{"id":"d625f845-06dc-4291-a9cc-ccbee7e518a0","arxiv_id":"2506.02645","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A run-and-tumble particle in a piecewise linear potential with thermal noise reaches a double-exponential steady state with two relaxation times and a moving relaxation front.","lead":"Run-and-tumble particles in a V-shaped trap, with added thermal noise, settle into a steady distribution that is a sum of two exponential curves with different decay lengths. The paper derives this exactly and approximates how each part relaxes, including a relaxation front that sweeps outward at constant speed.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (24) is not merely an unverified interpolation: in the Brownian limit it contradicts the exact root λ(s)=(b+√(b²+4sT))/(2T), so the central relaxation-time claim τ_i=(λ_i²T)^{-1} can be off by a factor 4.","rationale":"The reader identified Eq. (24) as the weakest assumption, and this stress-test confirms that identification while making it concrete: the interpolation is not merely unsupported outside three parameter sets, it fails in the Brownian limit that is included in the model. The exact Brownian root quoted by the paper has a branch point at s=-b²/(4T), whereas Eq. (24) implies s=-b²/T; the resulting relaxation time differs by a factor of 4. This affects the time-dependent claims (τ_i, front speed, MSD) but not the exact steady-state result (11), the entropy-production formula (38)-(39), or the qualitative distinction between two modes. The paper already labels the time-dependent solution as approximate, but the abstract presents τ_i as an identified result without the needed caveat. The correct response is not rejection: the steady-state contribution is solid, and the approximate time-dependent formulas may remain useful in the tested regimes. The reader's CONDITIONAL verdict is therefore appropriate, and no reclassification is needed; the condition should explicitly require either validating Eq. (24) across parameter extremes or restricting the relaxation-time claims to regimes where it has been tested.","tokens_in":18059,"tokens_out":14612,"duration_ms":151656,"concrete_test":"Take the Brownian limit of the model (b=T=1, r→10^{-4}, v0=√(2rTac) with Tac=1), solve the quartic Eq. (18) numerically for λ_i(s), and compare λ_i(s)²−λ_i(0)² to s/T over s∈[10^{-4},10]. If the ratio deviates by more than about 20% near s=b²/(4T)=0.25, Eq. (24) and the derived τ_i, front speed, and MSD formula are not reliable in an included parameter regime; additionally, a Langevin simulation with these parameters should show a local relaxation time at x=0 approaching 4T/b²=4 rather than τ=T/b²=1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the interpolation ansatz (24), because the relaxation time τ_i=(λ_i²T)^{-1}, the front speed v_i^*=2Tλ_i, and the MSD formula (34) all pass through it. The problem is sharper than 'unverified': Eq. (24) is already wrong in a limit that is part of the model. For r→0 with v0→0 the active noise vanishes and the exact Laplace-space root for the half-line is λ_Br(s)=(b+√(b²+4sT))/(2T), as the paper itself states before Eq. (22). This root has its branch point at s=-b²/(4T), hence a relaxation time 4T/b². Eq. (24) with λ_i=b/T gives √(s/T+b²/T²), branch point s=-b²/T and τ=T/b², a factor of 4 too fast. The same factor appears in the white-active-noise limit v0,r→∞ with Tac fixed. Equation (24) matches only the s→0 and s→∞ asymptotics, which do not control the intermediate-time exponential e^{-t/τ_i}. The numerical checks in Fig. 3 all use Tac comparable to T and therefore cannot detect this failure. Consequently the abstract's unqualified identification of τ_i=(λ_i²T)^{-1} overstates what is established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a one-dimensional run-and-tumble particle in a piecewise linear potential U(x)=b|x|, with both telegraphic active noise and thermal white noise. It obtains an exact steady-state position distribution as a sum of two Laplace distributions, whose inverse length scales are the two negative eigenvalues of a 3x3 matrix (Sec. 2). It then analyzes relaxation from a delta-function initial condition. The Laplace-space solution reduces to a quartic eigenvalue problem (Sec. 3); the authors approximate the two relevant eigenvalues by Eq. (24) and the mode amplitudes by Eq. (25), and use these to derive an approximate time-dependent distribution, MSD, local relaxation times, and a propagating relaxation-front velocity (Sec. 4). The paper also derives the steady-state entropy production rate in terms of the mode weights (Sec. 5).","tokens_in":18410,"tokens_out":5559,"duration_ms":50834,"significance":"The steady-state derivation is an elegant, exact result that is verified against Langevin simulations and has no free parameters; this part is a genuine contribution. The time-dependent analysis, if valid, gives a simple two-mode relaxation picture with testable predictions (front speed, MSD, local relaxation time). However, the time-dependent conclusions rest on two uncontrolled approximations, and one of them (Eq. (24)) demonstrably fails in a limit that is part of the model, so the unqualified relaxation-time claim in the abstract is not supported. The paper is worth revising, not rejecting: the exact steady-state core and the simulation methodology are sound.","major_comments":[{"comment":"The interpolation ansatz Eq. (24) is not merely unverified away from the three parameter sets: it is provably incorrect in the thermal (weak-activity) limit that is part of the model. For r→0 and v0→0, the exact Laplace-space root of the quartic (18) tends to the Brownian half-line root λ_Br(s)=(b+√(b²+4sT))/(2T), which the authors themselves write before Eq. (22). That root has a branch point at s=-b²/(4T), giving a relaxation time 4T/b². Substituting λ_i=b/T into Eq. (24) gives √(s/T+b²/T²), with branch point at s=-b²/T and hence τ_i=T/b², a factor of 4 smaller. Because τ_i=(λ_i²T)^{-1} (Eq. (28)), the front speed v_i^*=2√(T/τ_i), and the MSD (34) all inherit this error, the paper's unqualified statement in the abstract that the mode relaxation time is (λ_i²T)^{-1} is not established in the weakly active regime. The numerical checks in Figs. 3 and 4 use Tac comparable to T and thus cannot detect the failure; the authors should either restrict the claim to the regime where the approximation is controlled or provide a better approximation with a rigorous error estimate.","section":"Sec. 4, Eq. (24)"},{"comment":"The approximation B_i(s)≈A_i/s+α_i/T with α_i=1/4 is introduced without derivation; the text immediately acknowledges (Figs. 3(c-d)) that it deviates at intermediate s and that the resulting P_app is not normalized at intermediate times. Since the mode amplitudes control the relative weights of the two Laplace components and the short-time Gaussian splitting, the approximate time-dependent distribution (26) and the MSD (34) inherit an uncontrolled error at the very times where the two modes exchange probability. The paper should quantify the error, or better, determine α_i from the exact expression (20) at a matching point (e.g., by matching the small-s expansion), rather than setting it by hand.","section":"Sec. 4, Eq. (25)"},{"comment":"The relaxation-front velocity v_i^*=2√(T/τ_i) is derived by saddle-point evaluation of the approximate integrals in Eq. (36), so it is a property of the interpolation ansatz, not a directly computed consequence of the exact quartic solution (18). The local relaxation time t(x) in Fig. 5 is defined by an arbitrary threshold ε=0.1 and tested for one parameter set; the non-monotonic dip is also present in the Brownian case of Fig. 5(b), so this observation does not by itself discriminate between the model and ordinary Brownian motion. I would like to see at least one additional parameter set for t(x) and a statement of how t(x) depends on ε.","section":"Sec. 4.3"}],"minor_comments":[{"comment":"The last term inside the second bracket appears to be t/τ1 e^{-t/τ2}; it should presumably be t/τ2 e^{-t/τ2}.","section":"Sec. 4.1, Eq. (34)"},{"comment":"There are several typos: 'Lapalce' (Sec. 3), 'normalization' (Sec. 2), 'Botzmann-like' (Sec. 6), 'confiding potential' (Sec. 5), and '10 6' should be '10^6' in figure captions.","section":"Throughout"},{"comment":"The statement that all eigenvalues are real 'as absence of any boundaries forbids oscillatory solutions' is not self-evident; for a non-Hermitian matrix M3, realness of eigenvalues should be justified (or it can be checked from the discriminant of the cubic).","section":"Sec. 2, after Eq. (8)"}],"recommendation":"major_revision","confidential_remarks":"The exact steady-state portion (Sec. 2) is a solid, self-contained contribution that I would be happy to see published. The time-dependent claims rest on two unquantified approximations; the Brownian-limit failure of Eq. (24) is concrete and load-bearing. If the authors can restrict the relaxation-time claims to the regime where the ansatz is controlled, or supply a corrected approximation that captures the Brownian branch point, the paper would be acceptable for publication. The current abstract overstates what is established."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper has one result worth keeping and one set of claims that is currently over-sold. The steady-state distribution of an RTP in a V-shaped trap with thermal noise is exactly a sum of two Laplace distributions, with decay rates given by the negative eigenvalues of a 3x3 matrix. That derivation is exact, the simulation match is clean, and I don't see this in the earlier literature. If you work on active particles in confinement, this is a useful, citable result.\n\nThe time-dependent part is a different story. The authors propose the interpolation λ_i(s) ≈ sqrt(s/T + λ_i²) for the Laplace-space eigenvalues, then build the relaxation times τ_i=(λ_i²T)^{-1}, the front speed, and the MSD on top of it. The stress-test note is right: in the limit r→0, v0→0 (pure Brownian motion in the same potential), the exact root is λ(s)=(b+√(b²+4sT))/(2T), which has its branch point at s=-b²/(4T), giving τ=4T/b². The interpolation with λ_i=b/T gives branch point at s=-b²/T, τ=T/b²—off by a factor of 4. The same factor appears in the white-active-noise limit. So Eq. (24) is not just unverified in those corners; it is wrong where the model reduces to a known solvable case. The numerical checks in Fig. 3 all use Tac comparable to T, so they cannot see this failure. The abstract's unqualified 'τ_i=(λ_i²T)^{-1}' is therefore an overstatement.\n\nThe other approximation, B_i(s)≈A_i/s + 1/(4T), is honestly labeled as asymptotic-only, and the paper admits intermediate-s deviations and non-normalization. That part is less troubling because the authors are transparent about it. The problem is the eigenvalue interpolation, which is presented as if it captures the whole s range.\n\nBottom line: the exact SSD and the entropy-production expression in terms of mode weights are solid. The relaxation dynamics claims need either exact roots of the quartic (which are available numerically, and the paper already computes them for Fig. 3) or a clearly stated validity regime. Send it to peer review, but it comes back with a required major revision on Section 4 and the abstract.","headline":"The exact steady-state double-exponential result is solid and new; the relaxation-time claims rest on an approximation that fails in the Brownian limit, so the time-dependent part needs a major caveat or rework.","tokens_in":18866,"tokens_out":4325,"would_cite":true,"duration_ms":41318,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["05.40.-a","05.10.Gg","05.70.Ln"],"model":"deepseek-v4-flash","headline":"For a run-and-tumble particle in a heat bath trapped by a V-shaped potential, the exact steady state is a sum of two exponential modes, and each mode relaxes on its own time scale.","keywords":["run-and-tumble particle","active matter","thermal noise","steady-state distribution","relaxation dynamics","piecewise linear potential","entropy production","Laplace distribution"],"falsifier":"Compute the two negative roots of the quartic equation (18) numerically for parameter sets outside the three tested ones, for example very small or very large temperature $T$, and compare $\\lambda_i(s)$ with $\\sqrt{s/T+\\lambda_i^2}$ over the full range of $s$; any substantial deviation at intermediate $s$ would invalidate $\\tau_i=(\\lambda_i^2 T)^{-1}$ and $v_i^*=2T\\lambda_i$. A complementary experimental check is to measure the local relaxation time $t(x)$ in Langevin simulations for such parameters and test whether the large-$|x|$ slope equals $1/v_i^*$.","tokens_in":17861,"feed_emoji":"🦠","tokens_out":13773,"duration_ms":121918,"temperature":0.7,"pith_summary":"This paper asks how a run-and-tumble particle—a swimmer that runs straight and randomly tumbles—relaxes when it is also in contact with a heat bath and confined by a piecewise linear potential $U(x)=b|x|$. The authors establish an exact steady-state result: the position distribution is not a single Boltzmann or Laplace form but a normalized sum of two exponential distributions, $P(x)=\\frac{A_1}{\\lambda_1}e^{-\\lambda_1|x|}+\\frac{A_2}{\\lambda_2}e^{-\\lambda_2|x|}$, with the inverse lengths $\\lambda_i$ fixed by the active speed, tumbling rate, potential slope, and temperature. The approach to this steady state is organized around the same two modes: each mode first spreads as a thermal Gaussian, then relaxes on a time $\\tau_i=(\\lambda_i^2 T)^{-1}$, with the distant tail reached by relaxation fronts moving at speed $2T\\lambda_i$. The steady-state entropy production rate is also expressible through the mode weights and the $\\lambda_i$, so a full non-equilibrium steady state and its relaxation are captured analytically for this active system.","feed_headline":"Two exponential modes shape a run-and-tumble particle's steady state","feed_subtitle":"A trapped active swimmer with thermal noise relaxes through two modes, each with its own time scale.","key_machinery":"The central object is the two-mode decomposition of the position distribution, built from the coupled Fokker-Planck equations for the right- and left-moving subpopulations. In sum-and-difference variables, the steady-state equations reduce to a linear first-order system with a $3\\times 3$ matrix $\\mathbf{M}_3$; the two negative eigenvalues of $\\mathbf{M}_3$ supply the inverse lengths $\\lambda_1,\\lambda_2$ of the two Laplace modes in the steady state. The time-dependent problem uses the same construction, producing a $4\\times 4$ matrix $\\mathbf{M}_4$ and the quartic eigenvalue equation whose two negative roots are the Laplace-space inverse lengths $\\lambda_i(s)$. The argument is carried by the interpolation ansatz $\\lambda_i(s)\\approx\\sqrt{s/T+\\lambda_i^2}$ together with the coefficient ansatz $B_i(s)=A_i/s+1/(4T)$, which match the Brownian limits at small and large times and allow the Laplace transform to be inverted explicitly into Gaussian decays plus integrals that saturate to the steady state.","core_discovery":"The central claim is that the one-dimensional position distribution of a run-and-tumble particle with telegraphic active noise and white thermal noise, in the V-shaped potential $U(x)=b|x|$, splits into two dynamically distinct modes. At steady state this is exact: $P(x)=\\frac{A_1}{\\lambda_1}e^{-\\lambda_1|x|}+\\frac{A_2}{\\lambda_2}e^{-\\lambda_2|x|}$, where $\\lambda_1,\\lambda_2$ are the absolute values of the two negative eigenvalues of the matrix $\\mathbf{M}_3$ that governs the steady Fokker-Planck system, and $A_1,A_2$ are fixed by symmetry at the origin and by normalization. In Laplace space the time-dependent solution has the same two-mode structure, with $s$-dependent inverse lengths $\\lambda_i(s)$ obtained from a quartic equation. Because an exact inverse Laplace transform is impractical, the paper proposes the interpolations $\\lambda_i(s)\\approx\\sqrt{s/T+\\lambda_i^2}$ and $B_i(s)=A_i/s+1/(4T)$, which are exact in the short- and long-time limits and reproduce the numerically computed eigenvalues for the three parameter sets tested. From these follow the mode relaxation times $\\tau_i=(\\lambda_i^2 T)^{-1}$, the Gaussian decaying part of the transient distribution, the relaxation fronts moving at $v_i^*=2T\\lambda_i$, a closed-form mean-square displacement, and the entropy-production rate expressed through the mode weights and inverse lengths.","pith_inferences":["An inference beyond the paper: if the interpolation ansatz $\\lambda_i(s)\\approx\\sqrt{s/T+\\lambda_i^2}$ holds beyond the three tested parameter sets, the same two-mode structure should appear for any piecewise-linear confining potential, and for smooth potentials one would expect a discrete or continuous spectrum of relaxation times with $\\tau\\sim(\\lambda^2 T)^{-1}$; this is a testable extension th","An inference beyond the paper: the non-monotonic local relaxation time, with its dip at intermediate distances, arises from the crossing of an overshooting core and an undershooting tail; this signature should be observable in experiments that release a confined active colloid nearly from a point and track the density toward steady state.","An inference beyond the paper: the entropy-production formula gives a stochastic-thermodynamic reading in which the two modes act like two reservoirs at effective temperatures $k_B T_i=b/\\lambda_i$; this suggests that steady-state dissipation could be inferred from density measurements alone, a connection the paper states but does not develop into a measurement protocol."],"forward_implications":["The steady state of a thermally coupled run-and-tumble particle in a linear trap is exactly a two-exponential mixture, so no single effective temperature reproduces the density; a second, shorter correlation length always contributes near the origin.","Since $\\tau_i=(\\lambda_i^2 T)^{-1}$, the mode with the larger $\\lambda_i$ relaxes faster; the minority short-range mode therefore decays quickly while the dominant tail mode controls the late-time approach to steady state.","Relaxation is center-outward: close to the origin the local relaxation time is constant, while at large distances it grows linearly with $|x|$ as a front propagates at speed $v_i^*=2T\\lambda_i$; the same pattern, including a non-monotonic dip, is found for a purely Brownian particle in the same potential.","The mean-square displacement interpolates between the short-time thermal law $\\langle x^2\\rangle\\approx 2T t$ and a steady-state value through a closed approximate formula that matches Langevin simulations for the parameter sets tested.","The steady-state entropy production rate is fixed by the two-mode splitting, $\\dot S = \\frac{v_0^2}{\\mu k_B T}+\\mu b^2\\left(\\frac{\\phi_1}{k_B T_1}+\\frac{\\phi_2}{k_B T_2}-\\frac{1}{k_B T}\\right)$, so the density shape alone determines the dissipation."],"supporting_citations":[{"why":"Introduces the equivalence between a run-and-tumble particle with thermal noise and a Brownian particle in a dichotomously fluctuating potential, and supplies the steady-state solution method for the V-shaped potential.","marker":"[56]"},{"why":"Provides the exact steady state and relaxation results for a run-and-tumble particle without thermal contact in confining potentials, the baseline this paper extends.","marker":"[39]"},{"why":"Supplies the coupled Fokker-Planck equations for the two running states from which the analysis starts.","marker":"[43]"},{"why":"Gives the trigonometric solution of the cubic equation used to obtain the steady-state eigenvalues lambda_i.","marker":"[60]"},{"why":"Provides the Laplace-transform pairs used to invert the approximate time-dependent distribution into Gaussian decays and steady-state integrals.","marker":"[62]"},{"why":"Supplies the exact Brownian relaxation solution in the same potential, used to compare the measured local relaxation time t(x) and identify the non-monotonic dip.","marker":"[63]"},{"why":"Supplies the steady-state entropy-production-rate formula that the paper evaluates with its two-mode steady-state distribution.","marker":"[64]"},{"why":"Gives the free-space long-time diffusion coefficient of a run-and-tumble particle, used to check the flat-potential limit and the distinction between configurational and kinetic temperature.","marker":"[65]"}],"fun_headline_variants":["Active particle in heat bath relaxes via two exponential modes","Run-and-tumble dynamics show dual-mode steady state and relaxation","Two modes set relaxation times for trapped run-and-tumble particle","Thermal bath and active noise produce two-mode particle relaxation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The time-dependent predictions rest on the approximation $\\lambda_i(s)\\approx\\sqrt{s/T+\\lambda_i^2}$ and on a hand-set initial amplitude of $1/4$ for each mode, which the paper checks numerically for only three parameter sets; if these fail elsewhere, the relaxation times, front speeds, and MSD formula are not guaranteed.","fun_headline_variants_meta":{"raw":{"variants":["Active particle in heat bath relaxes via two exponential modes","Run-and-tumble dynamics show dual-mode steady state and relaxation","Two modes set relaxation times for trapped run-and-tumble particle","Thermal bath and active noise produce two-mode particle relaxation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000295,"raw_usage":{"total_tokens":1839,"prompt_tokens":1194,"completion_tokens":645,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":810,"completion_tokens_details":{"reasoning_tokens":588}},"tokens_in":810,"tokens_out":645,"duration_ms":6651,"temperature":1.0,"reasoning_tokens":588,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:18:41.628337+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the two negative roots of the quartic equation (18) numerically for parameter sets outside the three tested ones, for example very small or very large temperature $T$, and compare $\\lambda_i(s)$ with $\\sqrt{s/T+\\lambda_i^2}$ over the full range of $s$; any substantial deviation at intermediate $s$ would invalidate $\\tau_i=(\\lambda_i^2 T)^{-1}$ and $v_i^*=2T\\lambda_i$. A complementary experimental check is to measure the local relaxation time $t(x)$ in Langevin simulations for such parameters and test whether the large-$|x|$ slope equals $1/v_i^*$.","supporting_citations":[{"cited_title":"Bier and R","cited_arxiv_id":null,"evidence_quote":"Introduces the equivalence between a run-and-tumble particle with thermal noise and a Brownian particle in a dichotomously fluctuating potential, and supplies the steady-state solution method for the V-shaped potential."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the exact steady state and relaxation results for a run-and-tumble particle without thermal contact in confining potentials, the baseline this paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the coupled Fokker-Planck equations for the two running states from which the analysis starts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the trigonometric solution of the cubic equation used to obtain the steady-state eigenvalues lambda_i."},{"cited_title":"Oberhettinger and L","cited_arxiv_id":null,"evidence_quote":"Provides the Laplace-transform pairs used to invert the approximate time-dependent distribution into Gaussian decays and steady-state integrals."},{"cited_title":"Chase, K","cited_arxiv_id":null,"evidence_quote":"Supplies the exact Brownian relaxation solution in the same potential, used to compare the measured local relaxation time t(x) and identify the non-monotonic dip."},{"cited_title":"Paoluzzi, A","cited_arxiv_id":null,"evidence_quote":"Supplies the steady-state entropy-production-rate formula that the paper evaluates with its two-mode steady-state distribution."},{"cited_title":"Malakar, V","cited_arxiv_id":null,"evidence_quote":"Gives the free-space long-time diffusion coefficient of a run-and-tumble particle, used to check the flat-potential limit and the distinction between configurational and kinetic temperature."}],"review_version":1}