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 1097–1196 of 2037. Source text is displayed literally.

1097motion of the class totals. \chk{O7}1098\end{enumerate}1099\end{proposition}1100\begin{proof}1101(a) The mean dynamic is $\dot x_i=x_i\sum_jx_j\bigl(r(\pi_j,\pi_i)-r(\pi_i,\pi_j)\bigr)$. The factor $x_i$ gives1102invariance of faces, and the sum depends on $x$ only through the distribution of payoffs, which is unchanged1103when clones are merged. (b) is immediate from the definitions. (c) In both protocols the inflow into strategy1104$i$ does not carry the factor $x_i$. A strategy with two names receives the inflow twice: for1105Rock--Paper--Scissors with Rock duplicated the discrepancy between the pushed-forward field and the field of1106the quotient game reached $0.56$ for Smith and $0.38$ for Brown--von Neumann--Nash over fifty random states.1107\end{proof}1108 1109\begin{remark}1110Imitation counts individuals; a direct protocol counts names. Under imitation a strategy is as visible as it1111is common, so a twin of a strategy that exists is invisible as a separate thing, and a dominated twin dies at1112the rate of its penalty. Under a direct protocol every name is a candidate at every moment whatever its1113frequency, so a twin is a second target for switching, and it is replenished each time the cycle makes it1114nearly optimal. \Cref{thm:hs} is the price of Nash stationarity: one cannot have both the invariance of1115\Cref{thm:natural} and rest points that are exactly the Nash equilibria.1116\end{remark}1117 1118\section{Populations as open systems}1119 1120In \eqref{eq:rep} the game and the learning rule are fused. They can be separated.1121 1122\begin{definition}1123A \term{population box} on $S$ is an open dynamical system with input a payoff vector $\pi\in\R^S$, state1124$x\in\Dist(S)$, output the state, and dynamics $\dot x=V(x,\pi)$. The \term{replicator box} has1125$V_i=x_i(\pi_i-x\cdot\pi)$.1126\end{definition}1127 1128\begin{proposition}[The replicator box integrates and normalizes]\label{prop:softmax}1129\Scl For any integrable input $\pi(\cdot)$ and interior initial state, the replicator box has the solution1130\[ x(t)=\operatorname{softmax}\Bigl(\log x(0)+\int_0^t\pi(\tau)\,d\tau\Bigr). \]1131In the coordinates $z=\log x$ modulo constants the box is the integrator $\dot z=\pi$. \chk{O6}1132\end{proposition}1133 1134\noindent All the nonlinearity of the replicator equation is in the game and in the readout, the map1135$\operatorname{softmax}$ from $\R^S/\R\one$ onto the open simplex; the box itself is linear. The Lyapunov function of \Cref{thm:kl} is1136the Bregman divergence of the convex function whose gradient is this readout.1137 1138A game is then a way of closing boxes.1139\begin{center}1140\begin{tikzpicture}[>=stealth,scale=0.95,every node/.style={font=\small}]1141\node[draw,thick,minimum width=1.9cm,minimum height=0.9cm] (p) at (0,0) {population};1142\node[draw,minimum width=1cm,minimum height=0.7cm] (a) at (0,-1.5) {$A$};1143\draw[->] (p.east) -- ++(0.8,0) |- node[pos=0.25,right]{$x$} (a.east);1144\draw[->] (a.west) -- ++(-1.25,0) |- node[pos=0.25,left]{$\pi=Ax$} (p.west);1145\node at (0,-2.4) {one population};1146\begin{scope}[xshift=6.3cm]1147\node[draw,thick,minimum width=1.5cm,minimum height=0.9cm] (r) at (-1.6,0) {row};1148\node[draw,thick,minimum width=1.5cm,minimum height=0.9cm] (c) at (1.6,0) {column};1149\node[draw,minimum width=0.9cm,minimum height=0.7cm] (a2) at (0,0.75) {$B^{\T}$};1150\node[draw,minimum width=0.9cm,minimum height=0.7cm] (b2) at (0,-0.75) {$A$};1151\draw[->] (r.north) |- node[pos=0.25,left]{$x$} (a2.west);1152\draw[->] (a2.east) -| (c.north);1153\draw[->] (c.south) |- node[pos=0.25,right]{$y$} (b2.east);1154\draw[->] (b2.west) -| (r.south);1155\node at (0,-2.4) {two populations};1156\end{scope}1157\end{tikzpicture}1158\end{center}1159 1160\begin{remark}\label{rem:wiring}1161\Srf[\cite{SSV20,Mye21}] Open dynamical systems are the algebras of an operad of wiring diagrams: a diagram1162with boxes of given interfaces determines a closed system once each box is filled. A symmetric game is the1163one-box diagram with feedback through $A$; a bimatrix game is the two-box diagram; a local-interaction model1164(\Cref{ch:local}) has one box per vertex of a graph. Replacing the replicator box by a1165Smith box changes the dynamics and not the diagram. The statement that the source's models differ ``only in1166the dynamics'' or ``only in the interaction structure'' is the statement that a model is a pair of a diagram1167and a filling.1168\end{remark}1169 1170\section{Two populations}1171 1172For a bimatrix game $(A,B)$ the two-box system with replicator boxes is1173\[ \dot x_i=x_i\bigl((Ay)_i-x^{\T}Ay\bigr),\qquad\dot y_j=y_j\bigl((B^{\T}x)_j-y^{\T}B^{\T}x\bigr) \]1174on $\Dist(S)\times\Dist(T)$. The single-population equation for $A$ is its restriction to the diagonal in the case1175$B=A^{\T}$: the diagonal is the fixed set of the symmetry exchanging the two boxes.1176 1177\begin{theorem}\label{thm:selten}1178\Scl[\cite{Sel80,RiWe95,HoSi98}] In the two-population replicator dynamics a rest point is asymptotically stable if and1179only if it is a strict Nash equilibrium. In particular no mixed equilibrium is asymptotically stable.1180\end{theorem}1181 1182\noindent After a change of velocity the flow in the interior preserves volume, which excludes interior1183attractors. For the source's Battle of the Sexes, with payoffs $A=\begin{psmallmatrix}2&0\\0&1\end{psmallmatrix}$ to1184Row and $B=\begin{psmallmatrix}1&0\\0&2\end{psmallmatrix}$ to Column, the mixed equilibrium $(x,y)=(\tfrac23,\tfrac13)$1185is a saddle with eigenvalues $\pm\tfrac23$, and orbits started near it end at one of the two pure1186equilibria. \chk{O5} The population model selects conventions and never the mixed equilibrium, which is the1187content of the source's discussion \src{\S3.4}.1188 1189\section*{Notes}1190 1191Revision protocols and mean dynamics are developed systematically in \cite{San10}; proportional imitation1192is due to Schlag \cite{Sch98} and, in a different form, to Bj\"ornerstedt and Weibull \cite{BjWe96}. The1193Brown--von Neumann--Nash dynamic is from \cite{BvN50} and the Smith dynamic from \cite{Smi84}. Elimination of1194dominated strategies under imitative dynamics is in \cite{Nac90,SaZh92}. For open dynamical systems and1195wiring diagrams see \cite{SSV20,Mye21}; for games assembled compositionally from open pieces, \cite{GHWZ18}1196and \cite{GIHC}.