Independent scholarship · Open accessThe monthly digest

Indexed text attachment

evolutionary-game-theory-in-higher-categories.tex

Source for Evolutionary Game Theory in Higher Categories

View paper · Download original file

Lines 1601–1700 of 3949. Source text is displayed literally.

1601positive time. Brouwer's fixed point theorem ensures a1602rest point for continuous \(F\). Every rest point is1603interior and satisfies \(x=B^\eta(x)\).1604 1605If \(\eta_k\to0\) through positive parameters and1606\(x^{(k)}=B^{\eta_k}(x^{(k)})\) converges, its limit is Nash.1607For a strategy strictly suboptimal at the limit, continuity1608gives a payoff gap of at least some \(\delta>0\) compared1609with a maximizing strategy there, for all large \(k\).1610Its probability in \(x^{(k)}\) is at most1611\(\exp(-\delta/\eta_k)\), hence is zero in the limit.1612Only maximizers can retain mass. This argument does not1613choose which Nash states arise along branches of fixed points.1614 1615The potential case gives a convergence result at a fixed1616noise level. Let \(V\) be a potential in the sense of1617Chapter~\ref{ch:lyapunov}, continuously differentiable on1618the simplex in its affine hull. On the interior the logit1619dynamics satisfy1620\begin{align*}1621 \frac{d}{dt}(V+\eta H)(x)1622 &=\eta\sum_i (B^\eta_i(x)-x_i)1623                  \bigl(\log B^\eta_i(x)-\log x_i\bigr)\\1624 &=\eta\bigl[D(B^\eta(x)\Vert x)+D(x\Vert B^\eta(x))\bigr].1625\end{align*}1626The first equality uses1627\(F_i(x)=\eta\log B^\eta_i(x)+\eta\log\sum_j e^{F_j(x)/\eta}\)1628and the fact that the velocity sums to zero. Thus \(V+\eta H\)1629strictly increases except at rest points. The interior bounds1630after positive time make the compactness argument of1631Chapter~\ref{ch:lyapunov} applicable. If \(V\) is concave,1632then \(V+\eta H\) is strictly concave on the simplex. Each1633interior logit rest point satisfies its constrained1634first-order condition, so there is only one. All trajectories1635converge to it. The existence, limit, and convergence1636statements use different hypotheses on fitness and on the1637response rule. More general perturbation and revision1638theories are developed by Sandholm \cite{sandholm-population}.1639\par1640\chapter{Interacting populations and spatial structure}1641\label{ch:structured}1642 1643The comparison neighborhood in Chapter~\ref{ch:revision} consisted of1644individuals drawn from the same population. An individual can instead1645interact frequently with nearby populations while drawing its behavioral1646replacement from its own population. An individual who moves changes the1647composition of both the source and destination. These mechanisms give1648different equations and lead to different forms of reduction. We study1649exact synchronized subspaces for a deterministic interaction model and1650then derive a quantitative synchronization criterion for migration.1651 1652\section{Product simplices and assortative encounters}1653 1654Let $I$ be a finite set of populations and $S_r$ a finite set of types1655for each $r\in I$. A state in1656$\mathcal D_I=\prod_{r\in I}\Delta(S_r)$ records conditional1657frequencies within the populations. The payoff to type $i\in S_r$ is1658$F_i^r(x)$, potentially a function of the full profile. Local1659normalization gives1660\[1661  \phi_r(x)=\sum_{i\in S_r}x_i^r F_i^r(x),1662  \qquad1663  \dot x_i^r=x_i^r\bigl(F_i^r(x)-\phi_r(x)\bigr).1664\]1665Each population has its own conserved mass. Assuming a locally1666Lipschitz extension near the product, the proof of1667Chapter~\ref{ch:replicator} applies simultaneously. The equation for each1668population's mass receives its own continuous coefficient $\phi_r(x)$,1669and each coordinate still has an integrating factor. Compactness of the1670product yields the global flow. A choice of one face per population1671gives an invariant product face. If every $F^r$ depends solely on $x^r$,1672the flow is the product of the separate population flows. If payoffs1673couple populations, a product face can remain invariant while the flow1674within it is coupled.1675 1676Take a common type set $S$, a matrix $A$ of row-player payoffs on $S$,1677and a row stochastic matrix $W=(W_{rs})$ on $I$. An encounter by a1678resident of $r$ draws its partner's population from row $r$ and then1679draws the type from that population. The vector of expected payoffs is1680\[1681  F^r_W(x)=A\sum_{s\in I} W_{rs}x^s.1682\]1683Diagonal weights quantify within-population encounters. Directed weights1684allow asymmetric access to partners. The state equations apply the1685selection step within each destination of the payoff map, retaining1686separate local averages even when all populations see the same1687distribution of partners. At a stationary profile each resident type in1688each population has the same payoff. Comparison with absent types is1689required to promote this equality to a population Nash condition.1690 1691Here $W$ is part of a deterministic model posed directly at the1692population level. For an individual-based graph process, replacing, for1693example, the conditional distribution of neighbors of type $i$ by the1694unconditional population mixture is an additional modeling step.1695Ohtsuki and Nowak derive update-dependent replicator approximations for1696regular graphs using weak selection and pair1697approximation~\cite{ohtsuki-nowak}. That derivation treats local1698correlations explicitly. The present equation follows directly from the1699specified partner sampling.1700