{"id":"0665f048-e9c2-4981-bdc6-9b7a149f9bad","arxiv_id":"1908.09957","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A gas of hard spheres expanding in Bjorken flow has a constant Knudsen number, which yields an exact analytical hydrodynamic attractor and the first convergent gradient expansion in an expanding relativistic system.","lead":"This paper derives the exact analytical solution for the shear stress in an ultrarelativistic gas of hard spheres under Bjorken expansion, and uses it to find the late-time hydrodynamic attractor. It proves that the gradient expansion converges in this setting, the first such example in a rapidly expanding system, and validates the result against a full Boltzmann simulation.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The full-Boltzmann 'attractor' is extracted only from equilibrium initial conditions; no run varies the initial momentum distribution, so initial-condition independence is unestablished.","rationale":"The reader's weakest_assumption concerned the constant-cross-section modeling assumption and the reliability of the numerical solver. Those are legitimate scope and error-quantification issues, but the more load-bearing internal gap is the absence of any test of initial-condition independence for the Boltzmann attractor. 'Attractor' means late-time behavior independent of initial conditions within a basin. The paper proves this analytically for the Israel-Stewart truncation via the decaying c0 term in Eq. (10), but for the full nonlinear Boltzmann equation it demonstrates only that a single equilibrium-initialized trajectory per Kn reaches a constant plateau. That is necessary but not sufficient. The scaling check at τ0 = 1 fm keeps the same equilibrium initial temperature and chemical potential, so it tests only the scaling variable Kn, not universality over initial momentum distributions. This matters because Kn is constant: the system does not approach local equilibrium as τ → ∞; it approaches a non-thermal scaling solution, and no proof is given that this solution is unique. At large Kn the collision rate per expansion time is small, so the finite-time plateaus observed in numerics could in principle retain memory of the initial anisotropy. A numerical experiment varying the initial momentum-space anisotropy is therefore the decisive check. The analytical convergence result and the claimed 20% agreement remain meaningful conditional on that check. Since the reader's verdict is already CONDITIONAL, and this concern specifies an additional condition rather than overturning the paper, I recommend no change to the verdict.","tokens_in":8829,"tokens_out":28504,"duration_ms":297361,"concrete_test":"Run the Boltzmann solver of Section 4 for the same Kn with at least two qualitatively different initial conditions: (i) the equilibrium Jüttner distribution used in Fig. 1, and (ii) an anisotropic distribution (e.g., Romatschke-Strickland with a large anisotropy parameter) normalized to the same n(τ0) and ε(τ0), hence the same Kn through n0τ0σT, but with a substantially different initial shear π(τ0). Evolve both to the same late plateau time used in Fig. 1 and compare the asymptotic χ for a large-Kn case (e.g., σT = 0.06 fm²) and a small-Kn case. Establish the solver's numerical error by increasing the test-particle number and reducing the time step. If the two asymptotic values agree within that error, the attractor claim is supported; if they differ, the Boltzmann attractor is initial-condition dependent and the central claim must be restricted to equilibrium-initialized solutions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The analytical Israel-Stewart part is solid: Eq. (12) is a genuine attractor of Eq. (9) because the c0-dependent term in Eq. (10) decays, and the convergence radius of Eq. (14) is correctly identified. The load-bearing gap is in Section 4, where the full-Boltzmann 'exact attractor' is claimed. Every numerical simulation in Fig. 1 starts from the same equilibrium distribution (τ0 = 0.1 fm, T = 0.5 GeV, vanishing chemical potential), and the only variation is σT; the scaling check with τ0 = 1 fm and ten-times-smaller σT preserves the same equilibrium initial temperature and chemical potential, so it tests only the functional dependence on Kn, not universality over initial distributions. Reaching a constant χ at late times for one initial condition demonstrates a fixed point of that trajectory, not an attractor. The text asserts that 'any dimensionless quantity constructed using moments of the Boltzmann distribution must asymptote to a constant' and that this defines the attractor, but no uniqueness argument or basin-of-attraction test is given. Since Kn is constant, the collision rate is only a fixed fraction of the expansion rate, and there is no automatic guarantee that memory of the initial momentum-space anisotropy is erased at large Kn. If the asymptotic χ depends on initial conditions, the headline claim to have 'exactly determined the hydrodynamic attractor' of the Boltzmann equation fails.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies an ultrarelativistic gas of hard spheres undergoing Bjorken flow, with particle number conservation and a constant total cross section. It shows that the Knudsen number is constant, which makes the Israel-Stewart viscous hydrodynamic equations analytically solvable. The resulting late-time attractor for the shear-stress ratio chi (Eq. 12) is shown to have a gradient expansion that converges absolutely in a finite range of Knudsen numbers (Eq. 14, with radius |Kn| < 3/(2 a sqrt(b)) for lambda = 0). The paper also extracts a late-time constant value of chi from numerical solutions of the full nonlinear Boltzmann equation for several cross sections, identifies this as the exact Boltzmann attractor, and compares it with the Israel-Stewart attractor, finding agreement within 20% even at large Knudsen numbers.","tokens_in":9101,"tokens_out":5324,"duration_ms":55811,"significance":"If the claims hold, this is a significant contribution: it provides the first example of a convergent gradient expansion in a relativistically expanding system and an exact analytical attractor for an Israel-Stewart theory, which is a valuable benchmark for far-from-equilibrium hydrodynamics. The analytical derivation is transparent and the convergence radius is correctly identified from the branch points of the solution. The comparison with the full Boltzmann equation, if the attractor claim is justified, would be an important test of hydrodynamic attractor universality. However, as discussed below, the Boltzmann-side claim currently lacks a basin-of-attraction test, so the significance is somewhat contingent.","major_comments":[{"comment":"The claim that the late-time constant chi extracted from the Boltzmann equation is the 'exact hydrodynamic attractor' is not supported by the numerical evidence. All simulations shown in Fig. 1 start from the same equilibrium distribution (tau0 = 0.1 fm, T = 0.5 GeV, vanishing chemical potential); the only variation is the total cross section, and the scaling check with tau0 = 1 fm and a smaller cross section preserves the same equilibrium initial conditions. These runs show that the late-time value depends only on Kn for this particular family of initial conditions, but they do not establish that the limit is independent of the initial momentum-space distribution. The argument in the text that 'any dimensionless quantity constructed using moments of the Boltzmann distribution must asymptote to a constant' only shows approach to a fixed point along a trajectory, not uniqueness across initial conditions. Because Kn is constant, the collision rate is a fixed fraction of the expansion rate, so there is no automatic mechanism that erases the memory of the initial momentum-space anisotropy at large Kn. To support the headline claim, the authors should either perform additional numerical experiments with different initial momentum-space distributions (e.g., anisotropic distributions, different chemical potentials) and demonstrate the same late-time chi, or explicitly soften the claim to a late-time fixed point for equilibrium initial conditions.","section":"Section 4, Fig. 1 and Fig. 2"},{"comment":"The abstract states that 'in this example the gradient expansion converges' without qualification, but the convergence proof in Section 3 applies only to the Israel-Stewart truncation of the gradient expansion, specifically to the solution of Eq. (9). Section 4 correctly notes that the gradient expansion for the full nonlinear Boltzmann equation remains an open question and even demonstrates divergence in the relaxation-time-approximation toy model. The abstract should be reworded to attribute the convergent series to the Israel-Stewart theory, otherwise it overstates the scope of the result and could mislead readers into thinking the convergence holds for the full Boltzmann equation.","section":"Abstract and Section 1"}],"minor_comments":[{"comment":"The numerical Boltzmann solver is described only by reference to Ref. [53]; no resolution checks (grid size, particle number, time-step convergence) or numerical error estimates are provided. Since the 'exact Boltzmann attractor' rests entirely on these numerics, a brief convergence test or a statement of numerical uncertainty would strengthen the presentation.","section":"Section 4, numerical method"},{"comment":"The spelling 'Mandelstan' should be 'Mandelstam'.","section":"Eq. (18) and surrounding text"},{"comment":"The caption says the dashed curves indicate the asymptotic values that determine the attractor; it would be clearer to state explicitly that these runs all use equilibrium initial conditions, since that is the basis of the attractor extraction.","section":"Fig. 1 caption"},{"comment":"The phrase 'deviations ... remain 20% at best' is ambiguous; 'never exceed 20%' would be clearer.","section":"Conclusions"}],"recommendation":"major_revision","confidential_remarks":"The analytical Israel-Stewart part is clean and likely correct; the main risk is the overstatement of the Boltzmann-side attractor claim. I would advise requesting the additional initial-condition scans before publication, as the current manuscript's central novelty ('exact hydrodynamic attractor of the full Boltzmann equation') depends on that evidence. The paper should also make the scope of the convergent gradient expansion unambiguous in the abstract."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: the exact Israel-Stewart solution and the converging gradient series are real and worth taking seriously. The Boltzmann-level attractor claim is weaker than the abstract suggests, because every simulation starts from the same equilibrium distribution; the paper demonstrates a late-time fixed point for one initial condition, not initial-condition independence.\n\nWhat is genuinely new: in a hard-sphere gas with particle-number conservation, the Knudsen number is constant, and that simple observation unlocks an exact analytical solution of the Israel-Stewart equation. The resulting attractor, Eq. (12), is a clean function of Kn, and the gradient series converges absolutely for |Kn| < 3/(2a√b) (approximately 0.5 with the 14-moment coefficients). That is indeed the first convergent gradient expansion in a rapidly expanding system, and the proof is straightforward: the series coefficients come from expanding a square root, and the radius follows from branch points. The authors also correctly note that this does not contradict the holographic and RTA divergence results; it just shows expansion alone does not force divergence. Credit where due: the transport-coefficient input, the handling of the removable singularity at Kn = 0, and the honest comparison with Israel-Stewart theory are all well done.\n\nThe soft spots, in order of importance. (1) The convergence proof is for the Israel-Stewart truncation, not for the full Boltzmann equation. The abstract's phrase \"in this example the gradient expansion converges\" is technically true only for that truncation; the full Boltzmann gradient series is left open, and the paper even says so in the conclusions. That's a significant scope limitation, and the abstract blurs it. (2) The full-Boltzmann \"attractor\" in Section 4 is extracted from runs that all start from the same equilibrium distribution at τ0 = 0.1 fm, T = 0.5 GeV, with only σ_T varied. The late-time constant χ is then plotted against Kn, but the universality that defines an attractor—independence from the initial momentum-space anisotropy—is never tested. The scaling check with τ0 = 1 fm and ten-times-smaller cross section changes Kn without changing the initial distribution shape; it confirms the functional dependence on Kn, not insensitivity to initial conditions. So as written, the Boltzmann curve is a late-time fixed point for one family of initial conditions, not a proven attractor. This is the main flaw. (3) Minor: no numerical error bars, no convergence check on the Boltzmann solver itself, and the claim that any dimensionless moment must asymptote to a constant is asserted without proof. These are addressable.\n\nWho should read this: anyone working on hydrodynamic attractors, gradient expansions, or kinetic-theory derivations of hydrodynamics. The analytical result is a useful counterexample to the folklore that expanding systems always have divergent gradient series. The Boltzmann attractor claim needs more work, but it is a testable, well-posed question.\n\nRecommendation: send it to peer review. The analytical core is solid, and the overclaim is fixable—either restrict the abstract to the Israel-Stewart convergence or run the Boltzmann solver with non-equilibrium initial conditions to establish universality. I would not desk-reject this.","headline":"The Israel-Stewart convergence result is solid and novel; the full-Boltzmann attractor claim is supported only for a single equilibrium initial condition, so the abstract overstates it.","tokens_in":9609,"tokens_out":2084,"would_cite":true,"duration_ms":24904,"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":"For a hard-sphere gas under Bjorken flow, the hydrodynamic attractor is exact and its gradient expansion converges.","keywords":["hydrodynamic attractor","Bjorken flow","Israel-Stewart theory","Boltzmann equation","hard-sphere gas","Knudsen number","gradient expansion","heavy-ion collisions"],"falsifier":"Run a high-precision numerical solution of the full nonlinear Boltzmann equation for the same hard-sphere gas at a fixed Knudsen number but with initial times spread over a wide range, adjusting the initial density and cross section to keep $Kn = 1/(n_0\\tau_0\\sigma_T)$ fixed, and read off the late-time value of $\\pi/(\\epsilon+P)$; if that value depends on $\\tau_0$, the claimed attractor is not a function of $Kn$ alone.","tokens_in":8620,"feed_emoji":"⚛️","tokens_out":10221,"duration_ms":96063,"temperature":0.7,"pith_summary":"This paper shows that an ultrarelativistic gas of hard spheres expanding in the boost-invariant Bjorken flow has an exact hydrodynamic attractor, and that its gradient expansion converges over a finite range of Knudsen numbers. Because particle number is conserved in binary collisions, the density falls as $1/\\tau$ and the mean free path grows linearly with $\\tau$, so $Kn = 1/(n_0 \\tau_0 \\sigma_T)$ is constant. That constancy reduces the Israel-Stewart equations to an autonomous ordinary differential equation that can be solved in closed form, giving an attractor $\\chi_{\\rm att}$; the same attractor, computed from the full nonlinear Boltzmann equation, agrees within 20% even at large Knudsen numbers. If correct, this overturns the expectation that gradient expansions always diverge in rapidly expanding systems and supports the practical use of hydrodynamics far from equilibrium.","feed_headline":"Exact hydrodynamic attractor found for expanding hard-sphere gas","feed_subtitle":"Constant Knudsen number makes hydro exactly solvable; the full Boltzmann attractor mirrors it within 20%.","key_machinery":"The load-bearing object is the Knudsen number of the hard-sphere gas. Binary collisions conserve particle number, so $n(\\tau)=n_0\\tau_0/\\tau$ exactly, and with a constant total cross section the mean free path $l_{\\rm mfp}=1/(n\\sigma_T)$ grows linearly with $\\tau$; because the macroscopic gradient scale in Bjorken flow is also $\\tau$, $Kn$ is a constant fixed by initial conditions. This converts the Israel-Stewart equation for the shear correction into an autonomous ODE in $\\chi$ alone, solvable in closed form, and the late-time fixed point is the exact attractor whose power series in $Kn$ is the convergent gradient expansion. For the microscopic comparison, the same attractor is obtained by solving the full nonlinear Boltzmann equation numerically in Milne coordinates until all dimensionless moments saturate, since a constant Knudsen number forces the attractor to be constant.","core_discovery":"The paper's central discovery is that particle-number conservation in a hard-sphere gas under Bjorken flow makes the Knudsen number constant, $Kn = l_{\\rm mfp}/\\tau = 1/(n_0 \\tau_0 \\sigma_T)$, rather than time dependent as in previously studied conformal systems. With the shear viscosity and relaxation time written as $\\eta = a T/\\sigma_T$, $\\tau_\\pi = b\\eta/(4P)$, and $\\tau_{\\pi\\pi}=3\\lambda\\tau_\\pi$, the Israel-Stewart equation for $\\chi=\\pi/(4P)$ becomes an autonomous first-order ODE whose general solution is analytic, and whose late-time fixed point, $\\chi_{\\rm att}=A - \\frac{3}{8}\\frac{4+ab\\lambda Kn}{ab Kn}$ with $A$ the square root in Eq. (11), is the attractor. The gradient expansion of $\\chi_{\\rm att}$ in powers of $Kn$ converges absolutely for $|Kn| < 3/(2a\\sqrt{b})$ (with the 14-moment values $a=4/3$, $b=5$ giving radius $\\approx 0.5$), the first proven convergence in an expanding system. The paper then extracts the same attractor from a numerical solution of the full nonlinear Boltzmann equation and shows that the Israel-Stewart and Boltzmann attractors agree to within 20% even when the Knudsen number is large.","pith_inferences":["An immediate extension the authors do not make: for an energy-dependent cross section the Knudsen number varies with time, so the exact ODE reduction is lost; this paper therefore predicts that conformal-like gases should retain the previously seen divergent gradient behavior under Bjorken flow.","Because the closed-form attractor depends only on $Kn$, it can serve as a benchmark: any approximate hydrodynamic closure aiming at far-from-equilibrium validity should reproduce Eq. (12) for this system rather than only the Navier-Stokes limit.","One could probe the analytic structure of $\\chi_{\\rm att}$ as a function of complex $Kn$; the finite convergence radius suggests singularities at complex values, and locating the analogous singularities in the Boltzmann attractor might diagnose the onset of non-hydrodynamic transient modes."],"forward_implications":["In this hard-sphere gas the gradient expansion converges absolutely for $|Kn| < 3/(2a\\sqrt{b})$, so a convergent hydrodynamic series is possible even in a rapidly expanding system.","The exact attractor of the full nonlinear Boltzmann equation is a function of the Knudsen number alone and is independent of the initial preparation, extending attractor results beyond the relaxation-time approximation.","Israel-Stewart theory reproduces the exact Boltzmann attractor to within 20% even at large Knudsen numbers, supporting the use of viscous hydrodynamics far from equilibrium in heavy-ion collision modeling.","The zeroth-order slow-roll approximation gives the exact attractor in this system, so the closed form replaces the need for higher-order gradient corrections."],"supporting_citations":[{"why":"Documents the divergent gradient expansions in earlier expanding systems that this paper contrasts with its convergent series.","marker":"[4]"},{"why":"Introduces the hydrodynamic attractor concept and the slow-roll scheme whose zeroth-order truncation is shown here to be exact.","marker":"[5]"},{"why":"Defines the Bjorken flow geometry that the calculation uses as its symmetry setting.","marker":"[29]"},{"why":"Provides the Boltzmann-derived Israel-Stewart equations and the 14-moment transport coefficients used for the analytical solution.","marker":"[42]"},{"why":"Supplies the numerical method used to solve the full nonlinear Boltzmann equation and extract its attractor.","marker":"[53]"},{"why":"Gives reference numerical Bjorken-flow solutions used to check the Boltzmann solver.","marker":"[54]"}],"fun_headline_variants":["Exact attractor for expanding hard-sphere gas","Gradient expansion converges for hard-sphere gas","Constant Knudsen yields exact hydro attractor","Hard-sphere attractor matches Boltzmann within 20%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction rests on the modeling assumption that the collision cross section is independent of energy, so the mean free path is set by the particle density and the Knudsen number stays constant; if the cross section depends on energy, as in conformal or QCD-like matter, the Knudsen number becomes time dependent and the exact solution and convergence proof no longer apply.","fun_headline_variants_meta":{"raw":{"variants":["Exact attractor for expanding hard-sphere gas","Gradient expansion converges for hard-sphere gas","Constant Knudsen yields exact hydro attractor","Hard-sphere attractor matches Boltzmann within 20%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000193,"raw_usage":{"total_tokens":1347,"prompt_tokens":937,"completion_tokens":410,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":553,"completion_tokens_details":{"reasoning_tokens":348}},"tokens_in":553,"tokens_out":410,"duration_ms":4445,"temperature":1.0,"reasoning_tokens":348,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:58:22.274740+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a high-precision numerical solution of the full nonlinear Boltzmann equation for the same hard-sphere gas at a fixed Knudsen number but with initial times spread over a wide range, adjusting the initial density and cross section to keep $Kn = 1/(n_0\\tau_0\\sigma_T)$ fixed, and read off the late-time value of $\\pi/(\\epsilon+P)$; if that value depends on $\\tau_0$, the claimed attractor is not a function of $Kn$ alone.","supporting_citations":[{"cited_title":"In this section we obtain the attractor of an ultrarelativistic gas of hard spheres now from a microscopic perspective using the Boltzmann equation [50]","cited_arxiv_id":null,"evidence_quote":"Documents the divergent gradient expansions in earlier expanding systems that this paper contrasts with its convergent series."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the hydrodynamic attractor concept and the slow-roll scheme whose zeroth-order truncation is shown here to be exact."},{"cited_title":"Santos, J","cited_arxiv_id":null,"evidence_quote":"Gives reference numerical Bjorken-flow solutions used to check the Boltzmann solver."}],"review_version":1}