MIT Open Access Articles
https://hdl.handle.net/1721.1/49433
Wed, 08 Apr 2020 01:36:34 GMT
20200408T01:36:34Z

A continuous analogue of the tensortrain decomposition
https://hdl.handle.net/1721.1/124519
A continuous analogue of the tensortrain decomposition
Gorodetsky, Alex; Karaman, Sertac; Marzouk, Youssef M.
We develop new approximation algorithms and data structures for representing and computing with multivariate functions using the functional tensortrain (FT), a continuous extension of the tensortrain (TT) decomposition. The FT represents functions using a tensortrain ansatz by replacing the threedimensional TT cores with univariate matrixvalued functions. The main contribution of this paper is a framework to compute the FT that employs adaptive approximations of univariate fibers, and that is not tied to any tensorized discretization. The algorithm can be coupled with any univariate linear or nonlinear approximation procedure. We demonstrate that this approach can generate multivariate function approximations that are several orders of magnitude more accurate, for the same cost, than those based on the conventional approach of compressing the coefficient tensor of a tensorproduct basis. Our approach is in the spirit of other continuous computation packages such as Chebfun, and yields an algorithm which requires the computation of “continuous” matrix factorizations such as the LU and QR decompositions of vectorvalued functions. To support these developments, we describe continuous versions of an approximate maximumvolume cross approximation algorithm and of a rounding algorithm that reapproximates an FT by one of lower ranks. We demonstrate that our technique improves accuracy and robustness, compared to TT and quanticsTT approaches with fixed parameterizations, of highdimensional integration, differentiation, and approximation of functions with local features such as discontinuities and other nonlinearities. ©2018
Sat, 01 Dec 2018 00:00:00 GMT
https://hdl.handle.net/1721.1/124519
20181201T00:00:00Z

Search for a new bottomonium state decaying to ϒ ( 1 S ) π + π − in pp collisions at √ s = 8 TeV
https://hdl.handle.net/1721.1/124518
Search for a new bottomonium state decaying to ϒ ( 1 S ) π + π − in pp collisions at √ s = 8 TeV
The results of a search for the bottomonium counterpart, denoted as X b , of the exotic charmonium state X(3872) is presented. The analysis is based on a sample of pp collisions at s=8 TeV collected by the CMS experiment at the LHC, corresponding to an integrated luminosity of 20.7 fb 1 . The search looks for the exclusive decay channel X b →Υ{hooked}(1S)π + π  followed by Υ{hooked}(1S)→μ + μ  . No evidence for an X b signal is observed. Upper limits are set at the 95% confidence level on the ratio of the inclusive production cross sections times the branching fractions to Υ{hooked}(1S)π + π  of the X b and the Υ{hooked}(2S). The upper limits on the ratio are in the range 0.95.4% for X b masses between 10 and 11 GeV. These are the first upper limits on the production of a possible X b at a hadron collider. © 2013 CERN.
Fri, 01 Nov 2013 00:00:00 GMT
https://hdl.handle.net/1721.1/124518
20131101T00:00:00Z

Roadmap on STIRAP applications
https://hdl.handle.net/1721.1/124517
Roadmap on STIRAP applications
Barnum, Timothy J.; Field, Robert W.
STIRAP (stimulated Raman adiabatic passage) is a powerful laserbased method, usually involving two photons, for efficient and selective transfer of populations between quantum states. A particularly interesting feature is the fact that the coupling between the initial and the final quantum states is via an intermediate state, even though the lifetime of the latter can be much shorter than the interaction time with the laser radiation. Nevertheless, spontaneous emission from the intermediate state is prevented by quantum interference. Maintaining the coherence between the initial and final state throughout the transfer process is crucial. STIRAP was initially developed with applications in chemical dynamics in mind. That is why the original paper of 1990 was published in The Journal of Chemical Physics. However, from about the year 2000, the unique capabilities of STIRAP and its robustness with respect to small variations in some experimental parameters stimulated many researchers to apply the scheme to a variety of other fields of physics. The successes of these efforts are documented in this collection of articles. In Part A the experimental success of STIRAP in manipulating or controlling molecules, photons, ions or even quantum systems in a solidstate environment is documented. After a brief introduction to the basic physics of STIRAP, the central role of the method in the formation of ultracold molecules is discussed, followed by a presentation of how precision experiments (measurement of the upper limit of the electric dipole moment of the electron or detecting the consequences of parity violation in chiral molecules) or chemical dynamics studies at ultralow temperatures benefit from STIRAP. Next comes the STIRAPbased control of photons in cavities followed by a group of three contributions which highlight the potential of the STIRAP concept in classical physics by presenting data on the transfer of waves (photonic, magnonic and phononic) between respective waveguides. The works on ions or ion strings discuss options for applications, e.g. in quantum information. Finally, the success of STIRAP in the controlled manipulation of quantum states in solidstate systems, which are usually hostile towards coherent processes, is presented, dealing with data storage in rareearth ion doped crystals and in nitrogen vacancy (NV) centers or even in superconducting quantum circuits. The works on ions and those involving solidstate systems emphasize the relevance of the results for quantum information protocols. Part B deals with theoretical work, including further concepts relevant to quantum information or invoking STIRAP for the manipulation of matter waves. The subsequent articles discuss the experiments underway to demonstrate the potential of STIRAP for populating otherwise inaccessible highlying Rydberg states of molecules, or controlling and cooling the translational motion of particles in a molecular beam or the polarization of angularmomentum states. The series of articles concludes with a more speculative application of STIRAP in nuclear physics, which, if suitable radiation fields become available, could lead to spectacular results. ©2019
Sun, 01 Sep 2019 00:00:00 GMT
https://hdl.handle.net/1721.1/124517
20190901T00:00:00Z

3Daware scene manipulation via inverse graphics
https://hdl.handle.net/1721.1/124516
3Daware scene manipulation via inverse graphics
Yao, Shunyu; Hsu, Tzu Ming; Zhu, JunYan; Wu, Jiajun; Torralba, Antonio; Freeman, William T.; Tenenbaum, Joshua B.
We aim to obtain an interpretable, expressive, and disentangled scene representation that contains comprehensive structural and textural information for each object. Previous scene representations learned by neural networks are often uninterpretable, limited to a single object, or lacking 3D knowledge. In this work, we propose 3D scene derendering networks (3DSDN) to address the above issues by integrating disentangled representations for semantics, geometry, and appearance into a deep generative model. Our scene encoder performs inverse graphics, translating a scene into a structured objectwise representation. Our decoder has two components: a differentiable shape renderer and a neural texture generator. The disentanglement of semantics, geometry, and appearance supports 3Daware scene manipulation, e.g., rotating and moving objects freely while keeping the consistent shape and texture, and changing the object appearance without affecting its shape. Experiments demonstrate that our editing scheme based on 3DSDN is superior to its 2D counterpart. ©2018 Poster presentation at the 32nd annual Conference on Neural Information Processing Systems (NIPS 2018), December 35, 2018, Montréal, Québec.
Mon, 01 Jan 2018 00:00:00 GMT
https://hdl.handle.net/1721.1/124516
20180101T00:00:00Z