Update docs/src/catalyst_applications/dynamical_systems.md

Co-authored-by: George Datseris <[email protected]>

Author: TorkelE

docs/src/catalyst_applications/dynamical_systems.md

@@ -51,7 +51,7 @@ Here, in addition to the basins of attraction, the system's three steady states
 More information on how to compute basins of attractions for ODEs using DynamicalSystems.jl can be found [here](https://juliadynamics.github.io/DynamicalSystemsDocs.jl/attractors/stable/basins/).
 
 ## [Computing Lyapunov exponents](@id dynamical_systems_lyapunov_exponents)
-[*Lyapunov exponents*](https://en.wikipedia.org/wiki/Lyapunov_exponent) are scalar values that can be computed for any state of an ODE. For an ODE with $n$ variables, for each state, a total of $n$ Lyapunov exponents can be computed (and these are collectively called the *Lyapunov spectrum*). Large Lyapunov exponents indicates that trajectories infinitesimally close to the given state diverge from each other. Conversely, small Lyapunov exponents suggests that trajectories converge to each others. 
+[*Lyapunov exponents*](https://en.wikipedia.org/wiki/Lyapunov_exponent) are scalar values that can be computed for any attractor of an ODE. For an ODE with $n$ variables, for each state, a total of $n$ Lyapunov exponents can be computed (and these are collectively called the *Lyapunov spectrum*). Positive Lyapunov exponents indicate that trajectories initially infinitesimally close diverge from each other. Conversely, negative Lyapunov exponents suggests that trajectories converge to each others. 
 
 While Lyapunov exponents can be used for other purposes, they are primarily used to characterise [*chaotic behaviours*](https://en.wikipedia.org/wiki/Chaos_theory) (where small changes in initial conditions has large effect on the resulting trajectories). Generally, an ODE exhibit chaotic behaviours in a point in phase space if that point have *at least one* positive Lyapunov exponent. Practically, Lyapunov exponents can be computed using DynamicalSystems.jl's `lyapunovspectrum` function. Here we will use it to investigate two models, one which exhibits chaos and one which do not.