{"id":"41d56602-1f23-4d27-b5a3-6ebd52f53277","arxiv_id":"2504.19285","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For thermodynamically consistent active field theories, the total heat including post-protocol relaxation is minimized at a finite protocol duration, and in a conserved active phase-separation model the optimal duration scales as V^{1/2} Delta-mu^{-1}.","lead":"Active materials, which constantly burn fuel, are cheapest to manipulate when the manipulation takes a finite, not an infinitely slow, time, if you count the heat released after the manipulation ends. The paper derives a general optimization method for active field theories and gives scaling rules for the best control speed in phase-separating active fluids.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quantitative predictions for tau_t and the crossover rest on DRT outside its controlled regime: for conserved CMB, Omega*tau_t ~ 1/(Delta_mu V^{3/2}), so the V^{1/2} scaling and the non-monotonic protocol crossover are not established by the expansion used to derive them, even though the…","rationale":"The reader's weakest assumption is the quantitative accuracy of DRT at the predicted optimal duration, and that is also the most load-bearing concern I find. I agree with the conditional verdict: the central qualitative message, a finite-duration minimum of total heat, is supported by general energetics, exact small-noise analytical integrals, and simulations, and I see no evidence of circularity or data fitting. However, the paper's headline quantitative results, the scalings tau_t ~ V^{1/2}/Delta_mu and the crossover between monotonic and non-monotonic optimal protocols, are derived from DRT in a regime where the theory is not controlled. For conserved CMB the slowest relaxation rate scales as V^{-2}, so the DRT expansion parameter Omega*tau_t decreases as V^{-3/2} as the system grows; this is precisely the direction in which the authors report growing relative errors. The supplemental material also discloses that optimal protocols crossing the spinodal are patched by hand, which further weakens the quantitative crossover prediction. None of this undermines the existence of a finite optimal duration, so I would not move the verdict to reject; the correct status remains conditional pending a check of whether the exact small-noise solution reproduces the DRT scalings at larger volumes and whether the crossover survives without the a posteriori clipping.","tokens_in":38740,"tokens_out":11797,"duration_ms":133799,"concrete_test":"Recompute the optimal duration tau_t for the linear protocol using the exact small-noise integrals of SM III.B (not DRT) at fixed Delta_mu/T = 0.5 and V/ell = 128, 256, 512; extract the effective exponent nu in tau_t ~ V^nu and evaluate Omega*tau_t at each point. If nu departs from 1/2, or if Omega*tau_t << 1 while DRT still matches the exact curves, the DRT-based scaling is uncontrolled and the quantitative claim needs revision. As a complementary check, rerun the SM IV.E optimal-protocol construction with and without the a posteriori clipping of protocols that cross the spinodal, and compare the resulting tau_t from Eq. (23); a material shift would show that the crossover prediction is an artifact of the clipping.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The qualitative claim that total heat Qt = Qp + Qr has a finite-duration minimum for active field theories is credible: it follows from the linear growth of active work at large tau, is realized in CMB by exact small-noise calculations and simulations, and the simulation code is public. The load-bearing weak point is the quantitative layer built on DRT. DRT requires Omega*tau >> 1, with Omega the slowest relaxation frequency. For CMB, the slowest mode has k_min = 2*pi/V, so Omega ~ V^{-2} (SM IV.B), while Eq. (14) and Eq. (23) predict tau_t ~ V^{1/2}/Delta_mu. Hence Omega*tau_t ~ 1/(Delta_mu V^{3/2}) at fixed activity: increasing V pushes the predicted optimum further outside the DRT regime, exactly where Figs. 4(c-d) show growing relative errors. The V^{1/2} and Delta_mu^{-1} scalings are therefore not controlled by the response-theory expansion used to derive them, even though they are compatible with exact small-noise data over the plotted range. A second symptom, flagged in SM IV.E, is that for tau above tau_hat the Euler-Lagrange optimal protocol would cross the spinodal, and the authors alter it a posteriori to stay in the homogeneous phase; the long-tau master curve gamma_inf and the crossover are thus not rigorously derived from the stated optimization. The qualitative minimum and the existence of a crossover may well survive, but the quantitative predictions for tau_t and for the optimal protocol shape are conditional.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a thermodynamically consistent framework for optimal control of active field theories, centered on the distinction between the protocol heat Q_p, dissipated only during manipulation, and the total heat Q_t, which also includes post-protocol relaxation. The central claim is that for active systems Q_t generically has a global minimum at a finite protocol duration, because Q_t is finite at τ = 0 (relaxation after an instantaneous quench) and grows linearly at large τ through the active fuel-consumption term. The authors use dynamical response theory (DRT) around the quasistatic regime to derive explicit expressions for Q_p and Q_t, Eq. (12), and for the optimal durations, Eq. (14). They apply the framework to Chemical Model B (CMB), a conserved scalar active field theory with a φ⁴ free energy and a linear coupling between density and fuel concentration, working in the homogeneous phase and in a small-noise expansion. They report scalings τ_t ~ (V/ℓ)^{1/2}(Δμ/T)^{-1}, a crossover between monotonic and non-monotonic optimal protocols, and support the predictions with small-noise analytical calculations, DRT, and numerical simulations. The simulation code is publicly available.","tokens_in":39068,"tokens_out":5124,"duration_ms":58523,"significance":"If the quantitative claims hold, this is a valuable contribution to stochastic thermodynamics and active matter: it identifies the total heat as the correct control objective, demonstrates a generic finite-time minimum for active systems, and provides parameter-free predictions for the optimal duration and protocol shape in a thermodynamically consistent field theory. The paper has notable strengths: no fitted parameters enter the predictions; the response functions, steady-state power, and boundary terms are computed from the model; the small-noise calculations and simulations are cross-checked; and the code is public. The qualitative argument for a finite-duration minimum of Q_t is robust, as it relies only on the finite value at τ = 0 and the linear large-τ growth of active work. However, the quantitative scalings and the protocol-shape crossover rest on DRT in a regime where its control parameter Ωτ is not asymptotically large, and the long-duration optimal protocol is altered a posteriori to avoid crossing the phase transition. These issues are load-bearing for the advertised predictions, though not for the existence of the finite-τ minimum.","major_comments":[{"comment":"The quantitative scalings of the optimal duration are derived within DRT, whose validity requires Ωτ ≫ 1 where Ω is the slowest relaxation frequency. For conserved CMB, SM IV.B gives Ω ≡ ω₁(a_min, k_min) ∼ V⁻² because k_min = 2π/V, while Eq. (14) predicts τ_t ∼ V^{1/2}/Δμ, so Ωτ_t ∼ 1/(Δμ V^{3/2}) at fixed Δμ. Thus increasing V pushes the predicted optimum outside the controlled regime, and Figs. 4(c-d) indeed show relative errors growing with V and Δμ. The paper itself states that DRT captures the global minimum of Q_t only when its validity regime contains τ_t. Consequently, the advertised V^{1/2} and Δμ⁻¹ scalings and the quantitative location of the crossover are not established by the expansion used to derive them. The existence of a finite-duration minimum is not affected, since it follows from general energetics, but the quantitative control predictions need either an explicit error bound, a higher-order estimate, or a reformulation as qualitative tendencies.","section":"Eq. (14), Eq. (23), SM IV.B, Figs. 4(c-d)"},{"comment":"SM IV.E shows that for τ > τ̂ the assumption of a monotonic optimal protocol is false, and the Euler-Lagrange solution would cross the phase-transition line; the authors then 'forcefully alter a posteriori all optimal protocols to never cross the phase transition.' This means the non-monotonic master curve γ_∞ and the associated crossover are not derived from the stated optimization problem. Because the crossover is a headline claim, the authors should either solve the constrained optimization within the homogeneous phase or provide numerical evidence that the a posteriori restriction does not change the minimal Q_t or its minimizing duration. The main text notes that crossing phase transitions is left for future work, but Eq. (23), which estimates the optimal duration using γ_∞, is precisely built from this constrained object, so the caveat is load-bearing for the quantitative predictions.","section":"SM IV.E and Fig. 5"}],"minor_comments":[{"comment":"The version I received has many symbol substitutions (e.g., 'k 1' for '≫ 1', 'Ä' for τ, '¼φ' for λ_φ), which makes verification unnecessarily difficult; please ensure the final typeset version uses correct mathematical symbols.","section":"Throughout (typeset version)"},{"comment":"Equation (23) is presented as an approximation obtained by matching asymptotics, and the text says it reproduces the numerically estimated value for the specific parameters, but no accuracy estimate is given; a brief discussion of how the accuracy degrades as Ωτ_t approaches the boundary of the DRT regime would strengthen the reproducibility of this central quantity.","section":"Eq. (23)"},{"comment":"The definition of the relative error ε_x = τ_x^DRT/τ_x^anal − 1 appears only in the caption; it would help to define it explicitly in the main text when the scalings are first discussed.","section":"Captions of Figs. 4(c-d)"},{"comment":"The public repository [68] is a strength; consider adding a version identifier or DOI so that the exact simulation parameters and post-processing scripts used for Figs. 3–5 can be cited unambiguously.","section":"Code availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is likely of interest to the readership, and the qualitative finite-duration minimum of Q_t is credible and well supported. The main gap is that the paper advertises quantitative scalings and a protocol-shape crossover as central results, but the controlling expansion is not valid at the predicted optimum and the long-τ protocol is constrained ad hoc. If the authors can reframe the quantitative statements as conditional, add a validity criterion with error estimates, or provide numerical evidence that the constraints do not change the optimum, the paper would be suitable for publication. I see no concerns about citation practices or overlap with prior work; the relevant literature, especially Ref. [4], is properly cited."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Key thing to know: this paper gives the field the right objective for thermodynamic control of active field theories—total heat including post-relaxation, not just protocol heat—and shows that this objective has a finite-duration minimum. That part is physically sound and well supported. The trouble is the quantitative layer: the scalings for the optimal duration and the crossover between protocol shapes are derived with dynamical response theory at points where the theory is not controlled.\n\nWhat is actually new: the distinction between protocol heat Qp and total heat Qt, the general argument that Qt must have a global minimum because active work grows linearly with duration, and the Chemical Model B case study with explicit scalings and a claimed crossover from monotonic to non-monotonic optimal protocols. This goes beyond the earlier Davis-Proesmans-Fodor framework by targeting total heat and by computing the post-relaxation contribution explicitly. The numerics are public, the analytics are checked against simulation in several regimes, and there are no fitted parameters. The demonstration that the spurious drift does not contribute to the active work is a careful piece of work.\n\nThe soft spot is exactly what the stress-test note says. DRT requires Omega*tau >> 1, where Omega is the slowest relaxation rate. For conserved CMB, Omega ~ V^{-2} while the predicted optimum has tau_t ~ V^{1/2}/Delta_mu, so Omega*tau_t ~ 1/(Delta_mu V^{3/2}). Increasing V pushes the predicted optimum further outside the controlled regime, and Figures 4(c-d) show the relative errors growing there. The V^{1/2} scaling and the crossover are not rigorously established by the expansion used to derive them. They may well be correct—the exact small-noise numerics support them over the plotted range—but the derivation is not controlled. A second symptom, disclosed by the authors themselves in SM IV.E, is that for long durations the Euler-Lagrange optimal protocol would cross the spinodal and they patch it a posteriori to remain in the homogeneous phase. The long-tau master curve and the crossover are therefore not fully derived. None of this breaks the central qualitative claim: the finite-duration minimum of total heat is robust and general.\n\nWho this is for: people working on active matter control, stochastic thermodynamics of field theories, and phase-separation kinetics. It deserves a serious referee. The right outcome is probably major revision, with a clearer statement of the DRT validity regime and an honest separation of what is asymptotically matched versus universally derived.\n\nSend it to review.","headline":"The total-heat objective is a real step forward for active-matter control, and the finite-duration minimum is robust; the quantitative scalings, however, are derived with response theory outside its controlled regime and should be treated as conditional.","tokens_in":39610,"tokens_out":1383,"would_cite":true,"duration_ms":16231,"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":"Active field theories dissipate least total heat at a finite control duration, and dynamical response theory predicts the optimum.","keywords":["active matter","stochastic thermodynamics","finite-time control","dynamical response theory","active field theory","heat dissipation","phase separation","optimal protocols"],"falsifier":"Simulate Chemical Model B at finite noise for a series of protocol durations and compare the numerically measured $Q_t(\\tau)$ with the DRT prediction; if at the $\\Delta\\mu/T$ and $V/\\ell$ values where $\\Omega\\tau_t$ is not much larger than one the numerical minimum lies outside the predicted $\\tau_t$, the quantitative claim fails. Alternatively, measure the total heat in an experimental active phase-separating system and test whether the minimum and its scaling with system size and activity match $\\tau_t \\sim (V/\\ell)^{1/2}(\\Delta\\mu/T)^{-1}$.","tokens_in":38482,"feed_emoji":"🔥","tokens_out":11318,"duration_ms":90479,"temperature":0.7,"pith_summary":"Active systems constantly dissipate energy even when unperturbed, so the standard passive rule—optimal control is quasistatic—does not apply. The paper argues that the right quantity to minimize is the total heat $Q_t$, which includes both the heat released during the manipulation and the heat released while the system relaxes afterwards, and that for thermodynamically consistent active field theories $Q_t$ always has a global minimum at a finite protocol duration. It then shows how to find that optimum with dynamical response theory, and predicts for an active phase-separation model (Chemical Model B) that the optimal duration scales as the square root of system size divided by activity, with the optimal protocol switching from monotonic to non-monotonic shape as the duration grows. If correct, this gives experimenters a systematic rule for controlling active materials at minimal dissipation in finite time.","feed_headline":"Active systems burn least heat at a finite control speed","feed_subtitle":"Including post-protocol relaxation, total heat bottoms out at a nonzero duration, not quasistatically.","key_machinery":"The load-bearing object is the dynamical response theory (DRT) expansion of the heat in powers of the protocol rate, which turns the total heat into a one-dimensional effective Lagrangian $L(a,\\dot a) = m(a)\\dot a^2 + \\langle P\\rangle_s(a)$, where $m(a)$ is an effective mass built from integrated response functions and $\\langle P\\rangle_s(a)$ is the steady-state active power. The optimization then reduces to minimizing this Lagrangian over the protocol shape, with the optimal durations given by explicit formulas such as $\\tau_t^2 = (\\int \\dot a^2 m\\,ds)/(P_0+\\int \\langle P\\rangle_s\\,ds)$, so that the crossover between monotonic and non-monotonic optimal protocols is governed by the competition between the kinetic term (dominant at short $\\tau$) and the steady-state power term (dominant at long $\\tau$).","core_discovery":"The paper's central claim is that for active field theories kept away from equilibrium by a chemical fuel, the total heat dissipated in a full control cycle—manipulation plus post-protocol relaxation—is a non-monotonic function of protocol duration, with a global minimum at an intermediate duration that is absent from the protocol heat alone. The existence of the minimum follows from two opposing contributions: at short durations the Lagrangian term $m(a)\\dot a^2$ makes heat diverge as $\\sim 1/\\tau$, while at long durations the background active power $P_0+\\langle P\\rangle_s(a)$ makes it grow linearly with $\\tau$. Applying this to Chemical Model B—a conserved scalar field theory with a linear Onsager coupling between density and fuel—the paper derives the scaling $\\tau_t \\sim (V/\\ell)^{1/2}(\\Delta\\mu/T)^{-1}$ for homogeneous states, and shows that the optimal protocol crosses over from a monotonic master curve to a non-monotonic one that lingers near the phase boundary, because the steady-state dissipation landscape shapes the optimal strategy.","pith_inferences":["This suggests the same trade-off should appear in experimental active systems whose steady-state dissipation grows with the control-parameter dwell time, such as light-activated colloids or ATP-driven emulsions, so the predicted finite-duration optimum could be checked with time-resolved measurements of heat or fuel consumption.","The effective-mass form of the Lagrangian hints at a geometric picture in which finite-time optimal protocols are geodesics in a parameter-space metric determined by response functions; extending the analogy to multi-parameter protocols could yield thermodynamic cycles with characteristic speeds.","The crossover protocol that hovers at the phase boundary suggests a general strategy: when control time is long, park the system in a low-dissipation state rather than interpolating monotonically; this may apply to guided self-assembly or membrane remodeling in active materials.","Extending the computation to protocols that cross phase transitions—flagged as future work—would require the Lagrangian to include interfacial contributions, and the present homogeneous-state predictions provide a baseline against which such extensions can be tested."],"forward_implications":["Total dissipation in active systems is minimized at a finite protocol duration, so optimal control does not require quasistatic protocols; the optimum can be located from response functions alone.","For conserved active fields, the optimal duration grows as the square root of system size and shrinks linearly with activity, $\\tau_t \\sim (V/\\ell)^{1/2}(\\Delta\\mu/T)^{-1}$, so larger or more active systems favor different operating speeds.","The shape of the optimal protocol crosses over from a monotonic master curve at short durations to a non-monotonic one that keeps the control parameter near the phase boundary at long durations, reflecting the landscape of steady-state dissipation.","The protocol heat $Q_p$ alone can be monotonic (no minimum), especially at high activity and large volume, whereas the total heat $Q_t$ always has a global minimum; hence optimizing total heat, not protocol heat, is the generally valid target."],"supporting_citations":[{"why":"Introduces the response-theory approach to controlling active matter that this paper extends from protocol heat to total heat.","marker":"[4]"},{"why":"Supplies the passive-system thermodynamic Lagrangian and effective-mass optimization method that is generalized here to active field theories.","marker":"[54]"},{"why":"Provides the thermodynamically consistent formulation of active field theories with reservoir coupling, fixing the heat and active-work definitions used throughout.","marker":"[38]"},{"why":"Sets the stochastic thermodynamics framework (path probabilities, heat, work) on which the total-heat balance is built.","marker":"[13]"},{"why":"Justifies extending stochastic-thermodynamics heat definitions to active-matter field theories.","marker":"[51]"},{"why":"Supplemental material contains the response-function derivations, small-noise expansions, and scaling analyses that produce the explicit predictions.","marker":"[61]"}],"fun_headline_variants":["Total heat dips at finite control speed","Active systems hit minimal heat mid-protocol","Optimal active control at finite duration","Least heat for active matter at nonzero pace"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the approximate method used to compute heat from slow-response theory remains accurate at the optimal protocol duration; if that approximation breaks down, the predicted optimum could move, though a finite-time minimum still exists.","fun_headline_variants_meta":{"raw":{"variants":["Total heat dips at finite control speed","Active systems hit minimal heat mid-protocol","Optimal active control at finite duration","Least heat for active matter at nonzero pace"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000145,"raw_usage":{"total_tokens":1185,"prompt_tokens":958,"completion_tokens":227,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":574,"completion_tokens_details":{"reasoning_tokens":174}},"tokens_in":574,"tokens_out":227,"duration_ms":2936,"temperature":1.0,"reasoning_tokens":174,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:57:25.877750+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate Chemical Model B at finite noise for a series of protocol durations and compare the numerically measured $Q_t(\\tau)$ with the DRT prediction; if at the $\\Delta\\mu/T$ and $V/\\ell$ values where $\\Omega\\tau_t$ is not much larger than one the numerical minimum lies outside the predicted $\\tau_t$, the quantitative claim fails. Alternatively, measure the total heat in an experimental active phase-separating system and test whether the minimum and its scaling with system size and activity match $\\tau_t \\sim (V/\\ell)^{1/2}(\\Delta\\mu/T)^{-1}$.","supporting_citations":[{"cited_title":"Fodor and M","cited_arxiv_id":null,"evidence_quote":"Provides the thermodynamically consistent formulation of active field theories with reservoir coupling, fixing the heat and active-work definitions used throughout."},{"cited_title":"Sorkin, H","cited_arxiv_id":null,"evidence_quote":"Justifies extending stochastic-thermodynamics heat definitions to active-matter field theories."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplemental material contains the response-function derivations, small-noise expansions, and scaling analyses that produce the explicit predictions."}],"review_version":1}