pith. sign in

arxiv: 2510.13354 · v2 · submitted 2025-10-15 · 🧮 math.OC

Target Controllability Scores for Actuation-Constrained Network Intervention

Kazuhiro Sato This is my paper

Pith reviewed 2026-05-18 07:50 UTC · model grok-4.3

classification 🧮 math.OC
keywords target controllability scoreoutput controllability Gramianvirtual systemactuator constraintsnetwork controlvolumetric scoreenergy controllabilitybrain networks
0
0 comments X

The pith

Target controllability scores assess node importance for specific targets under actuator constraints using convex optimization on virtual system Gramians.

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

The paper introduces target controllability scores to rank which nodes to control when only certain parts of a network matter and actuators face limits. These scores split into volumetric and average energy versions, each found by solving a convex optimization problem built around the output controllability Gramian of a specially constructed virtual system. Because the scores focus only on the targets, they can produce different importance rankings than conventional full-state controllability measures. The authors also develop reduced versions of the virtual system that approximate the scores accurately when connections between targets and other nodes are weak and the system matrix has suitable properties, and they test this on human brain networks to show how well the approximations hold over different time scales.

Core claim

The central claim is that the target volumetric controllability score and the target average energy controllability score are well-defined as the unique optimal solutions to convex programs involving the output controllability Gramian in a virtual system, and that these scores can differ qualitatively from their full-state analogs precisely because restricting attention to target nodes modifies the Gramian. The work further shows that a target-only reduced virtual system yields non-asymptotic error bounds for these scores when cross terms are small and the logarithmic norm of the dynamics matrix is low or negative, with particular accuracy at short or moderate time horizons.

What carries the argument

The output controllability Gramian of the virtual system, which encodes the reachable energy to the targets and serves as the basis for convex optimization problems that define the target VCS and AECS.

Load-bearing premise

The virtual system formulation and the associated output controllability Gramian correctly model the impact of actuator constraints and the restriction to designated target nodes.

What would settle it

A numerical counterexample in which the target scores fail to differ from full-state scores despite a nontrivial projection onto the targets, or an experiment where the reduced-system approximation errors violate the stated non-asymptotic bounds.

Figures

Figures reproduced from arXiv: 2510.13354 by Kazuhiro Sato.

Figure 1
Figure 1. Figure 1: Comparison between the conventional and practical settings. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Illustration of the block decomposition of [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Boxplots of the top 5 nodes: target AECS (top) and its reduced-system [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Boxplots of the top 5 nodes: target VCS (top) and its reduced-system [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Boxplots of the top 5 nodes: target VCS (top) and its reduced-system [PITH_FULL_IMAGE:figures/full_fig_p015_6.png] view at source ↗
read the original abstract

We introduce the target controllability score (TCS), a concept for evaluating node importance under actuator constraints and designated target objectives, formulated within a virtual system setting. The TCS consists of the target volumetric controllability score (VCS) and the target average energy controllability score (AECS), each defined as an optimal solution to a convex optimization problem associated with the output controllability Gramian. We establish existence and uniqueness (for almost all time horizons), develop a projected gradient method for computation, and show that target VCS/AECS can behave qualitatively differently from their standard full-state counterparts because projection onto the target nodes changes the underlying Gramian structure. To enable scalability, we construct a target-only reduced virtual system and derive non-asymptotic bounds showing that weak cross-coupling and a low or negative logarithmic norm of the system matrix yield accurate approximations of target VCS/AECS, particularly over short or moderate time horizons. Experiments on human brain networks reveal a clear trade-off: at short horizons, both target VCS and target AECS are well approximated by their reduced formulations, while at long horizons, target AECS remains robust but target VCS deteriorates.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript introduces target controllability scores (TCS) for actuation-constrained network intervention, consisting of the target volumetric controllability score (VCS) and target average energy controllability score (AECS). These are defined as optimal solutions to convex optimization problems involving the output controllability Gramian within a virtual system setting that incorporates designated target nodes and actuator constraints. The authors prove existence and uniqueness for almost all time horizons, develop a projected gradient method for computation, derive non-asymptotic error bounds for a target-only reduced virtual system under weak cross-coupling and logarithmic-norm conditions, and demonstrate via experiments on human brain networks that the target scores can differ qualitatively from full-state counterparts due to the Gramian projection while exhibiting horizon-dependent approximation quality.

Significance. If the central claims hold, the paper provides a scalable, theoretically supported framework for assessing node importance in targeted control problems with actuator constraints. Strengths include the explicit non-asymptotic bounds that enable reduced-order approximations, the projected gradient solver, and the clear attribution of qualitative differences to changes in the underlying Gramian structure induced by target projection. These elements, together with the brain-network experiments, offer practical value for applications in complex networked systems such as neuroscience.

major comments (1)
  1. [Target-only Reduced Virtual System] Target-only Reduced Virtual System section: the non-asymptotic error bounds for approximating target VCS/AECS rely on stated conditions of weak cross-coupling and low/negative logarithmic norm of the system matrix; while these yield accurate short-to-moderate horizon approximations, the manuscript does not provide a sharpness analysis or counterexamples when the conditions are violated, which bears on the generality of the scalability claims.
minor comments (2)
  1. [Existence and Uniqueness] The statement of existence and uniqueness 'for almost all time horizons' is standard but would benefit from a brief remark on the exceptional set (e.g., its measure-zero character) to aid readers unfamiliar with generic controllability results.
  2. [Experiments] In the brain-network experiments, reporting the specific network dimensions, number of target nodes, and actuator constraint sets used would improve reproducibility and allow readers to assess how the observed trade-offs scale with problem size.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their careful review and supportive recommendation for minor revision. We address the major comment below.

read point-by-point responses

  1. Referee: [Target-only Reduced Virtual System] Target-only Reduced Virtual System section: the non-asymptotic error bounds for approximating target VCS/AECS rely on stated conditions of weak cross-coupling and low/negative logarithmic norm of the system matrix; while these yield accurate short-to-moderate horizon approximations, the manuscript does not provide a sharpness analysis or counterexamples when the conditions are violated, which bears on the generality of the scalability claims.

    Authors: We appreciate the referee's observation on the conditional nature of the error bounds. These non-asymptotic bounds are derived explicitly under the assumptions of weak cross-coupling and low or negative logarithmic norm, which permit precise quantification of the approximation error between the full target VCS/AECS and their target-only reduced counterparts, particularly for short-to-moderate horizons. The conditions are clearly stated and are motivated by structural properties that hold in the brain-network examples where the reduced approximations are shown to perform well. While a dedicated sharpness analysis or counterexamples outside these regimes would provide additional insight into bound tightness, the manuscript's contribution centers on establishing sufficient conditions together with explicit, computable error estimates that support practical scalability. To clarify the scope of the scalability claims, we will add a concise discussion in the revised manuscript noting the dependence on the stated conditions and the potential for degraded approximation quality when they are violated. revision: partial

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper defines target VCS and AECS directly as optimal solutions to convex optimization problems using the output controllability Gramian in a virtual system. It then proves existence/uniqueness for almost all horizons via standard arguments, supplies a projected gradient solver, and derives explicit non-asymptotic bounds for the target-only reduction from stated conditions on cross-coupling and logarithmic norm. These steps constitute an independent derivation chain grounded in control theory definitions and explicit mathematical reductions; the qualitative difference from full-state scores is attributed to the direct effect of target projection on the Gramian structure, with no reduction to fitted parameters, self-definitional loops, or load-bearing self-citations.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies insufficient detail to enumerate free parameters or invented entities; standard linear-system assumptions are implicit but not listed explicitly.

pith-pipeline@v0.9.0 · 5722 in / 1090 out tokens · 39930 ms · 2026-05-18T07:50:58.732145+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Relationship Between Controllability Scoring and Optimal Experimental Design

    math.OC 2026-02 unverdicted novelty 7.0

    Finite-time controllability scoring matches approximate OED, with VCS corresponding to D-optimality and AECS to A-optimality, plus a unique optimizer and long-horizon node downweighting.

Reference graph

Works this paper leans on

51 extracted references · 51 canonical work pages · cited by 1 Pith paper

  1. [1]

    Social diversity reduces the complexity and cost of fostering fairness,

    T. Cimpeanu, A. Di Stefano, C. Perret, and T. A. Han, “Social diversity reduces the complexity and cost of fostering fairness,”Chaos, Solitons & Fractals, vol. 167, p. 113051, 2023

    work page 2023

  2. [2]

    Resilience-based component im- portance measures for critical infrastructure network systems,

    Y .-P. Fang, N. Pedroni, and E. Zio, “Resilience-based component im- portance measures for critical infrastructure network systems,”IEEE Transactions on Reliability, vol. 65, no. 2, pp. 502–512, 2016

    work page 2016

  3. [3]

    Multi-criteria node criticality assessment framework for critical infrastructure networks,

    L. Faramondi, G. Oliva, and R. Setola, “Multi-criteria node criticality assessment framework for critical infrastructure networks,”International Journal of Critical Infrastructure Protection, vol. 28, p. 100338, 2020

    work page 2020

  4. [4]

    Influence of the first- mover advantage on the gender disparities in physics citations,

    H. Kong, S. Martin-Gutierrez, and F. Karimi, “Influence of the first- mover advantage on the gender disparities in physics citations,”Com- munications Physics, vol. 5, no. 1, p. 243, 2022

    work page 2022

  5. [5]

    Sporns,Networks of the Brain

    O. Sporns,Networks of the Brain. MIT press, 2010

    work page 2010

  6. [6]

    Network hubs in the human brain,

    M. P. Van den Heuvel and O. Sporns, “Network hubs in the human brain,”Trends in cognitive sciences, vol. 17, no. 12, pp. 683–696, 2013

    work page 2013

  7. [7]

    Centrality measures in networks,

    F. Bloch, M. O. Jackson, and P. Tebaldi, “Centrality measures in networks,”Social Choice and Welfare, pp. 1–41, 2023

    work page 2023

  8. [8]

    Network centrality: an introduction,

    F. A. Rodrigues, “Network centrality: an introduction,”A mathematical modeling approach from nonlinear dynamics to complex systems, pp. 177–196, 2019

    work page 2019

  9. [9]

    Controllability of complex networks,

    Y .-Y . Liu, J.-J. Slotine, and A.-L. Barab´asi, “Controllability of complex networks,”Nature, vol. 473, no. 7346, pp. 167–173, 2011

    work page 2011

  10. [10]

    Vital nodes identification in complex networks,

    L. L ¨u, D. Chen, X.-L. Ren, Q.-M. Zhang, Y .-C. Zhang, and T. Zhou, “Vital nodes identification in complex networks,”Physics reports, vol. 650, pp. 1–63, 2016

    work page 2016

  11. [11]

    Network alignment,

    R. Tang, Z. Yong, S. Jiang, X. Chen, Y . Liu, Y .-C. Zhang, G.-Q. Sun, and W. Wang, “Network alignment,”Physics Reports, vol. 1107, pp. 1–45, 2025

    work page 2025

  12. [12]

    On the general theory of control systems,

    R. E. Kalman, “On the general theory of control systems,” inPro- ceedings First International Conference on Automatic Control, Moscow, USSR, 1960, pp. 481–492

    work page 1960

  13. [13]

    Mathematical description of linear dynamical systems,

    R. E. Kalman, “Mathematical description of linear dynamical systems,” Journal of the Society for Industrial and Applied Mathematics, Series A: Control, vol. 1, no. 2, pp. 152–192, 1963

    work page 1963

  14. [14]

    Structural controllability,

    C.-T. Lin, “Structural controllability,”IEEE Transactions on Automatic Control, vol. 19, no. 3, pp. 201–208, 1974

    work page 1974

  15. [15]

    Dilation choice sets, dulmage– mendelsohn decomposition, and structural controllability,

    C. Commault and J. van der Woude, “Dilation choice sets, dulmage– mendelsohn decomposition, and structural controllability,”IEEE Trans- actions on Control of Network Systems, vol. 11, no. 2, pp. 1046–1055, 2024

    work page 2024

  16. [16]

    Minimal controllability problems,

    A. Olshevsky, “Minimal controllability problems,”IEEE Transactions on Control of Network Systems, vol. 1, no. 3, pp. 249–258, 2014

    work page 2014

  17. [17]

    A framework for structural input/output and control configuration selection in large-scale systems,

    S. Pequito, S. Kar, and A. P. Aguiar, “A framework for structural input/output and control configuration selection in large-scale systems,” IEEE Transactions on Automatic Control, vol. 61, no. 2, pp. 303–318, 2015

    work page 2015

  18. [18]

    An overview of structural systems theory,

    G. Ramos, A. P. Aguiar, and S. Pequito, “An overview of structural systems theory,”Automatica, vol. 140, p. 110229, 2022

    work page 2022

  19. [19]

    Minimal controllability problems on linear structural descriptor systems,

    S. Terasaki and K. Sato, “Minimal controllability problems on linear structural descriptor systems,”IEEE Transactions on Automatic Control, vol. 11, no. 5, pp. 2522–2528, 2022

    work page 2022

  20. [20]

    Minimal controllability problem on linear structural descriptor systems with forbidden nodes,

    S. Terasaki and K. Sato, “Minimal controllability problem on linear structural descriptor systems with forbidden nodes,”IEEE Transactions on Automatic Control, vol. 69, no. 1, pp. 527–534, 2024

    work page 2024

  21. [21]

    Controllability of large-scale networks: The control energy exponents,

    G. Baggio and S. Zampieri, “Controllability of large-scale networks: The control energy exponents,”IEEE Transactions on Control of Network Systems, vol. 11, no. 2, pp. 808–820, 2023

    work page 2023

  22. [22]

    Controllability metrics, limitations and algorithms for complex networks,

    F. Pasqualetti, S. Zampieri, and F. Bullo, “Controllability metrics, limitations and algorithms for complex networks,”IEEE Transactions on Control of Network Systems, vol. 1, no. 1, pp. 40–52, 2014

    work page 2014

  23. [23]

    On submodularity and controllability in complex dynamical networks,

    T. H. Summers, F. L. Cortesi, and J. Lygeros, “On submodularity and controllability in complex dynamical networks,”IEEE Transactions on Control of Network Systems, vol. 3, no. 1, pp. 91–101, 2016

    work page 2016

  24. [24]

    Energy-aware controlla- bility of complex networks,

    G. Baggio, F. Pasqualetti, and S. Zampieri, “Energy-aware controlla- bility of complex networks,”Annual Review of Control, Robotics, and Autonomous Systems, vol. 5, no. 1, pp. 465–489, 2022

    work page 2022

  25. [25]

    On the role of network centrality in the controllability of complex networks,

    N. Bof, G. Baggio, and S. Zampieri, “On the role of network centrality in the controllability of complex networks,”IEEE Transactions on Control of Network Systems, vol. 4, no. 3, pp. 643–653, 2017

    work page 2017

  26. [26]

    Controllability maximization of large-scale sys- tems using projected gradient method,

    K. Sato and A. Takeda, “Controllability maximization of large-scale sys- tems using projected gradient method,”IEEE Control Systems Letters, vol. 4, no. 4, pp. 821–826, 2020

    work page 2020

  27. [27]

    Centrality measures and the role of non- normality for network control energy reduction,

    G. Lindmark and C. Altafini, “Centrality measures and the role of non- normality for network control energy reduction,”IEEE Control Systems Letters, vol. 5, no. 3, pp. 1013–1018, 2021

    work page 2021

  28. [28]

    Eigenvalue clustering, control energy, and logarithmic capacity,

    A. Olshevsky, “Eigenvalue clustering, control energy, and logarithmic capacity,”Systems & Control Letters, vol. 96, pp. 45–50, 2016

    work page 2016

  29. [29]

    Re: Warnings and caveats in brain controllability,

    F. Pasqualetti, S. Gu, and D. S. Bassett, “Re: Warnings and caveats in brain controllability,”NeuroImage, vol. 197, pp. 586–588, 2019

    work page 2019

  30. [30]

    Brain controllability: Not a slam dunk yet,

    S. Suweis, C. Tu, R. P. Rocha, S. Zampieri, M. Zorzi, and M. Corbetta, “Brain controllability: Not a slam dunk yet,”NeuroImage, vol. 200, pp. 552–555, 2019

    work page 2019

  31. [31]

    Warnings and caveats in brain controllability,

    C. Tu, R. P. Rocha, M. Corbetta, S. Zampieri, M. Zorzi, and S. Suweis, “Warnings and caveats in brain controllability,”NeuroImage, vol. 176, pp. 83–91, 2018

    work page 2018

  32. [32]

    Controllability scores for selecting control nodes of large-scale network systems,

    K. Sato and S. Terasaki, “Controllability scores for selecting control nodes of large-scale network systems,”IEEE Transactions on Automatic Control, vol. 69, no. 7, pp. 4673–4680, 2024

    work page 2024

  33. [33]

    Uniqueness Analysis of Controllability Scores and Their Application to Brain Networks,

    K. Sato and R. Kawamura, “Uniqueness Analysis of Controllability Scores and Their Application to Brain Networks,”IEEE Transactions on Control of Network Systems, 2025 (accepted)

    work page 2025

  34. [34]

    Energy scaling of targeted optimal control of complex networks,

    I. Klickstein, A. Shirin, and F. Sorrentino, “Energy scaling of targeted optimal control of complex networks,”Nature communications, vol. 8, no. 1, p. 15145, 2017

    work page 2017

  35. [35]

    Target control of complex networks,

    J. Gao, Y .-Y . Liu, R. M. D’souza, and A.-L. Barab ´asi, “Target control of complex networks,”Nature communications, vol. 5, no. 1, pp. 1–8, 2014

    work page 2014

  36. [36]

    A novel algorithm for finding optimal driver nodes to target control complex networks and its applications for drug targets identification,

    W.-F. Guo, S.-W. Zhang, Q.-Q. Shi, C.-M. Zhang, T. Zeng, and L. Chen, “A novel algorithm for finding optimal driver nodes to target control complex networks and its applications for drug targets identification,” Bmc Genomics, vol. 19, no. Suppl 1, p. 924, 2018

    work page 2018

  37. [37]

    Controlling target brain regions by optimal selection of input nodes,

    K. K. H. Manjunatha, G. Baron, D. Benozzo, E. Silvestri, M. Corbetta, A. Chiuso, A. Bertoldo, S. Suweis, and M. Allegra, “Controlling target brain regions by optimal selection of input nodes,”PLOS Computational Biology, vol. 20, no. 1, p. e1011274, 2024

    work page 2024

  38. [38]

    On the concepts of controllability and observability of linear systems,

    E. Kreindler and P. Sarachik, “On the concepts of controllability and observability of linear systems,”IEEE Transactions on Automatic Control, vol. 9, no. 2, pp. 129–136, 1964

    work page 1964

  39. [39]

    Fast projection onto the simplex and theL 1 ball,

    L. Condat, “Fast projection onto the simplex and theL 1 ball,”Mathe- matical Programming, vol. 158, no. 1, pp. 575–585, 2016

    work page 2016

  40. [40]

    D. P. Bertsekas,Nonlinear Programming, Third Edition. Athena Scientific, 2016

    work page 2016

  41. [41]

    R. A. Horn and C. R. Johnson,Matrix analysis. Cambridge university press, 2012

    work page 2012

  42. [42]

    Distributed control strategies for microgrids: An overview,

    E. Espina, J. Llanos, C. Burgos-Mellado, R. Cardenas-Dobson, M. Martinez-Gomez, and D. Saez, “Distributed control strategies for microgrids: An overview,”IEEe Access, vol. 8, pp. 193 412–193 448, 2020

    work page 2020

  43. [43]

    Modular and hierarchi- cally modular organization of brain networks,

    D. Meunier, R. Lambiotte, and E. T. Bullmore, “Modular and hierarchi- cally modular organization of brain networks,”Frontiers in neuroscience, vol. 4, p. 200, 2010

    work page 2010

  44. [44]

    Hierarchical organization of modularity in metabolic networks,

    E. Ravasz, A. L. Somera, D. A. Mongru, Z. N. Oltvai, and A.- L. Barab ´asi, “Hierarchical organization of modularity in metabolic networks,”science, vol. 297, no. 5586, pp. 1551–1555, 2002

    work page 2002

  45. [45]

    Model reduction based approximation of the output controllability gramian in large-scale networks,

    G. Casadei, C. Canudas-de Wit, and S. Zampieri, “Model reduction based approximation of the output controllability gramian in large-scale networks,”IEEE Transactions on Control of Network Systems, vol. 7, no. 4, pp. 1778–1788, 2020

    work page 2020

  46. [46]

    Beck,First-order methods in optimization

    A. Beck,First-order methods in optimization. SIAM, 2017

    work page 2017

  47. [47]

    Human brain structural connec- tivity matrices–ready for modelling,

    A. ˇSkoch, B. Reh ´ak Bu ˇckov´a, J. Mare ˇs, J. Tint ˇera, P. Sanda, L. Jajcay, J. Hor ´aˇcek, F. ˇSpaniel, and J. Hlinka, “Human brain structural connec- tivity matrices–ready for modelling,”Scientific Data, vol. 9, no. 1, p. 486, 2022

    work page 2022

  48. [48]

    Nonlinear responses in fMRI: the Balloon model, V olterra kernels, and other hemodynamics,

    K. J. Friston, A. Mechelli, R. Turner, and C. J. Price, “Nonlinear responses in fMRI: the Balloon model, V olterra kernels, and other hemodynamics,”NeuroImage, vol. 12, no. 4, pp. 466–477, 2000

    work page 2000

  49. [49]

    A guide to group effective connectivity analysis, part 1: First level analysis with DCM for fMRI,

    P. Zeidman, A. Jafarian, N. Corbin, M. L. Seghier, A. Razi, C. J. Price, and K. J. Friston, “A guide to group effective connectivity analysis, part 1: First level analysis with DCM for fMRI,”NeuroImage, vol. 200, pp. 174–190, 2019

    work page 2019

  50. [50]

    Network diffusion accurately models the relationship between structural and functional brain connec- tivity networks,

    F. Abdelnour, H. U. V oss, and A. Raj, “Network diffusion accurately models the relationship between structural and functional brain connec- tivity networks,”Neuroimage, vol. 90, pp. 335–347, 2014

    work page 2014

  51. [51]

    Controllability of structural brain networks,

    S. Gu, F. Pasqualetti, M. Cieslak, Q. K. Telesford, B. Y . Alfred, A. E. Kahn, J. D. Medaglia, J. M. Vettel, M. B. Miller, S. T. Grafton, and D. S. Bassett, “Controllability of structural brain networks,”Nature communications, vol. 6, no. 1, pp. 1–10, 2015

    work page 2015