master-thesis/tex/talk/talk.org

2.8 KiB

These are the planning and notes for the talk for "Wissenschaftliches Arbeiten".

There is a total time of 20 minutes to fill, which is going to be a pretty tough upper bound.

  (defun total-time ()
    (interactive)
    (let ((total 0))
      (message
       (number-to-string
        (apply
         #'+
         (org-map-entries
          #'(lambda ()
              (string-to-number
               (or (cdr (assoc "DURATION" (org-entry-properties))) "0")))))))))
  • other seeds
  • only analyt. sol
  • no code comments :(: done, no we can :P
  • convergence, only one/two
  • other projects, less point but more pictures
  • "norm based"
  • outlook too much text

Introduction

DONE General Model   WAITING

  • hamiltonian multiple baths
  • (def. BCFs)

DONE Motivation

  • weak coupling/equilibrium well settled
  • strong coupling: no consensus (lots of citations)
  • won't be settled in a masters thesis but ->

    • access to bath/interaction observables with exact dynamics would be helpful
  • this where hopsflow will come in
  • also: nice maintainable tooling would be good

DONE NMQSD

  • nmqsd equation
  • not much more

DONE HOPS

  • hops for multiple baths
  • with new normalization

Bath Energy

DONE Formulas and Generalizations

  • short derivation
  • explain hurdles
  • explain nonzero

DONE Interaction Energy

  • straight forward -> remove dot

Demonstration one Bath

  • explain model
  • show some plots of the single qubit

Demonstration two Baths

  • explain model
  • show some plots of the double qubit
  • note that flow converges faster than system energy

Analytical Verification

Model

  • for one HO
  • give valentin credit

Solution

  • give valentin credit
  • general idea
  • formula for laplace transform of exponential
  • trick with mathematica

Comparison

  • one bath case for zero and nonzero temperature

Quick review of the two HO case

  • explain that generalization is straight forward
  • problems with numerics (sensitive to bcf)
  • show comparison

HOPS Related, No Details

  • code and docs and tests and poetry
  • auto norm
  • new cutoff
  • tree hops

Outlook

  • concrete models

    • sweep
    • three baths
  • steady state detection -> pseudo eigenstates?
  • higher order moments of bath energy

    • may bring insight into role of aux states
    • better cutoff for multi bath (not as many cross terms)
  • concrete importance sampling