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 3501–3600 of 3949. Source text is displayed literally.

3501types. Equivariant contractibility of the whole simplex gives3502its quotient type. Each subspace requires a separate computation.3503 3504For a smooth equivariant field, the space of trajectories3505on a fixed time interval carries the action by transport3506of each trajectory. Initial and terminal evaluation maps3507are equivariant. In a well-posed deterministic system the3508initial evaluation is an equivariant homeomorphism to3509the initial state space, using continuous dependence on3510initial conditions. Terminal evaluation is then the flow3511map expressed through this identification. For a3512set-valued best response or a multivalued update (see3513Chapter~\ref{ch:revision}), the corresponding initial3514evaluation can have multiple trajectories over a state.3515In that situation its fibers encode a choice of resolution3516of ties or admissible continuations.3517 3518\section{Coherent actions in an \texorpdfstring{$\infty$}{infinity}-topos}3519 3520An internal group object in an $\infty$-topos3521$\mathcal E$ has a reduced bar groupoid $G_\bullet$ and a3522delooping $BG$. Effectivity identifies $G_\bullet$ with3523the \v{C}ech nerve of $1\to BG$. An action on $X$ can be3524specified by a cartesian transformation into3525$G_\bullet$ with degree-zero object $X$. Its realization3526is the map $X//G\to BG$. Conversely any object over $BG$3527pulls back along the \v{C}ech nerve to a cartesian3528transformation. Descent for slices shows that these3529assignments are inverse \cite[\S6.1.3]{lurie-htt,nss}.3530In particular,3531\[3532 \operatorname{Act}_G(\mathcal E)\simeq \mathcal E_{/BG}.3533\]3534The corresponding square3535\[3536 \begin{tikzcd}[column sep=large,row sep=large]3537 X\arrow[r]\arrow[d] & X//G\arrow[d]\\3538 1\arrow[r] & BG3539 \end{tikzcd}3540\]3541is a homotopy pullback. This describes quotient families3542while retaining their lifts and all compatibility3543conditions for change of context.3544 3545An evaluator $P\times X\to Y$ between equivariant objects3546is represented in this slice, so its local evaluations3547must transform both their parameters and inputs. For a3548discrete group acting on spaces, the space of equivariant3549maps from $X$ to $Y$ is the homotopy fixed-point space of3550$\operatorname{Map}(X,Y)$ with conjugation transport.3551For example, the cosimplicial description has in degree3552$n$ a family of maps for each string in $G^n$. Transporting3553the domain along the string identifies it with the usual3554cosimplicial diagram for the homotopy limit of the3555conjugation action. Totalization imposes naturality and3556its higher coherence. The assertion concerns the space3557of possible equivariant maps. The existence of a3558natural dynamical invariant, a distinguished equilibrium,3559or a numerical update requires its own mathematical3560conditions.3561 3562In these constructions the context can be an ordinary3563parameter space, a classifying groupoid, or an internal3564base for locally varying types. Equivariance and descent3565describe how locally constructed fitness and revision3566laws assemble. They do not determine a revision law3567without assumptions on its behavioral or statistical3568meaning. Weight updates provide a particularly direct3569case where those assumptions and the resulting equations3570can be specified together.3571\par3572\chapter{Learning rules and populations of agents}3573\label{ch:agent-learning}3574 3575Selection can act on candidate policies or programs as well3576as on inherited traits. In a learning system, three choices3577enter explicitly: what counts as a strategy, how its score3578is obtained, and how that score changes its frequency.3579The replicator field is an infinitesimal revision rule.3580A discrete-time learner can approximate it or use a3581different update with its own statistical guarantee.3582 3583\section{An update from relative entropy}3584 3585Let $x$ be a probability vector with positive coordinates3586on a finite set $I$. Given a payoff vector $a$ and a step3587$\eta>0$, exponential weights defines3588\[3589 x_i^+=\frac{x_i e^{\eta a_i}}{\sum_j x_j e^{\eta a_j}}.3590\]3591A common additive shift of the payoff cancels. The3592update uniquely maximizes on the simplex3593\[3594 y\longmapsto3595 \langle y,a\rangle-\eta^{-1}D(y\Vert x),3596 \qquad3597 D(y\Vert x)=\sum_i y_i\log(y_i/x_i),3598\]3599where a zero summand is interpreted as zero. Indeed,3600substitution of $x^+$ writes the objective as a constant