pith. sign in

arxiv: 2604.08730 · v1 · submitted 2026-04-09 · 📡 eess.SY · cs.SY

Invariance of Competition Outcomes in Hypergraph Competitive Dynamics

Qi Zhao , Shaoxuan Cui , Baolin Zhang , Junwei Du , Yuanshi Zheng This is my paper

Pith reviewed 2026-05-10 16:54 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords hypergraph competitive dynamicsLotka-Volterrawinner-take-allhigher-order networksstability analysistensor algebramulti-way inhibition
0
0 comments X

The pith

Competitive selection outcomes on hypergraphs depend only on inhibition ratios and inputs, not on the detailed higher-order structure.

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

The paper examines a Lotka-Volterra model of competition extended to hypergraphs, where pairwise inhibition is joined by uniform multi-way terms from hyperedges. Mathematical analysis proves that equilibria exist and are stable in the same categories seen in ordinary graphs, with outcomes classified as winner-take-all, winner-share-all, or variant forms. The central result is that these outcomes are set by a handful of scalar quantities such as the self-inhibition to lateral-inhibition ratio and the external drive values. This finding indicates that the taxonomy of selection behaviors remains robust even when group interactions are present, which matters for predicting or engineering competition in systems where agents interact in clusters rather than pairs alone.

Core claim

In the proposed hypergraph Lotka-Volterra system, classical pairwise inhibition is augmented by multi-way interaction terms induced by the hyperedges of uniform hypergraphs. The existence, uniqueness, and stability of equilibria are established through stability theory and tensor algebra. The eventual selection outcome is relatively insensitive to hyperedge order and the specific higher-order coupling structure, and is instead determined by a small set of interpretable scalar parameters such as the ratio between self-inhibition and lateral-inhibition and the external inputs.

What carries the argument

The hypergraph-augmented Lotka-Volterra model with uniform multi-way inhibition, whose equilibria and stability are characterized by tensor algebra and classical stability theory.

If this is right

  • The same parameter conditions that produce winner-take-all in pairwise graphs continue to do so when hyperedges are added.
  • Higher-order terms change convergence speed and some steady-state values but preserve the overall outcome taxonomy of WTA, WSA, and VWTA.
  • Equilibria can be classified and their stability decided from scalar parameters without needing the full hypergraph adjacency details.
  • Numerical simulations confirm that the model produces the same qualitative selection modes as standard graphs despite added group interactions.

Where Pith is reading between the lines

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

  • Effective low-dimensional models based on inhibition ratios may suffice to predict competition results even when real networks contain many higher-order cliques.
  • This invariance could simplify the design of robust selection mechanisms in multi-agent or ecological systems by focusing tuning effort on a few ratios rather than topology.
  • The approach may extend to time-varying or non-uniform hypergraphs, where deviations from invariance would mark the breakdown of simple scalar control.

Load-bearing premise

That higher-order interactions can be faithfully captured by adding uniform multi-way inhibition terms to the classical pairwise Lotka-Volterra equations.

What would settle it

Fix the inhibition ratio and external inputs, then vary hyperedge order or coupling structure; if the stable outcome type changes from winner-take-all to winner-share-all or vice versa, the invariance claim is false.

Figures

Figures reproduced from arXiv: 2604.08730 by Baolin Zhang, Junwei Du, Qi Zhao, Shaoxuan Cui, Yuanshi Zheng.

Figure 2
Figure 2. Figure 2: Fig.2. It is easy to see that the system ultimately con [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 1
Figure 1. Figure 1: The state trajectories of all neurons in system (12) [PITH_FULL_IMAGE:figures/full_fig_p013_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: The state trajectories of all neurons in system (12) [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 6
Figure 6. Figure 6: The state trajectories of all neurons in system (22) [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The state trajectories of all neurons in system (22) [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The state trajectories of all neurons in system (22) [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 11
Figure 11. Figure 11: The state trajectories of all neurons in pairwise [PITH_FULL_IMAGE:figures/full_fig_p015_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: The state trajectories of all neurons in pairwise [PITH_FULL_IMAGE:figures/full_fig_p015_12.png] view at source ↗
Figure 10
Figure 10. Figure 10: The state trajectories of all neurons in pairwise [PITH_FULL_IMAGE:figures/full_fig_p015_10.png] view at source ↗
read the original abstract

Winner-take-all (WTA)--type selection is a fundamental mechanism in networked competition, yet its dependence on higher-order interactions remains insufficiently understood. We study a Lotka--Volterra competitive dynamics on higher-order networks, where classical pairwise inhibition is augmented by multi-way interaction terms induced by hyperedges of uniform hypergraphs. The proposed model shows multiple competitive outcomes, including WTA, winner-share-all (WSA), and variant winner-take-all (VWTA). The existence, uniqueness and stability of equilibria are rigorously proved through mathematical analysis, which relies on classical stability theory and recent advances in tensor algebra. We show that the eventual selection outcome is relatively insensitive to the hyperedge order and the specific higher-order coupling structure, and is instead determined by a small set of interpretable scalar parameters, such as the ratio between self-inhibition and lateral-inhibition and the external inputs. Numerical experiments support the theory by showing that higher-order interactions affect convergence and steady states, yet yield the similar outcome taxonomy (WTA/WSA/VWTA) as in standard graphs. These results provide a network-scientific explanation of the robustness of WTA-type outcomes under complex group interactions and offer principled guidance for designing selection mechanisms on higher-order networks.

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

2 major / 2 minor

Summary. The manuscript studies Lotka-Volterra competitive dynamics on uniform hypergraphs, augmenting classical pairwise inhibition with multi-way terms induced by hyperedges. It identifies three outcome classes (WTA, WSA, VWTA), proves existence/uniqueness/stability of equilibria via classical stability theory and tensor algebra, and claims that the eventual selection outcome is insensitive to hyperedge order and higher-order coupling details, being governed instead by a small set of scalar parameters (self-to-lateral inhibition ratio and external inputs). Numerical experiments are presented to show that higher-order interactions alter convergence rates and steady-state values but preserve the same outcome taxonomy as the pairwise case.

Significance. If the central invariance result and the supporting proofs hold, the work supplies a network-scientific account of why WTA-type selection remains robust under group interactions, together with interpretable parameters that classify regimes. The explicit use of tensor algebra for the higher-order analysis and the numerical confirmation of the taxonomy are concrete strengths that could guide mechanism design on hypergraphs in ecology, neural computation, and social systems.

major comments (2)
  1. [Model formulation] Model construction (around the augmented LV equations): the uniform multi-way inhibition assumption is load-bearing for the invariance claim, yet the manuscript does not compare it against non-uniform or weighted hyperedge couplings; a short sensitivity check would be needed to confirm that the scalar-parameter governance survives when the higher-order terms deviate from uniformity.
  2. [Mathematical analysis] Stability and uniqueness proofs: while classical theory plus tensor algebra are invoked, the text should explicitly delineate the parameter regimes (inhibition ratio and external-input values) for which uniqueness and asymptotic stability are guaranteed; boundary cases where the ratio approaches 0 or infinity could alter the claimed outcome taxonomy and must be addressed to support the insensitivity statement.
minor comments (2)
  1. The precise definition of 'variant winner-take-all (VWTA)' relative to standard WTA and WSA should be stated explicitly when the taxonomy is first introduced, to avoid ambiguity for readers.
  2. A specific citation or brief explanation of the 'recent advances in tensor algebra' used for the equilibrium analysis would improve traceability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and the recommendation for minor revision. The comments have helped us improve the clarity of the manuscript regarding model assumptions and mathematical analysis. We provide point-by-point responses below.

read point-by-point responses

  1. Referee: [Model formulation] Model construction (around the augmented LV equations): the uniform multi-way inhibition assumption is load-bearing for the invariance claim, yet the manuscript does not compare it against non-uniform or weighted hyperedge couplings; a short sensitivity check would be needed to confirm that the scalar-parameter governance survives when the higher-order terms deviate from uniformity.

    Authors: We acknowledge that the uniform assumption is central to our analysis, enabling the application of tensor algebra for proving the invariance of outcomes. The manuscript does not include comparisons to non-uniform couplings because the focus is on uniform hypergraphs to derive the clean invariance result. A comprehensive sensitivity check would require additional simulations and analysis beyond the current scope. In the revised version, we have added a discussion in the model section explaining why uniformity is assumed and noting that deviations could affect the scalar governance, with a brief remark on expected robustness for small perturbations. revision: partial

  2. Referee: [Mathematical analysis] Stability and uniqueness proofs: while classical theory plus tensor algebra are invoked, the text should explicitly delineate the parameter regimes (inhibition ratio and external-input values) for which uniqueness and asymptotic stability are guaranteed; boundary cases where the ratio approaches 0 or infinity could alter the claimed outcome taxonomy and must be addressed to support the insensitivity statement.

    Authors: We are grateful for the suggestion to delineate the parameter regimes more explicitly. The original proofs assume an inhibition ratio greater than one to guarantee the negative definiteness of the Jacobian and the uniqueness via tensor properties. We have revised the mathematical analysis section to include explicit statements: uniqueness and asymptotic stability are guaranteed when the self-to-lateral inhibition ratio r satisfies r > 1 and the external inputs are positive with their sum less than a critical value derived from the equilibrium equations. For the boundary cases, as r → 0, the system exhibits reduced competition leading to WSA outcomes, and as r → ∞, it recovers the pairwise model with WTA dominance. These cases are now addressed to confirm that the outcome taxonomy remains consistent, supporting the insensitivity claim within the studied regimes. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper derives invariance of competitive outcomes (WTA/WSA/VWTA taxonomy) via mathematical analysis of equilibria in an augmented Lotka-Volterra system on uniform hypergraphs. Existence, uniqueness, and stability are established using classical stability theory and tensor algebra applied to the model equations; the scalar parameters (self/lateral inhibition ratio, external inputs) enter as free inputs that classify equilibria rather than being fitted or defined to force the invariance result. No step reduces by construction to its own inputs, no predictions are statistically forced from subsets of data, and no load-bearing self-citations or ansatzes are invoked. The derivation is self-contained against external benchmarks of dynamical systems theory.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 1 invented entities

The model augments standard Lotka-Volterra with hyperedge-induced terms but introduces no new physical entities; it relies on classical dynamical systems assumptions plus a small number of scalar parameters that classify equilibria.

free parameters (2)
  • ratio between self-inhibition and lateral-inhibition
    Scalar parameter that determines which of WTA/WSA/VWTA occurs.
  • external inputs
    Additional scalars that shift the selection outcome.
axioms (2)
  • domain assumption Lotka-Volterra competitive dynamics with added multi-way inhibition terms
    Core modeling choice that higher-order interactions act as uniform hyperedge augmentations to pairwise terms.
  • standard math Existence, uniqueness and stability follow from classical stability theory and tensor algebra
    Invoked to prove properties of equilibria without further justification in the abstract.
invented entities (1)
  • variant winner-take-all (VWTA) no independent evidence
    purpose: New category in the outcome taxonomy for the hypergraph model
    Introduced to describe a partial-winner steady state distinct from classical WTA and WSA.

pith-pipeline@v0.9.0 · 5523 in / 1550 out tokens · 61689 ms · 2026-05-10T16:54:59.665863+00:00 · methodology

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

36 extracted references · 36 canonical work pages

  1. [1]

    Neuronal competition and selection during memory formation,

    J. H. Han, S. A. Kushner, A. P. Yiu, et al., “Neuronal competition and selection during memory formation,”Science, vol. 316, no. 5823, pp. 457–460, 2007

    work page 2007

  2. [2]

    Dynamic ensemble balance in direct-and indirect-pathway striatal projection neurons underlying decision-related action selection,

    S. Tang, L. Cui, J. Pan, and N. L. Xu, “Dynamic ensemble balance in direct-and indirect-pathway striatal projection neurons underlying decision-related action selection,”Cell Reports, vol. 43, no. 9, p. 114726, 2024

    work page 2024

  3. [3]

    The neuropsychology of ventral prefrontal cortex: decision-making and reversal learning,

    L. Clark, R. Cools, and T. Robbins, “The neuropsychology of ventral prefrontal cortex: decision-making and reversal learning,”Brain and Cognition, vol. 55, no. 1, pp. 41–53, 2004

    work page 2004

  4. [4]

    Analytical note on certain rhythmic relations in organic systems,

    A. J. Lotka, “Analytical note on certain rhythmic relations in organic systems,”Proc. Natl. Acad. Sci. U.S.A., vol. 6, no. 7, pp. 410–415, 1920

    work page 1920

  5. [5]

    Variations and fluctuations of the number of individuals in animal species living together,

    V. Volterra, “Variations and fluctuations of the number of individuals in animal species living together,”ICES J. Mar. Sci., vol. 3, no. 1, pp. 3–51, 1928

    work page 1928

  6. [6]

    Mapping the ecological networks of microbial communities,

    Y. Xiao, M. T. Angulo, J. Friedman, et al., “Mapping the ecological networks of microbial communities,”Nat. Commun., vol. 8, no. 1, p. 2042, 2017

    work page 2042

  7. [7]

    Lyapunov-based boundary control for a class of hyperbolic Lotka–Volterra systems,

    L. Pavel and L. Chang, “Lyapunov-based boundary control for a class of hyperbolic Lotka–Volterra systems,”IEEE Trans. Autom. Control, vol. 57, no. 3, pp. 701–714, 2011

    work page 2011

  8. [8]

    Analog integrated circuits for the Lotka-Volterra competitive neural networks,

    T. Asai, M. Ohtani, and H. Yonezu, “Analog integrated circuits for the Lotka-Volterra competitive neural networks,” IEEE Trans. Neural Netw., vol. 10, no. 5, pp. 1222–1231, 1999

    work page 1999

  9. [9]

    Dynamic stability conditions for Lotka- Volterra recurrent neural networks with delays,

    Z. Yi and K. K. Tan, “Dynamic stability conditions for Lotka- Volterra recurrent neural networks with delays,”Phys. Rev. E, vol. 66, no. 1, Art. no. 011910, 2002

    work page 2002

  10. [10]

    Oscillating dynamics of Lotka–Volterra neural networks with time-varying delays and distributed delays,

    M. Es-saiydy and M. Zitane, “Oscillating dynamics of Lotka–Volterra neural networks with time-varying delays and distributed delays,”Ricerche di Matematica, vol. 73, no. 5, pp. 2779–2799, 2024

    work page 2024

  11. [11]

    Lotka–Volterra like dynamics in phase oscillator networks,

    C. Bick, “Lotka–Volterra like dynamics in phase oscillator networks,” inAdvances in Dynamics, Patterns, Cognition: Challenges in Complexity, vol. 20, pp. 115–125, 2017

    work page 2017

  12. [12]

    A simple neural network exhibiting selective activation of neuronal ensembles: From winner-take- all to winners-share-all,

    T. Fukai and S. Tanaka, “A simple neural network exhibiting selective activation of neuronal ensembles: From winner-take- all to winners-share-all,”Neural Comput., vol. 9, no. 1, pp. 77–97, 1997

    work page 1997

  13. [13]

    Using competitive layer model implemented by Lotka–Volterra recurrent neural networks for detecting brain activated regions from fMRI data,

    B. Zheng and Z. Yi, “Using competitive layer model implemented by Lotka–Volterra recurrent neural networks for detecting brain activated regions from fMRI data,”Neural Comput. Appl., vol. 22, no. 1, pp. 395–404, 2013

    work page 2013

  14. [14]

    A winner-take-all Lotka–Volterra recurrent neural network with only one winner in each row and each column,

    B. Zheng, “A winner-take-all Lotka–Volterra recurrent neural network with only one winner in each row and each column,” Neural Comput. Appl., vol. 24, no. 1, pp. 1749–1757, 2014

    work page 2014

  15. [15]

    Scientific machine learning in ecological systems: A study on the predator-prey dynamics,

    R. Devgupta, R. A. Dandekar, R. Dandekar, and S. Panat, “Scientific machine learning in ecological systems: A study on the predator-prey dynamics,” arXiv preprint arXiv:2411.06858, 2024

    work page arXiv 2024

  16. [16]

    Enhanced species coexistence in Lotka-Volterra competition models due to nonlocal interactions,

    G. A. Maciel and R. Martinez-Garcia, “Enhanced species coexistence in Lotka-Volterra competition models due to nonlocal interactions,”J. Theor. Biol., vol. 530, p. 110872, 2021

    work page 2021

  17. [17]

    Arguments in favor of higher order interactions,

    P. A. Abrams, “Arguments in favor of higher order interactions,”Am. Nat., vol. 121, no. 6, pp. 887–891, 1983

    work page 1983

  18. [18]

    An SIS diffusion process with direct and indirect spreading on a hypergraph,

    S. Cui, F. Liu, L. Liang, H. Jard´ on-Kojakhmetov, and M. Cao, “An SIS diffusion process with direct and indirect spreading on a hypergraph,”Automatica, vol. 158, p. 111035, 2025

    work page 2025

  19. [19]

    Higher-order interactions capture unexplained complexity in diverse communities,

    M. M. Mayfield and D. B. Stouffer, “Higher-order interactions capture unexplained complexity in diverse communities,”Nat. Ecol. Evol., vol. 1, no. 3, p. 0062%, 2017

    work page 2017

  20. [20]

    Dynamical systems on hypergraphs,

    T. Carletti, D. Fanelli, and S. Nicoletti, “Dynamical systems on hypergraphs,”J. Phys. Complexity, vol. 1, no. 3, p. 035006, 2020

    work page 2020

  21. [21]

    The physics of higher-order interactions in complex systems,

    F. Battiston, E. Amico, A. Barrat, et al., “The physics of higher-order interactions in complex systems,”Nat. Phys., vol. 17, no. 10, pp. 1093–1098, 2021

    work page 2021

  22. [22]

    On discrete-time polynomial dynamical systems on hypergraphs,

    S. Cui, G. Zhang, H. Jard´ on-Kojakhmetov, and M. Cao, “On discrete-time polynomial dynamical systems on hypergraphs,” IEEE Control Syst. Lett., vol. 8, pp. 1078–1083, 2024

    work page 2024

  23. [23]

    On Metzler positive systems on hypergraphs,

    S. Cui, G. Zhang, H. Jard´ on-Kojakhmetov, and M. Cao, “On Metzler positive systems on hypergraphs,”IEEE Trans. Control Netw. Syst., vol. 12, no. 2, pp. 345–356, 2025

    work page 2025

  24. [24]

    The mechanistic basis for higher-order interactions and non-additivity in competitive communities,

    A. D. Letten and D. B. Stouffer, “The mechanistic basis for higher-order interactions and non-additivity in competitive communities,”Ecol. Lett., vol. 22, no. 3, pp. 423–436, 2019

    work page 2019

  25. [25]

    Higher-order interactions in random Lotka-Volterra communities,

    L. Sidhom and T. Galla, “Higher-order interactions in random Lotka-Volterra communities,” arXiv preprint arXiv:2409.10990, 2024

    work page arXiv 2024

  26. [26]

    Hyperedge prediction and the statistical mechanisms of higher-order and lower-order interactions in complex networks,

    M. Sales-Pardo, A. Marin´ e-Tena, and R. Guimer` a, “Hyperedge prediction and the statistical mechanisms of higher-order and lower-order interactions in complex networks,”Proc. Natl. Acad. Sci. U.S.A., vol. 120, no. 50, e2303887120, 2023

    work page 2023

  27. [27]

    Higher order interactions and species coexistence,

    P. Singh and G. Baruah, “Higher order interactions and species coexistence,”Theor. Ecol., vol. 14, no. 1, pp. 71–83, 2021

    work page 2021

  28. [28]

    Coexistence in diverse communities with higher-order interactions,

    T. Gibbs, S. A. Levin, and J. M. Levine, “Coexistence in diverse communities with higher-order interactions,”Proc. Natl. Acad. Sci. U.S.A., vol. 119, no. 43, e2205063119, 2022

    work page 2022

  29. [29]

    On the analysis of a higher-order Lotka-Volterra model: An application of S-tensors and the polynomial complementarity problem,

    S. Cui, Q. Zhao, G. Zhang, H. Jardon-Kojakhmetov, and M. Cao, “On the analysis of a higher-order Lotka-Volterra model: An application of S-tensors and the polynomial complementarity problem,” IEEE Transactions on Automatic Control, 2025

    work page 2025

  30. [30]

    Directed hypergraphs and applications,

    G. Gallo, G. Longo, S. Pallottino, and S. Nguyen, “Directed hypergraphs and applications,”Discrete Appl. Math., vol. 42, nos. 2–3, pp. 177–201, 1993

    work page 1993

  31. [31]

    M-tensors and nonsingular m-tensors,

    W. Ding, L. Qi, and Y. Wei, “M-tensors and nonsingular m-tensors,”Linear Algebra Appl., vol. 439, no. 10, pp. 3264– 3278, 2013

    work page 2013

  32. [32]

    On metzler positive systems on hypergraphs,

    S. Cui, G. Zhang, H. Jardon-Kojakhmetov, and M. Cao, “On metzler positive systems on hypergraphs,” arXiv preprint arXiv:2401.03652, 2024

    work page arXiv 2024

  33. [33]

    Existence and uniqueness of positive solution forH +-tensor equations,

    X. Wang, M. Che, and Y. Wei, “Existence and uniqueness of positive solution forH +-tensor equations,”Appl. Math. Lett., vol. 98, pp. 191–198, 2019

    work page 2019

  34. [34]

    Eigenvalues of a real supersymmetric tensor,

    L. Qi, “Eigenvalues of a real supersymmetric tensor,”J. Symb. Comput., vol. 40, no. 6, pp. 1302–1324, 2005

    work page 2005

  35. [35]

    Statistical mechanics of nervous nets,

    J. D. Cowan, “Statistical mechanics of nervous nets,” inNeural Networks: Proceedings of the School on Neural Networks - June, Berlin, Germany: Springer Berlin Heidelberg, 1968

    work page 1968

  36. [36]

    A statistical mechanics of nervous activity,

    J. D. Cowan, “A statistical mechanics of nervous activity,” inLectures on Mathematics in the Life Sciences, vol. 2, pp. 1–57, 1970. 16

    work page 1970