Pith. sign in

REVIEW 6 minor 18 references

PIPO plants and PITO controllers keep positive feedback control signals bounded.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Persistent-input/persistent-output (PIPO) plants connected to persistent-input/transient-output (PITO) controllers have bounded control signals in positive feedback loops.

T0 review reviewed 2026-08-04 challenge →

load-bearing objection A clean new input-output criterion for bounded control signals in positive interconnections; the core theorem is sound, the examples are concrete, and the paper deserves a serious referee.

arxiv 2608.01699 v1 pith:AIMJEIQO submitted 2026-08-03 eess.SY cs.SY

On input-output persistency and the interconnection of positive nonlinear systems

Arthut Castello B. de Oliveira , Moh K. Wafi , Eduardo D. Sontag This is my paper
classification eess.SY cs.SY MSC 93C1093C2893D25
keywords positive nonlinear systemsinput-output propertiesPIPOPITOboundednessantithetic integral controllermonotone systemsanti-windup
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks when a positive nonlinear feedback loop—plant and controller that keep their states and signals in the nonnegative orthant—can be guaranteed not to produce an unbounded control signal. The authors introduce two complementary input–output properties: a plant is persistent-input/persistent-output (PIPO) if any input that stays above a threshold eventually forces the output above any prescribed level, uniformly over initial states; a controller is persistent-input/transient-output (PITO) if any input that stays above a threshold eventually drives its output below any prescribed tolerance. Their main theorem proves that interconnecting a PIPO plant with a PITO controller, under forward completeness and a one-sided affine growth bound on the controller output, makes the control signal bounded for every initial condition. This matters because boundedness is the first step before studying asymptotic behavior of integral-feedback interconnections used in biomolecular control, and the result needs no linearity, monotonicity, or shared Lyapunov function.

Core claim

Theorem 1 states: for positive nonlinear plant and controller in feedback, if the plant is PIPO, the controller is PITO, the closed loop is forward complete, and the controller output satisfies the one-sided affine bound dw/dt ≤ k w + ω, then the control signal is bounded on [0,∞). The proof runs by contradiction: if the control signal ever reached a sufficiently large value R, the PIPO plant would turn the persistently large input into a persistently large plant output, which as controller input would trigger the PITO property and force the controller output below a small ε within a fixed time; however, the chosen thresholds and the growth bound force that same output to stay above ε during

What carries the argument

The machinery is the pair of qualitative properties PIPO and PITO, plus the one-sided affine growth bound. PIPO is the plant-side 'amplifier' property: inputs never below U(κ) eventually force outputs never below κ, uniformly over initial states. PITO is the controller-side 'attenuator' property: inputs never below V(ε) eventually force outputs to fall below ε, with transient time depending on initial output only. Their roles in the proof are complementary: the plant converts a persistently large control signal into a persistently large measurement, and the controller converts that measurement into a small control signal. The one-sided affine bound on the controller output lets a comparison

Load-bearing premise

The load-bearing premise is the one-sided affine growth bound on the controller output, w_dot ≤ k w + ω; if the controller output can grow faster than linearly in its own value, the comparison step used to limit it during the plant's response delay fails, and the contradiction proof of Theorem 1 no longer holds.

What would settle it

Run a numerical search over positive PIPO plants and PITO controllers satisfying forward completeness and the affine growth bound, looking for an initial condition whose control signal exceeds the proof's predicted bound R; if any trajectory exceeds R, Theorem 1 is false. Alternatively, exhibit a PIPO/PITO pair whose controller output obeys a superlinear growth bound, for example w_dot = w^2, and whose closed-loop control signal escapes to infinity; that would show the affine growth assumption is doing essential work.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Any PIPO plant connected to any PITO controller with forward completeness and the affine growth bound will have a bounded control signal, for every initial condition, with no assumption that either subsystem is stable in isolation.
  • If the plant is also BIBS stable, boundedness of the control signal upgrades to boundedness of the plant state.
  • Every positive monotone plant with a globally asymptotically stable equilibrium for constant inputs and a class-K∞ input–output characteristic automatically satisfies PIPO and BIBS, so the framework covers a broad structural class.
  • For the antithetic integral controller, PITO is satisfied with explicit gains, so the AIC output is bounded in feedback with any positive PIPO plant; boundedness of the second AIC species is left open.
  • The nonlinear-II integral-feedback motif is verified with closed-form PIPO/PITO gains, recovering boundedness under a multiplicative growth bound.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • A testable extension is to certify PIPO/PITO empirically: with a bench experiment, hold the input above candidate thresholds and measure whether output settles above or below prescribed levels; such measurements would supply the gain functions needed to invoke the theorem without a model.
  • The theorem's scope is bounded by the one-sided affine growth condition; controllers whose output can grow superlinearly in its own value (for example, through multiplicative state-dependent rates) will need a different argument or a coordinate change, since the comparison step is the only place the bound enters.
  • Because the result is purely input–output, it may transfer to stochastic or discrete-time positive systems if the PIPO/PITO definitions are adapted to almost-sure or sample-path persistence; the paper does not pursue this direction.
  • The open z2-boundedness issue for the AIC suggests a natural next question: identify minimal dissipativity or detectability conditions on the plant that turn controller-output boundedness into full state boundedness.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 6 minor

Summary. The paper studies feedback interconnections of positive nonlinear SISO systems and introduces two complementary input-output properties: PIPO (persistent input drives the plant output eventually above any prescribed level, uniformly in the initial state) and PITO (persistent input drives the controller output eventually below any prescribed level, with transient time depending on the initial output). The central result, Theorem 1, states that if the closed loop is forward complete, the controller output satisfies a one-sided affine growth bound, the plant is PIPO, and the controller is PITO, then the control signal is bounded. Corollary 1 adds plant-state boundedness under BIBS stability. The paper then gives a structural sufficient condition for PIPO for positive monotone plants with a class-K∞ steady-state characteristic (Lemma 1), proves positivity, forward completeness, and explicit PITO gains for the antithetic integral controller (Lemmas 2–3), and applies the framework to the nonlinear-II integral-feedback motif with explicit PIPO/PITO gains (Lemmas 4–6, Corollary 3). The paper candidly states that the framework does not generally guarantee boundedness of the full controller state, in particular z2 in the antithetic controller.

Significance. If correct, the paper provides a modular, architecture-independent sufficient condition for boundedness of the control signal in positive feedback interconnections, without requiring linearity or monotonicity once PIPO and PITO are verified. The main theorem is self-contained and the worked examples supply explicit gain functions, which is practically useful for synthetic-biology control design. The limitations are clearly acknowledged: the result is a controller-output boundedness result, not a full-state boundedness result, and the nonlinear-II example recovers an existing conclusion rather than proving new asymptotic behavior. The framework itself is novel and appears to be a genuine step beyond the earlier case-specific boundedness analysis in [17]. The paper would be of interest to the systems-biology control community and the broader positive-systems community.

minor comments (6)
  1. [Section 3, Theorem 1] The statement requires 'there exist k, ω > 0' for the affine growth bound, but the proof explicitly handles k=0 and Corollary 2 (AIC) relies on k=0, ω=α1. The condition should be restated as k, ω ≥ 0, not both zero, or the k=0 case should be separated into a remark.
  2. [Section 3, proof of Theorem 1] The shifted applications of PIPO and PITO are justified by an implicit causality argument: the output at a finite time depends only on the input over the preceding interval, so an input can be extended beyond the interval of interest before applying the definition. This is correct, but the proof should spell it out explicitly; the current phrase 'By causality' is terse for such a delicate step.
  3. [Equation (2), Section 4.2] After the controller equation, the text says 'where z=(z1,z2)∈R^2_≥0, y∈R_≥0', but the controller input is v. Replace y by v to avoid confusing the controller input with the plant output.
  4. [Section 4.2, Corollary 2 proof] The phrase 'one-sided PITO property' appears to be a typo; it should read 'PITO property'.
  5. [Section 4.3] The use of u for both the fixed external stimulus and the closed-loop plant input is confusing. Rename the external parameter, for example μ, to avoid collision with the interconnection variable u=w.
  6. [Section 5] The limitation that z2 boundedness is not guaranteed by the framework is clearly stated in Section 4.2. A forward reference from Corollary 2 to this discussion would help readers avoid over-interpreting the corollary.

Circularity Check

0 steps flagged

No significant circularity: Theorem 1 and its supporting lemmas derive their conclusions from explicit PIPO/PITO hypotheses and self-contained proofs, not from their own conclusions.

full rationale

The central derivation is self-contained. Theorem 1 constructs a finite bound R for w(t) from the PIPO gain U(Vε), the PITO time T(ε,δ), and the one-sided growth bound φ; the contradiction argument applies the shifted PIPO and PITO implications with explicit causal extensions, so the conclusion is not an input to the proof. Lemma 1 proves PIPO using monotone comparison with the trajectory x(t,0,Uκ) and convergence to the K∞ characteristic ky(Uκ)>2κ, which is an assumed structural property rather than the desired output lower bound. Lemma 3 verifies PITO for the antithetic controller directly from the equations, with explicit V0 and T formulas, using only positivity and Gronwall's inequality. The nonlinear-II example verifies PIPO and PITO from explicit solution formulas, and Corollary 3 explicitly states that it recovers a conclusion already implied by Corollary 5.2 of [15], so the self-citation there is transparently a worked illustration rather than a hidden premise. The paper also explicitly disclaims boundedness of z2 in Corollary 2, further confirming that it is not overclaiming. Self-citations to [3,4], [15], and [17] are contextual or are used only for standard definitions and forward-completeness of a previously known example; none carries the weight of the main theorem. No fitted parameters are relabeled as predictions, and no uniqueness claim is imported from prior work to force the chosen framework.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The result rests on explicit domain assumptions: forward completeness, a one-sided affine growth bound, and (for the structural PIPO condition) monotonicity with a K∞ steady-state characteristic. No parameters are fitted to data, and no new physical entities are introduced.

axioms (5)
  • domain assumption The closed-loop interconnection is forward complete (each subsystem and the interconnection have globally defined solutions for locally essentially bounded positive inputs).
    Assumed in Theorem 1 and Corollary 2; verified for the nonlinear-II example via [15].
  • domain assumption The controller output satisfies a one-sided affine growth bound ẇ(t) ≤ k w(t) + ω for all t ≥ 0, with k, ω ≥ 0.
    Stated in Section 2.1 as a standing assumption for Theorem 1; holds for AIC with k=0, ω=α1 and for the nonlinear-II controller with k=αy0, ω=0.
  • domain assumption The plant is positive, monotone with respect to state and input, and the output map is order preserving (Assumption 1).
    Used in Lemma 1 to derive PIPO and BIBS; adapted from Angeli-Sontag [3,4].
  • domain assumption For every constant input, the plant has a globally asymptotically stable equilibrium, and the associated input-output characteristic k_y is class K∞ (Assumption 2).
    Used in Lemma 1 to construct the threshold U(κ) via k_y(U_κ) > 2κ; this is the quantitative content of the structural PIPO condition.
  • standard math The nonlinear-II closed loop is globally asymptotically stable for fixed ū > 0 (Lemma 5.1 of [15]).
    Invoked in Corollary 3 to certify forward completeness of the closed loop.

reviewed 2026-08-04 · how reviews work

0 comments
Cite this review

Pith. "Pith review of On input-output persistency and the interconnection of positive nonlinear systems." pith.science (2026). https://pith.science/paper/AIMJEIQO

@misc{pith2026260801699,
  author       = {Pith},
  title        = {Pith review of: On input-output persistency and the interconnection of positive nonlinear systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AIMJEIQO}},
  note         = {Machine review of arXiv:2608.01699}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In this paper we study feedback interconnections of positive nonlinear SISO systems and introduce two complementary properties: persistent-input/persistent-output (PIPO) for the plant and persistent-input/transient-output (PITO) for the controller. Assuming forward completeness of the closed loop and a one-sided affine growth bound on the controller output, we show that PIPO and PITO jointly imply boundedness of the control signal. The result is input-output in nature and does not require linearity or monotonicity of the interconnected subsystems, although it does not in general guarantee boundedness of the full controller state. We provide a structural PIPO condition for positive monotone plants with a class-\(\mathcal K_\infty\) steady-state characteristic, establish PITO for the antithetic integral controller with explicit gains, and illustrate the framework on a nonlinear integral-feedback motif with multiplicative controller growth.

Figures

Figures reproduced from arXiv: 2608.01699 by Arthut Castello B. de Oliveira, Eduardo D. Sontag, Moh K. Wafi.

Figure 1
Figure 1. Figure 1: Closed-loop interconnection between the plant [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The PIPO property – persistently large inputs eventually drive the system’s output to also be proportionally persistently [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The PITO property – persistently large inputs eventually drive the system’s output to be proportionally persistently [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Intuition on PIPO/PITO interconnections. The input [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages · 1 internal anchor

  1. [1]

    D. K. Agrawal, R. Marshall, M. Ali Al-Radhawi, V . Noireaux, and E. D. Sontag. Some remarks on robust gene regulation in a biomolecular integral controller. InProc. 2019 IEEE Conf. Decision and Control, pages 2820–2825, Nice, France, 2019

    work page 2019

  2. [2]

    D. K. Agrawal, R. Marshall, V . Noireaux, and E. D. Sontag. In vitro implementation of robust gene regulation in a synthetic biomolecular integral controller.Nature Communications, 10:5760, 2019

    work page 2019

  3. [3]

    Angeli and E

    D. Angeli and E. D. Sontag. Monotone control systems.IEEE Trans. Automat. Control, 48:1684–1698, 2003. 16

    work page 2003

  4. [4]

    Angeli and E

    D. Angeli and E. D. Sontag. Interconnections of monotone systems with steady-state characteristics. In M. S. de Queiroz, M. A. Malisoff, and P. R. Wolenski, editors,Optimal control, stabilization and nonsmooth analysis, volume 301 ofLecture Notes in Control and Inform. Sci., pages 135–154, Berlin, 2004. Springer

    work page 2004

  5. [5]

    Aoki, Gabriele Lillacci, Ankit Gupta, Armin Baumschlager, David Schweingruber, and Mustafa Khammash

    Stephanie K. Aoki, Gabriele Lillacci, Ankit Gupta, Armin Baumschlager, David Schweingruber, and Mustafa Khammash. A universal biomolecular integral feedback controller for robust perfect adaptation.Nature, 570:533–537, 2019

    work page 2019

  6. [6]

    Internal models in control, bioengineering, and neuroscience.Annual Review of Control, Robotics, and Autonomous Systems, 5:55–79, 2022

    Michelangelo Bin, Jie Huang, Alberto Isidori, Lorenzo Marconi, Matteo Mischiati, and Eduardo Sontag. Internal models in control, bioengineering, and neuroscience.Annual Review of Control, Robotics, and Autonomous Systems, 5:55–79, 2022

    work page 2022

  7. [7]

    Briat, A

    C. Briat, A. Gupta, and M. Khammash. Antithetic integral feedback ensures robust perfect adaptation in noisy biomolecular networks.Cell Syst., 2:15–26, 2016

    work page 2016

  8. [8]

    A biology-inspired approach to the positive integral control of positive systems: The antithetic, exponential, and logistic integral controllers.SIAM J

    Corentin Briat. A biology-inspired approach to the positive integral control of positive systems: The antithetic, exponential, and logistic integral controllers.SIAM J. Applied Dynamical Systems, 19(1):619–664, 2020

    work page 2020

  9. [9]

    A tutorial on modern anti-windup design

    Sergio Galeani, Sophie Tarbouriech, Matthew Turner, and Luca Zaccarian. A tutorial on modern anti-windup design. European Journal of Control, 15(3-4):418–440, 2009

    work page 2009

  10. [10]

    A quasi-integral controller for adaptation of genetic modules to variable ribosome demand.Nature Communications, 9(1):5415, 2018

    H-H Huang, Y Qian, and D Del Vecchio. A quasi-integral controller for adaptation of genetic modules to variable ribosome demand.Nature Communications, 9(1):5415, 2018

    work page 2018

  11. [11]

    Huang, A

    J. Huang, A. Isidori, L. Marconi, M. Mischiati, E. D. Sontag, and W. M. Wonham. Internal models in control, biology and neuroscience. InProc. 2018 IEEE Conf. Decision and Control, pages 5370–5390, 2018

    work page 2018

  12. [12]

    Instability of equilibrium and convergence to periodic orbits in strongly 2-cooperative systems.Journal of Differential Equations, 444:113651, 2025

    Rami Katz, Giulia Giordano, and Michael Margaliot. Instability of equilibrium and convergence to periodic orbits in strongly 2-cooperative systems.Journal of Differential Equations, 444:113651, 2025

    work page 2025

  13. [13]

    Margaliot and E.D

    M. Margaliot and E.D. Sontag. Compact attractors of an antithetic integral feedback system have a simple structure (preprint version).bioRxiv, 2019. DOI: 10.1101/868000

    work page doi:10.1101/868000 2019

  14. [14]

    Michael Margaliot, Chengshuai Wu, and Eduardo D. Sontag. Compact attractors of an antithetic integral feedback system have a simple structure. In2025 IEEE 64th Conference on Decision and Control (CDC), pages 2880–2885, 2025

    work page 2025

  15. [15]

    Symmetry invariance for adapting biological systems.SIAM journal on applied dynamical systems, 10(3):857–886, 2011

    Oren Shoval, Uri Alon, and Eduardo Sontag. Symmetry invariance for adapting biological systems.SIAM journal on applied dynamical systems, 10(3):857–886, 2011

    work page 2011

  16. [16]

    Sontag.Mathematical Control Theory

    E.D. Sontag.Mathematical Control Theory. Deterministic Finite-Dimensional Systems, volume 6 ofTexts in Applied Math- ematics. Springer-Verlag, New York, second edition, 1998

    work page 1998

  17. [17]

    Boundedness of solutions in feedback systems with antithetic controllers

    Moh Kamalul Wafi, Arthur CB de Oliveira, and Eduardo D Sontag. Boundedness of solutions in feedback systems with antithetic controllers.arXiv preprint arXiv:2604.27290, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  18. [18]

    Weiss and M

    E. Weiss and M. Margaliot. A generalization of linear positive systems with applications to nonlinear systems: Invariant sets and the Poincaré-Bendixson property.Automatica, 123:109358, 2021. 17

    work page 2021

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.