Indexed text attachment
evolutionary-game-theory-in-higher-categories.tex
Source for Evolutionary Game Theory in Higher Categories
View paper · Download original file
Lines 2949–3048 of 3949. Source text is displayed literally.
2949basis vector. If the equalities of columns hold, define the2950coarse column using any microscopic representative of the2951block. The sums of its entries vanish because the columns2952of \(Q\) sum to zero. Its off-diagonal entries are nonnegative2953because the corresponding target blocks contain no source state.2954Thus the resulting matrix is a Markov generator. Equivalently,2955the intertwining holds for the matrix exponentials at every2956nonnegative time, by applying it to each power and summing2957the exponential series. The implication in the reverse2958direction follows by differentiating the exponentials at2959zero. For open Markov processes, preservation of interfaces2960and fluxes adds constraints to the condition on the generators2961\cite{baez-courser-markov}. In the nonlinear selection equation2962the fiber test applies to the centered payoff averages, whose2963weights are themselves microscopic state variables.2964 2965\section{Observed equilibria and stability}2966 2967Semiconjugacy carries a rest point of \(F\) to a rest point of2968\(G\). Over a target rest point the condition says2969\(Dq_xF(x)=0\), allowing motion tangent to a fiber. If \(V\)2970is a Lyapunov function for \(G\), then2971\(D(V\circ q)_xF(x)=DV_{q(x)}G(q(x))\). Thus the same2972monotonicity is inherited by the pullback, with strictness2973only where the coarse Lie derivative is strict. These facts2974identify the limit of equilibrium and stability information2975that can be recovered from aggregated coordinates alone.2976 2977\section{Finite-time error under a nonzero defect}2978 2979Suppose the target is represented in a Euclidean domain, as2980for an affine simplex. The defect of an observation is2981\[2982 r_q(x)=Dq_xF(x)-G(q(x)).2983\]2984Suppose \(G\) is \(L\)-Lipschitz on a region containing both2985the observed source solution \(q(x(t))\) and a target solution2986\(z(t)\) over \([0,T]\). If the defect along the source2987solution has norm at most \(\epsilon\), then for2988\(0\leq t\leq T\),2989\[2990 \|q(x(t))-z(t)\|2991 \le e^{Lt}\|q(x(0))-z(0)\|2992 +\begin{cases}2993 \epsilon(e^{Lt}-1)/L,& L>0,\\2994 \epsilon t,& L=0.2995 \end{cases}2996\]2997The integral equations for the two curves, followed by the2998triangle inequality, bound their distance at \(t\) by its2999initial value plus the integral of \(L\) times the current3000distance and \(\epsilon\). The integral form of Gronwall's3001inequality yields the claim. For polynomial fields, formal3002approximate differential equivalences give more specific3003ways to find and certify reductions \cite{cardelli-lumping}.3004 3005If \(h:N\to P\) compares \(G\) to a third field \(J\), the3006chain rule gives3007\[3008 r_{h\circ q}(x)=Dh_{q(x)}r_q(x)+r_h(q(x)).3009\]3010In particular a bound on \(Dh\) controls the magnification3011of the first reduction's defect under composition. The error3012estimate uses a specified time interval and hypotheses about3013the region visited by the comparison trajectories. Long-time3014conclusions need additional information. For example, if the3015target satisfies the contraction inequality3016\[3017 \langle u-v,G(u)-G(v)\rangle\leq -\lambda\|u-v\|^23018 \qquad (\lambda>0)3019\]3020on the comparison region, the error in Euclidean norm satisfies3021\[3022 \|q(x(t))-z(t)\|3023 \leq e^{-\lambda t}\|q(x(0))-z(0)\|3024 +\frac{\epsilon}{\lambda}(1-e^{-\lambda t}).3025\]3026For positive distance, differentiate its square, apply the3027contraction inequality and Cauchy--Schwarz, and divide by3028twice the distance. At distance zero the upper right3029derivative of the distance is at most \(\epsilon\), since3030the difference of velocities is then the defect itself.3031Integrating the resulting differential inequality gives the3032bound. If the two trajectories remain in the region, the3033bound holds on their whole common forward time domain.3034\par3035 3036\bookpart{V}{Descent, compression, and learning}3037\chapter{Fitness systems over varying contexts}3038\label{ch:contextual-games}3039 3040A population in an ecological or strategic context uses a set3041of available types, a state of its interactions, and a rule for3042evaluating prospective behavior. When contexts vary, these data3043can have nontrivial transition maps. A model can then be a family3044of games without having a single global numbering of the types.3045This chapter uses sheaves to state exactly what must agree between3046presentations.3047 3048\section{An internal strategy object}