1: \begin{abstract}

2: 

3: We provide several characterizations of convergence to unstable equilibria in nonlinear systems. Our current contribution is three-fold. First we present

4:  simple algebraic conditions for establishing local convergence of non-trivial solutions of nonlinear systems to unstable equilibria. The conditions are based on the earlier work \cite{Tyukin:SIAM:2013} and can be viewed as an extension of the Lyapunov's first method in that they apply to systems in which the corresponding Jacobian has one zero eigenvalue. Second, we show that for a relevant subclass of systems, persistency of excitation of a function of time in the right-hand side of the equations governing dynamics of the system ensure existence of an attractor basin such that solutions passing through this basin in forward time converge to the origin exponentially. Finally we demonstrate that conditions developed in  \cite{Tyukin:SIAM:2013} may be remarkably tight.

5: %

6: %for a class of systems

7: %with unstable invariant sets. Examples of systems to

8: %which the results may apply include interconnections of stable

9: %subsystems with one-dimensional unstable dynamics or critically stable

10: %dynamics. Systems of this type arise in  problems of nonlinear

11: %output regulation, parameter estimation and adaptive control.

12: %In addition to providing boundedness and convergence criteria

13: %the results allow to derive domains of initial conditions

14: %corresponding to solutions leaving a given neighborhood of the

15: %origin at least once. In contrast to other works addressing

16: %convergence issues in unstable systems, our results require

17: %neither  input-output characterizations for

18: %the stable part nor estimates of convergence rates.

19: %The results are illustrated with examples, including the analysis of phase synchronization of neural oscillators with heterogenous coupling.

20: \end{abstract}