Solitary states in spiking oscillators with higher-order interactions
Vladimir V. Semenov, Subhasanket Dutta, Stefano Boccaletti, Charo I. del Genio, Sarika Jalan, Anna Zakharova · 2025
Read Solitary states in spiking oscillators with higher-order interactions by Vladimir V. Semenov, Subhasanket Dutta, Stefano Boccaletti, Charo I. del Genio, Sarika Jalan, Anna Zakharova (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-19016.txt.
Introduction
Vladimir Semenov and co-authors study lone rebels in a crowd of rhythm cells. Their system links many FitzHugh-Nagumo units, a standard math model of a spiking nerve cell. When the units sync up, most fire together, but a few may break away and fire on their own, forming what are called solitary states.
The twist is that the units interact not just in pairs but also in groups, a setup called higher-order interaction. The team shows these group links change how the crowd shifts between synced and unsynced states. Near that shift, lone units appear when the system is tuned either up or down at weak coupling. At stronger coupling, they appear only on the way up, while the way down breaks apart all at once.
Catalog record
| CREATOR | Vladimir V. Semenov, Subhasanket Dutta, Stefano Boccaletti, Charo I. del Genio, Sarika Jalan, Anna Zakharova |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-19016
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Solitary states in spiking oscillators with higher-order interactions Vladimir V. Semenov, Subhasanket Dutta, Stefano Boccaletti, Charo I. del Genio, Sarika Jalan, Anna Zakharova Abstract We study a system of globally coupled FitzHugh-Nagumo oscillators, showing that the presence of higher-order interactions affects the character of the transition between synchronous and asynchronous states. In particular, we demonstrate that, around the synchronization transition, solitary states emerge due to the presence of second-order interactions. In difference to the phenomenology observed in systems of phase oscillators, we show that, at low coupling strengths, solitary states appear for both transition directions, whereas for higher couplings they only occur in the forward direction, with the backwards one characterized by explosive desynchronization. I Introduction In the past decade, the presence of higher-order interactions has been recognized as having a substantial impact on the behaviour of a large number of complex systems Bat20 ; Bat22 . Their effects are particularly relevant because in the general case they cannot be recovered via the superposition of pairwise interactions. This has caused a strong return of interest in the study of the static and dynamical properties of the structures that represent them, namely hypergraphs and simplicial complexes Boc23 ; del24 ; Ma25 . The influence of higher-order interactions on collective dynamics has been investigated in particular depth for networks of oscillators, where their relation to phenomena such as synchronization and resonance has been well established Tla19 ; Gal22 ; Par22 ; Sem25 . In such systems, one of their most striking effects is that they can give rise to abrupt desynchronization transitions in the absence of abrupt synchronization counterparts. Additionally, they can induce extreme multistability, whereby infinitely many stable and partially synchronized states coexist Ska19 . Such coexistence relates to the remarkable phenomenon of solitary states, in which one or more dynamical units split off and behave differently from the others despite the coupling being homogeneous Jar18 ; Maj19 ; Hel20 ; Sch21 . The effect of partial synchronization caused by higher-order interactions can also result in the formation of chimera states Kun22 and in the induction of abrupt synchronization transitions with hysteresis and bistability of synchronized and incoherent states Ska20 ; Kac22 , ultimately causing explosive synchronization Boc16 . However, these complex dynamical behaviours have so far been observed and studied only in networks of phase oscillators. In this article, we show how higher-order interactions have a significant effect also on the transitions occurring in globally coupled systems of spiking oscillators. More specifically, we show how higher-order interactions cause the emergence of solitary states in both directions along the route to and from synchronization. We then characterize the specific effects of the different orders of interaction on the phenomenology of these systems. II The model We study a system of globally coupled FitzHugh-Nagumo oscillators in the presence of both pairwise and second-order interactions. Its general evolution is described by the following system: \displaystyle\varepsilon\dot{x}_{i} \displaystyle=x_{i}-\frac{x_{i}^{3}}{3}-y_{i}+\frac{\sigma_{p}}{N}\sum_{\begin{subarray}{c}j=1\\ j\neq i\end{subarray}}^{N}\tanh(x_{j}-x_{i}) (1) \displaystyle+\frac{\sigma_{h}}{2N^{2}}\sum_{\begin{subarray}{c}j=1\\ j\neq i\end{subarray}}^{N}\sum_{\begin{subarray}{c}k=1\\ k\neq j\end{subarray}}^{N}\tanh(x_{j}+x_{k}-2x_{i})\>, \displaystyle\dot{y}_{i} \displaystyle=x_{i}+a_{i}-by_{i}\>. Here, x_{i} and y_{i} are the state variables of the i th oscillator, the parameter \varepsilon\ll 1 is responsible for the time scale separation of the state variables, N is the number of oscillators, and \sigma_{p} and \sigma_{h} are the strengths of pairwise and higher-order interactions, respectively. We consider the system in Eq.( 1 ) in the particular case of at b=0 , so that the equations are akin to those of a van der Pol model. Note that, in this interpretation, \varepsilon is the inverse of the nonlinear damping. Furthermore, we consider the oscillators to be slightly different from each other, by imposing that a_{i} be random variables following a normal distribution with mean 0.5 and variance 0.001 . The model as defined involves nonlinear couplings in terms of the tanh function. This specific choice allows one to account for a saturation effect for large values of the function argument, a behaviour that is often desired when modelling biological processes such as neural cell dynamics Sni20 ; Hej21 . Figure 1: The transition to synchronization is continuous in the absence of higher-order interactions. (a) The time-averaged order parameter \left\langle r\right\rangle changes continuously with the two-body interaction strength \sigma_{p} . The forward curve (solid red line) and the backwards one (solid black line) can be superimposed, and no hysteresis cycle occurs. The character of the transition remains the same if one uses a linear coupling, rather than the nonlinear one of Eq. ( 1 ) (dashed black line). Note the logarithmic scale for \sigma_{p} . (b)–(d) Space-time snapshots of the system, taken at the points 1–3 in panel (a), illustrate how, as \sigma_{p} increases above a critical value, the oscillators achieve global synchronization. In these panels, point 1 corresponds to \sigma_{p}=0.004 , point 2 to \sigma_{p}=0.02 and point 3 to \sigma_{p}=0.4 . The upper insets show the final states of the system, which are used in the simulations as the initial conditions for the next points. We study the system via numerical simulations, using N=100 oscillators and \varepsilon=-0.01 , starting from uniformly distributed random initial conditions x_{i}\in[-1.5,1.5] and y_{i}\in[-1,1] . The steady state of a run for a given value of the coupling strengths is used as the initial state of the next run. At every time step, we compute the value of the order parameter r(t)=\frac{1}{N}\left|\sum_{i=1}^{N}\mathrm{e}^{\mathrm{i}\theta_{i}}\right| , where \theta_{i}=\arctan(\frac{y_{i}}{x_{i}}) is the angle of the i th oscillator. The range of the time-averaged order parameter \left\langle r\right\rangle , averaged over the steady state of the system, is between 0 and 1, with 0 corresponding to an anynchronous state, and 1 to fully synchronized oscillations. Additionally, we also compute the mean phase velocity for each oscillator, \omega_{i}=2\pi\frac{M_{i}}{\Delta T} , where M_{i} is the number of complete rotations around the origin performed by the i th oscillator during a time interval of size \Delta T . III Results III.1 Exclusively pairwise interactions Figure 2: Higher-order interactions make the synchronization transition discontinuous and induce solitary states. (a) A hysteresis cycle appears in the transition to synchronization: the system switches to a synchronized state at \sigma_{p}=-0.6 (red line), but it reverts back to the anynchronous state only when \sigma_{p} is lowered to -0.69 (black line), indicating that the transition is first-order. (b)–(d) Space-time snapshots of the system, taken at the points 1–3 in panel (a), show the appearance of solitary states on the route to synchronization and to desynchronization. In these panels, point 1 corresponds to \sigma_{p}=-0.73 , point 2 to \sigma_{p}=-0.58 and point 3 to \sigma_{p}=-0.58 . The upper insets show the final states of the system, which are used in the simulations as the initial conditions for the next points. (e)–(f) The phase portraits of the synchronized oscillators (black) are limit cycles, whereas those of the oscillators corresponding to the solitary state (red) are open trajectories. The upper insets show the mean phase velocities of each individual oscillator. Panel (e) corresponds to point 2 and panel (f) to point 3. We start by considering a system with only pairwise interaction, i.e., with \sigma_{h}=0 . In such a case, a continuous transition between synchronous and asynchronous dynamics occurs as a function of the interaction strength. As shown in Fig. 1 (a), no hysteresis appears in the time-averaged order parameter, and