A slight variant $\tilde Mot_{num}(k,\mathbb{Q})$ of the category of pure motives $Mot_{num}(k,\mathbb{Q})$ is a Tannakian category equivalent to a category of representations of some algebraic group $GMot_k$ (the Motivic galois group):

$$ Mot_{num}(k,\mathbb{Q}) \simeq Rep(GMot_k) \,. $$


What it is known about the Derived category version (The derived category of motives) of this equivalence?

Equivalence between the sub-category of compact objects of (say for example) étale Motivic sheaves and the derived category of representations of finite dimension of (Ayoub) Motivic galois groups


1 Answer 1


Let $k$ be a field and $\operatorname{DM}_{gm}(k)_{\mathbb Q}$ the ∞-category of rational geometric motives over $k$. A mixed Weil cohomology theory induces a symmetric monoidal exact functor

$$ R: \operatorname{DM}_{gm}(k)_{\mathbb Q} \to D_c(\mathbb Q), $$

where $D_c(\mathbb Q)$ are the compact=dualizable objects in the derived ∞-category of $\mathbb Q$. For abstract reasons, there exists a best possible commutative Hopf algebra $H$ in $D(\mathbb Q)$ such that, if $G=\operatorname{Spec}(H)$, then $R$ lifts to

$$ \tilde R: \operatorname{DM}_{gm}(k)_{\mathbb Q} \to \operatorname{Rep}_{D_c(\mathbb Q)}(G). $$

This general construction can be found in this paper by Iwanari. As you know, Ayoub has also constructed $H$, but only in the homotopy category of $D(\mathbb Q)$, which is not enough to define $\tilde R$.

Let me now assume that $k\subset \mathbb C$ and that $R$ is the Betti realization. Ayoub then proves that $H$ is connective, so that $\tau_{\leq 0}H$ is a Hopf algebra in rational vector spaces. If $G^{cl}=\operatorname{Spec}(\tau_{\leq 0}H)$, we then have a closed immersion $i\colon G^{cl}\hookrightarrow G$, whence

$$ i^*\circ \tilde R: \operatorname{DM}_{gm}(k)_{\mathbb Q} \to \operatorname{Rep}_{D_c(\mathbb Q)}(G^{cl}). $$

Choudhury and Gallauer Alves de Souza proved that the group scheme $G^{cl}$ is isomorphic to Nori's motivic Galois group.

The strongest conjecture one can make here is that $i^*\circ\tilde R$ is an equivalence of ∞-categories. This conjecture is equivalent to the conjunction of two conjectures: $\tilde R$ is an equivalence, and $G^{cl}=G$. It also clearly implies the conservativity conjecture (that $R$ is conservative) and the existence of the motivic $t$-structure. Conversely, the existence of a $t$-structure such that $R$ is $t$-exact implies that $G^{cl}=G$. The conservativity of $R$ together with the existence of a suitable $t$-structure should imply that $\tilde R$ is an equivalence, but I cannot find an applicable reference at the moment.

  • 1
    $\begingroup$ The functor $\tilde{R}$ has been also constructed by Pridham in arxiv.org/abs/1309.0637. $\endgroup$
    – tttbase
    Dec 31, 2016 at 21:40
  • $\begingroup$ @MarcHoyois Could you help me understand, or point to a reference, what does "geometric" mean for a motive. I familiar with the general construction of taking nisnevich sheaves and localizing by $A^1$-equivalences and stabilizing then rationalizing. What do I need to do from this point to get "geometric motives? Persumably those should correspond to the numerical equivalence relation, how is it encoded in this language? $\endgroup$ Dec 13, 2018 at 6:55
  • 1
    $\begingroup$ @SaalHardali Geometric motives are the compact objects in the ∞-category of motives. $\endgroup$ Jan 19, 2019 at 7:40

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.