You can find the details of Viola's presentation here:

Worlds are not denotations
The main goal of this talk is negative: I will argue that possible worlds don’t have object-language representations, as (i) readings that have been argued to require such object-language representations are more plausibly derived without them (this is based on current joint work with Ido Benbaji  and Clemens Mayr) and (ii)`ontological symmetry’ (as discussed in Schlenker 2006) breaks down when considering plurality. Beyond providing evidence for the negative claim, I will argue that these data raise questions for our conception of how language makes use of possible worlds, even when considered as parameters of the interpretation function.