A simplicial set is called finite if it has only a finite number of non degenerate simplicies (or equivalently, if its number of $n$-simplicies grows polynomially with $n$). In an answer to a previous question of mine Exponentiation in finite simplicial sets it was shown that for any $n\geq 4$ the simplicial set $(\Delta_n/\partial \Delta_n)^{\Delta_1}$ is not finite. Following one of the comments there, I wanted to ask:

Is the simplicial set $(\Delta_3/\partial \Delta_3)^{\Delta_1}$ finite?

  • $\begingroup$ What about (Δ^2/∂Δ^2)^{Δ^1}, is it finite or not? $\endgroup$ – Dmitri Pavlov Jul 22 '15 at 11:33
  • $\begingroup$ It is finite. See comments to the answer. $\endgroup$ – Ilan Barnea Jul 22 '15 at 20:59

No, it is not; here is a slightly simpler argument than the one indicated in my comment. As in the answer to the previous question, let $X=\Delta_3/\partial\Delta_3$. It suffices to show that the cardinality of $A_n=\operatorname{Hom}(\Delta_1\times \Delta_n, X)$ grows faster than any polynomial as a function of $n$. As in that answer, an element of $A_n$ can be identified with a sequence $(x_0,\dots,x_n)$ of elements of $X_{n+1}$ such that $d_{i+1}x_i=d_{i+1}x_{i+1}$ for $0\leq i<n$.

Let $B_n\subset A_n$ be the set of such sequences for which $d_{i+1}x_i=[\partial\Delta_3]$ is the basepoint of $X_n$ for all $0\leq i<n$ and $x_0=x_n=[\partial\Delta_3]$ is the basepoint of $X_{n+1}$. If $(x_0,\dots,x_n)\in B_n$, then for each $0<i<n$, $x_i$ must be either $[\partial\Delta_3]$ or the (image in $X$ of) the unique $(n+1)$-simplex $\sigma_i$ in $\Delta_3$ such that $\sigma_i$ is not in $\partial\Delta_3$ but $d_i\sigma_i$ and $d_{i+1}\sigma_i$ are in $\partial\Delta_3$. Explicitly, this simplex is the map $\sigma_i:[n+1]\to[3]$ such that $\sigma_i(j)=0$ for $j<i$, $\sigma_i(i)=1$, $\sigma_i(i+1)=2$, and $\sigma_i(j)=3$ for $j>i+1$.

So to define an element of $B_n$, we have exactly two choices for each $x_i$ for $0<i<n$, and every possible sequence of such choices does define an element of $B_n$. Thus $B_n$ has exactly $2^{n-1}$ elements. In particular, $|A_n|\geq 2^{n-1}$ grows exponentially.

  • 2
    $\begingroup$ Very nice! Thanks! What about $\Delta_2/\partial\Delta_2$? Your argument certainly doesn't apply then. $\endgroup$ – Ilan Barnea Jul 14 '15 at 16:38
  • $\begingroup$ It is not hard to show $(\Delta_2/\partial\Delta_2)^{\Delta_1}$ is finite (in fact, $4$-dimensional if I'm not mistaken). Just enumerate what a sequence $(x_0,\dots,x_n)$ can look like, and you'll only find $O(n^4)$ possibilities (this is a little messy but not hard, the point being that $x_i$ very nearly determines $x_{i+1}$). $\endgroup$ – Eric Wofsey Jul 14 '15 at 17:06
  • $\begingroup$ In fact, I think a similar argument can show that $X^K$ is finite whenever $\dim X\leq 2$ and $X$ and $K$ are finite, by induction on the number of nondegenerate simplices in $X$. The point is that given $(x_0,\dots, x_n)$ as above, the sequence $(x_0, d_1x_0, x_1, d_2x_1,\dots, x_n)$ can only exit and reenter the interior of any maximal face of $X$ a bounded number of times (as opposed to the situation above for $\Delta_3/\partial\Delta_3$, where all the $d_{i+1}x_i$ could be in $\partial\Delta_3$ while none of the $x_i$ are). $\endgroup$ – Eric Wofsey Jul 14 '15 at 17:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.