The two-eigenvalue problem and density of Jones representation of braid groups

  • Michael Freedman ,
  • Michael J. Larsen ,
  • Wang Zhenghan

MSR-TR-2001-42 |

In 1983 V. Jones discovered a new family of representations ρ of the
braid groups. They emerged from the study of operator algebras (type Π1
factors) and unlike earlier braid representations had no naive homological
interpretation. Almost immediately he found that the trace or “Markov”
property of ρ allowed new link invariants to be defined and this ushered in
the era of quantum topology. There has been an explosion of link and 3-
manifold invariants with beautiful inter-relations, asymptotic formulae, and
enchanting connections to mathematical physics: Chern-Simons theory and
2-dimensional statistical mechanics. While many sought to bend Jones’ theory
toward classical topological objectives, we have found that the relation
between the Jones polynomial and physics allows potentially realistic models
of quantum computation to be created [FKW][FLW][FKLW][F]. Unitarity, a
hidden locality, and density of the Jones representation are central to computational
applications. With this application in mind, we have returned
to some of Jones’ earliest questions about these representations and the distributions
of his invariants. A few concise answers are stated here in the
introduction. Question 9 of Jones in [J2] asked for the closed images of the
irreducible components of his representation. We answer Jones’ question,
and also identified the closed images for the general SU(N) case completely.