update: Here is the abstract for the third talk, by Robin Blume-Kohout.
------
A gauge symmetry appears when we can transform the parameters of a
theory, yet leave every observable property invariant. This means
that the theory contains redundant parameters. The theory of quantum
information processing (QIP) -- in which elementary operations are
completely positive maps, and any viable algorithm is described by a
quantum circuit that specifies a sequence of those operations -- has
a gauge. For "forward" QIP this is no big deal; it just provides
multiple equivalent ways to describe the same physical device. But
for "inverse" QIP -- i.e., tomography and characterization -- this is
surprisingly inconvenient! I will discuss our recent progress on
device characterization via gate-set tomography, chronicle our ongoing
struggles with the gauge, and plead for your assistance in taming it.
On Mon, May 19, 2014 at 11:44 AM, aram harrow <aram(a)mit.edu> wrote:
Dear quanta,
We have an exciting roster of speakers and visitors coming up this week.
Visitors:
Steve Flammia (Sydney): May 20-27
Robin Blume-Kohout (Sandia) : May 22-23
Nicole Yunger Halpern (Caltech): May 22
If you want to meet any of them, either email me, or email them
directly. Their office locations are on the blackboard at the
entrance of the CTP.
Talks, each in 6c-442:
Wed, 11am: Kevin Zatloukal (MIT), Subset sum and the hidden subgroup
problem for nilpotent groups
Thurs, 4pm: Nicole Yunger Halpern, Beyond heat baths: Resource
theories for small-scale thermodynamics
Fri, 1:30pm: Robin Blume-Kohout, Quantum information is
(inconveniently!) a gauge theory
There is (inconveniently!) no abstract for the last talk. Abstracts
for the first two talks are:
1. Zatloukal - HSP
In this talk, we demonstrate that there exists an efficient quantum
algorithm for solving the hidden subgroup problem (HSP) in k-nilpotent
groups if we are provided with an oracle for solving average-case
subset sum. The same fact was already known for the dihedral group. In
fact, solving HSP for the dihedral group is thought to be equivalent
to solving average-case subset sum. Thus, our result can be
interpreted as showing that the HSP for any k-nilpotent group is no
harder than the HSP for the dihedral group.
Using classical algorithms for solving subset sum (even just the
brute-force algorithm), we are able to get a non-trivial speedup for
all k-nilpotent groups. In particular, for 2-groups, the subset sum
problem is easy, so we get efficient quantum algorithms for
k-nilpotent 2-groups. This is an especially interesting case because
the dihedral group can be a 2-group as well, albeit one of high
nilpotency class.
2. Halpern - thermodynamics
How can thermodynamics, which involves vast numbers of particles,
describe cutting-edge technology, which involves tiny scales? Heat
exchanges have been modeled with the resource-theory framework famous
for elucidating entanglement. “Helmholtz” resource theories have shown
that the work cost W_cost of creating one copy of a state can differ
from the work W_dist distillable from the state. This “one-shot”
result contrasts with the thermodynamic limit, in which W_cost =
W_dist. I will generalize the resource-theory framework from heat
exchanges to more-arbitrary thermodynamic interactions. I will
introduce resource theories of non-equilibrium, a family of models
that includes the Helmholtz theories. In addition to shedding new
light on Helmholtz theories, the non-equilibrium framework expands the
resource theories’ potential to describe experiments and quantum
phenomena. I will quantify the distinction between W_cost and W_dist
in terms of the hypothesis-testing entropy D_H. This research is
motivated by the interplay between information and energy on small
scales exemplified by single-molecule experiments, nanoscale engines,
and molecular motors.
This work was conducted partially (arXiv:1309.6586) at the Perimeter
Institute for Theoretical physics with Markus Müller and Rob Spekkens,
as well as with Gilad Gour and Varun Narasimhachar. Part of this work
is being conducted with Joe Renes.
_______________________________________________
qip mailing list
qip(a)mit.edu
http://mailman.mit.edu/mailman/listinfo/qip