Indexed text attachment
evolutionary-game-theory-in-higher-categories.tex
Source for Evolutionary Game Theory in Higher Categories
View paper · Download original file
Lines 3281–3380 of 3949. Source text is displayed literally.
3281For an arbitrary test vector $u\in\mathbb R^J$,3282\begin{align*}3283 C_A(x)&=\sum_i x_i3284 \bigl(\operatorname{diag}(a_i)-a_i a_i^\top\bigr),\\3285 u^\top C_A(x)u&=\sum_i x_i3286 \left(\sum_j (a_i)_j u_j^2-3287 \left(\sum_j(a_i)_j u_j\right)^2\right).3288\end{align*}3289The summands are conditional variances and are nonnegative.3290A variance is zero precisely when $u$ is constant on the3291support of the corresponding column. For a positive3292semidefinite real symmetric matrix $C$, $u^\top C u=0$3293is equivalent to $Cu=0$ by diagonalization.3294Taking $u=g(Ax)$ proves the equality criterion. Every3295deterministic column has zero covariance. Conversely choose3296$x$ concentrated on any specified $i$ and allow a constant3297target fitness taking an arbitrary vector value. Universal3298vanishing then forces every coordinate indicator to have zero3299variance under $a_i$, so every coordinate probability is either3300zero or one. Their sum is one, proving determinism.3301\end{proof}3302 3303Deterministic compression groups types without spreading any3304type across distinct observed labels. Its exactness is the3305coordinate expression of summing replicator derivatives3306over the fibers of a function, when the payoffs agree within3307fibers. In stochastic compression, equality can still hold3308for particular payoffs or particular support faces. It is3309then a condition on the conditional distribution and the3310payoff together.3311 3312\section{Composition of two compressions}3313 3314For a second channel $B:J\to K$ the defects obey3315\[3316 C_{BA}(x)=C_B(Ax)+B C_A(x)B^\top.3317\]3318Indeed, first add and subtract3319$B\operatorname{diag}(Ax)B^\top$ in the expression for3320$C_{BA}(x)$. The two terms are the covariance from the3321second channel and the covariance of its conditional means3322induced by the first. This is the law of total covariance3323for the original sampling sequence. The formula treats3324loss of a selection signal as a two-stage correlation3325calculation.3326 3327The two summands are positive semidefinite. If the final3328fitness is $v\in\mathbb R^K$, exact pushforward through3329the composite holds at $x$ precisely when3330$C_B(Ax)v=0$ and $C_A(x)B^\top v=0$. To see this,3331test the displayed expression in the quadratic form at3332$v$. It is a sum of two nonnegative numbers, each with the3333zero criterion of Theorem~\ref{thm:channel-defect}.3334The two equality conditions correspond to asking that3335the final payoff be constant on the occupied conditional3336supports of the second stage and that its first-stage3337conditional mean be constant on the relevant supports of the3338first. This criterion can be applied before composing the3339channels.3340 3341\[3342 \begin{tikzcd}[column sep=large,row sep=large]3343 \Delta(I) \arrow[r,"A"] \arrow[dr,"BA"'] &3344 \Delta(J) \arrow[d,"B"]\\3345 &\Delta(K)3346 \end{tikzcd}3347\]3348If each step is deterministic, the maps and the exact3349fitness pullbacks compose without defect. This yields a3350class of morphisms of selection systems. The same category3351of stochastic maps can be used for compression with a3352nonzero defect. In that case the formula identifies the3353specific obstruction to the map being a morphism of3354replicator dynamics.3355 3356\section{Trajectory error on a bounded time interval}3357 3358Write $x(t)$ for the fine solution and $z(t)$ for the3359coarse solution starting at $Ax(0)$. Suppose both remain3360on their simplices and that $R_g$ is Lipschitz there with3361constant $L$ for a fixed norm. Define3362$c_t=C_A(x(t))g(Ax(t))$. The projected fine solution3363satisfies3364\[3365 \frac{d}{dt}Ax(t)=R_g(Ax(t))-c_t .3366\]3367Subtracting the coarse equation and applying the3368integral form of Gr\"onwall's inequality yields3369\[3370 \|Ax(t)-z(t)\|3371 \leq\int_0^t e^{L(t-s)}\|c_s\|\,ds .3372\]3373One derivation is to integrate the difference of the3374two derivatives, bound the target field difference by3375$L\|Ax-z\|$, and iterate the resulting integral inequality.3376Its Volterra series sums to the displayed exponential.3377If the defect is identically zero, uniqueness gives exact3378semiconjugacy. If the defect is merely small, the estimate3379controls finite-time approximation. It does not give a3380uniform long-time guarantee when the bound grows with $t$.