-
Controlling Floquet states on ultrashort time scales
Authors:
Matteo Lucchini,
Fabio Medeghini,
Yingxuan Wu,
Federico Vismarra,
Rocío Borrego-Varillas,
Aurora Crego,
Fabio Frassetto,
Luca Poletto,
Shunsuke A. Sato,
Hannes Hübener,
Umberto De Giovannini,
Ángel Rubio,
Mauro Nisoli
Abstract:
The advent of ultrafast laser science offers the unique opportunity to combine Floquet engineering with extreme time resolution, further pushing the optical control of matter into the petahertz domain. However, what is the shortest driving pulse for which Floquet states can be realised remains an unsolved matter, thus limiting the application of Floquet theory to pulses composed by many optical cy…
▽ More
The advent of ultrafast laser science offers the unique opportunity to combine Floquet engineering with extreme time resolution, further pushing the optical control of matter into the petahertz domain. However, what is the shortest driving pulse for which Floquet states can be realised remains an unsolved matter, thus limiting the application of Floquet theory to pulses composed by many optical cycles. Here we ionized Ne atoms with few-femtosecond pulses of selected time duration and show that a Floquet state can be established already within 10 cycles of the driving field. For shorter pulses, down to 2 cycles, the finite lifetime of the driven state can still be explained using an analytical model based on Floquet theory. By demonstrating that the population of the Floquet sidebands can be controlled not only with the driving laser pulse intensity and frequency, but also by its duration, our results add a new lever to the toolbox of Floquet engineering.
△ Less
Submitted 2 May, 2022;
originally announced May 2022.
-
Won't you see my neighbor?: User predictions, mental models, and similarity-based explanations of AI classifiers
Authors:
Kimberly Glasgow,
Jonathan Kopecky,
John Gersh,
Adam Crego
Abstract:
Humans should be able work more effectively with artificial intelligence-based systems when they can predict likely failures and form useful mental models of how the systems work. We conducted a study of human's mental models of artificial intelligence systems using a high-performing image classifier, focusing on participants' ability to predict the classification result for a particular image. Pa…
▽ More
Humans should be able work more effectively with artificial intelligence-based systems when they can predict likely failures and form useful mental models of how the systems work. We conducted a study of human's mental models of artificial intelligence systems using a high-performing image classifier, focusing on participants' ability to predict the classification result for a particular image. Participants viewed individual labeled images in one of two classes and then tried to predict whether the classifier would label them correctly. In this experiment we explored the effect of giving participants additional information about an image's nearest neighbors in a space representing the otherwise uninterpretable features extracted by the lower layers of the classifier's neural network. We found that providing this information did increase participants' prediction performance, and that the performance improvement could be related to the neighbor images' similarity to the target image. We also found indications that the presentation of this information may influence people's own classification of the target image -- that is, rather than just anthropomorphizing the system, in some cases the humans become "mechanomorphized" in their judgements.
△ Less
Submitted 31 January, 2022;
originally announced January 2022.
-
Laser-based X-ray and electron source for X-ray fluorescence studies
Authors:
F. Valle Brozas,
A. Crego,
L. Roso,
A. Peralta Conde
Abstract:
In this work we present a modification to conventional X-rays fluorescence using electrons as excitation source, and compare it with the traditional X-ray excitation for the study of pigments. For this purpose we have constructed a laser-based source capable to produce X-rays as well as electrons. Because of the large penetration depth of X-rays, the collected fluorescence signal is a combination…
▽ More
In this work we present a modification to conventional X-rays fluorescence using electrons as excitation source, and compare it with the traditional X-ray excitation for the study of pigments. For this purpose we have constructed a laser-based source capable to produce X-rays as well as electrons. Because of the large penetration depth of X-rays, the collected fluorescence signal is a combination of several material layers of the artwork under study. However electrons are stopped in the first layers allowing therefore a more superficial analysis. We show that the combination of both excitation sources can provide extremely valuable information about the structure of the artwork.
△ Less
Submitted 13 May, 2016;
originally announced May 2016.
