Library / Physical Thinking DPF
Jump to passage
In this reading

Link to current text

Published source confirmed at last check

Source changed 2026-10-03 08:25:59 UTC · snapshot created 2026-10-03 10:17:34 UTC · last check 2026-10-03 10:20:08 UTC

PHY.10:11 - SoTA-Echoing

Huan, Jagalur and Marzouk, Optimal experimental design: Formulations and computations, Acta Numerica (2024), with corrections and clarifications in the 2026 author revision, distinguishes design for parameter information, prediction and model discrimination. Its sections 2.2 and 6.1 retain uncertain influences and the possibility of model misspecification. These distinctions enter :4.1 and :4.4-.5. A design based on a supplied model set does not establish that the set contains an adequate account.

For the drag comparison, the disjoint force ranges give a sufficient design without a numerical optimizer. When feasible inputs are numerous and predicted records are uncertain, an experimental-design method can compare their expected contribution. Its advantage depends on a useful objective and credible response models. The cost of constructing and evaluating that optimization matters alongside the cost of the experiment.

Vezvaee et al., Fourier transform noise spectroscopy, npj Quantum Information (2024), develops noise reconstruction from free-induction and spin-echo measurements. The adopted physical contribution is changing how a prepared probe couples to temporal structure. The paper’s spectral reconstruction relies on its pure-dephasing and stationary Gaussian-noise setting; those assumptions are not imposed on the Lorentzian static-offset construction in :5.3.

For a noise question within that setting, the paper compares reconstruction from simple sequences with methods requiring many pulses. It also exposes different limitations: finite pulse timing, a restricted accessible frequency range and sensitivity of time derivatives to measurement error. Choosing more elaborate control is therefore not a general improvement. The simpler sequence can be sufficient for the needed distinction, while a different noise or control question calls for another method.

The three numerical cases are constructed comparisons under stated physical laws. Their purpose is to exhibit the changed preparation, response and interpretation; they are not reports of performed experiments. New physical effects or apparatus limits change the corresponding case premises.