maxwellian shift plus Lyapunov bounds put Landau-type Carleman VP inside physical ν≲1; two-stream and time-resolved extraction remain soft.","tokens_in":16317,"tokens_out":511,"would_cite":true,"duration_ms":10708,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Shifting the distribution around Maxwellian lets Carleman embedding of Vlasov–Poisson converge at physically allowed collision frequencies.","keywords":["Vlasov-Poisson","Carleman embedding","quantum algorithms","kinetic plasma","Fourier-Hermite","collision frequency","Landau damping","Lyapunov matrix"],"falsifier":"Directly compute the Carleman truncation error for the shifted Fourier–Hermite system at ν = 1 with N_v ≳ 20 and two-stream initial data; if the error fails to decay exponentially with truncation level, the physical-regime claim does not hold.","tokens_in":16211,"feed_emoji":"⚛️","tokens_out":944,"duration_ms":34638,"temperature":0.7,"pith_summary":"The paper shows that a quantum algorithm using Carleman embedding can solve the nonlinear Vlasov–Poisson equations of kinetic plasma physics with collision frequencies that nature actually permits, rather than the unphysically large values required by earlier work. By rewriting the equations for the deviation from a Maxwellian equilibrium and expanding in Fourier–Hermite modes, the authors derive analytical and numerically optimized lower bounds on the needed collision strength; those bounds sit inside the physical window for Landau-damping initial data. They also prove that, for a wide family of basis functions, the collision frequency must still grow with velocity resolution if convergence is to be guaranteed. The resulting quantum algorithm extracts time-averaged transport coefficients far more efficiently than fully time-resolved snapshots. Predictive fluid models of fusion and laser plasmas need reliable kinetic closures; this result indicates that quantum hardware could supply them without artificial dissipation.","feed_headline":"Quantum plasma solver converges at realistic collisions","feed_subtitle":"Shifted Fourier-Hermite Carleman embedding works for Landau damping without unphysical dissipation","key_machinery":"The Carleman convergence ratio R_P rewritten as a lower bound ν_P on collision frequency after the Maxwellian shift; the bound is minimized over Hermitian positive-definite Lyapunov matrices P that certify dissipation of the linear part.","core_discovery":"After shifting the electron distribution by the Maxwellian and expanding in a Fourier–Hermite basis, Carleman embedding of the resulting quadratic ODE system converges for collision frequencies ν ≲ 1 that are physically allowed. The guarantee follows from a sufficient Lyapunov criterion that is evaluated both analytically for the identity matrix and by numerical minimization over positive-definite Hermitian matrices; the minimized bounds lie below the physical ceiling for perturbed-Maxwellian (Landau-damping) data and remain near it for two-stream data. The same analysis shows that any basis expansion forces the required collision frequency to increase with the number of velocity modes retai","pith_inferences":["The same Lyapunov-minimization approach could be tried on Vlasov–Maxwell or kinetic-ion systems to test whether physical collisions remain sufficient once magnetic fields or ion motion are restored.","Tighter necessary-and-sufficient Carleman criteria, if found, might pull the two-stream bounds that currently sit slightly above ν = 1 into the physical window without changing the basis.","The exponential cost of time-resolved readout points toward hybrid workflows that query the quantum device only for averaged moments and advance the fluid equations classically between calls.","Velocity bases whose derivative matrices have slower-growing spectral norms could push the physical-resolution limit well beyond the Hermite scaling."],"forward_implications":["Time-averaged heat flux and other fluid transport coefficients can be extracted with a near-quadratic query advantage over classical FFT time-stepping.","Landau-damping regimes remain inside the physical collision window even with hundreds to thousands of Hermite modes.","Any common velocity basis forces the required collision frequency to grow at least as fast as the square root of the number of modes.","Fully time-resolved final-state readout incurs an exponential-in-time cost factor created by the same Maxwellian shift that enables convergence.","Fluid plasma codes can in principle replace free-parameter closures with self-consistent kinetic moments obtained from the quantum algorithm."],"fun_headline_variants":["Carleman quantum plasma solver converges at physical collisions","Shifted Fourier-Hermite extends Carleman Vlasov-Poisson convergence","Quantum algorithm reaches realistic ν for Landau-damping plasmas","Carleman embedding works for Vlasov-Poisson without excess dissipation","Basis analysis ties required collisions to velocity mode count"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The argument rests on a sufficient mathematical test for truncation error that may still demand more collisions than nature supplies once velocity resolution is high or the plasma starts far from equilibrium.","fun_headline_variants_meta":{"raw":{"variants":["Carleman quantum plasma solver converges at physical collisions","Shifted Fourier-Hermite extends Carleman Vlasov-Poisson convergence","Quantum algorithm reaches realistic ν for Landau-damping plasmas","Carleman embedding works for Vlasov-Poisson without excess dissipation","Basis analysis ties required collisions to velocity mode count"]},"model":"grok-4.5","effort":"low","cost_usd":0.003576,"raw_usage":{"total_tokens":1091,"prompt_tokens":688,"num_sources_used":0,"completion_tokens":68,"cost_in_usd_ticks":35764000,"prompt_tokens_details":{"text_tokens":688,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":335,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":688,"tokens_out":68,"duration_ms":6117,"temperature":1.0,"reasoning_tokens":335,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T07:44:11.310189+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Directly compute the Carleman truncation error for the shifted Fourier–Hermite system at ν = 1 with N_v ≳ 20 and two-stream initial data; if the error fails to decay exponentially with truncation level, the physical-regime claim does not hold.","supporting_citations":[],"review_version":1}