http://mw.hh.se/wg211/index.php?action=history&feed=atom&title=WG211%2FM7NameWG211/M7Name - Revision history2026-04-05T23:01:07ZRevision history for this page on the wikiMediaWiki 1.43.5http://mw.hh.se/wg211/index.php?title=WG211/M7Name&diff=235&oldid=prevAdmin: 1 revision2011-12-12T10:06:27Z<p>1 revision</p>
<p><b>New page</b></p><div>[[Category:WG211]]<br />
<br />
<br />
===Theory of mind and bounded rationality without interpretive overhead===<br />
Oleg Kiselyov and Chung-chieh Shan<br />
<br />
Computers and humans that work well together have beliefs about each<br />
other's intentions, about each other's desires about each other's<br />
beliefs, and so on. To practise such a _theory of mind_, agents need<br />
to slip easily into each other's shoes. Ideally, when Agent A reasons<br />
about Agent B with complete certainty, Agent A should simulate Agent B's<br />
mind as efficiently as if that simulation were reality. Modeling agents<br />
as programs, we want Agent A to interpret Agent B's program _without<br />
interpretive overhead_, that is, as efficiently as if that program ran<br />
directly. A programming language with _delimited control operators_<br />
lets us eliminate interpretive overhead in a computational model of<br />
bounded-rational agents that reason about each other probabilistically.<br />
The key is to reify stochastic programs as probability distributions<br />
using the increasingly popular _finally tagless_ technique for embedding<br />
programming languages. We demonstrate the idea with a simplistic model<br />
of plausibly deniable bribing.</div>Admin