Bell Labs

About: Bell Labs is a based out in . It is known for research contribution in the topics: Laser & Optical fiber. The organization has 36499 authors who have published 59862 publications receiving 3190823 citations. The organization is also known as: Bell Laboratories & AT&T Bell Laboratories.

Transcription against an applied force

Abstract: The force produced by a single molecule of Escherichia coli RNA polymerase during transcription was measured optically. Polymerase immobilized on a surface was used to transcribe a DNA template attached to a polystyrene bead 0.5 micrometer in diameter. The bead position was measured by interferometry while a force opposing translocation of the polymerase along the DNA was applied with an optical trap. At saturating nucleoside triphosphate concentrations, polymerase molecules stalled reversibly at a mean applied force estimated to be 14 piconewtons. This force is substantially larger than those measured for the cytoskeletal motors kinesin and myosin and exceeds mechanical loads that are estimated to oppose transcriptional elongation in vivo. The data are consistent with efficient conversion of the free energy liberated by RNA synthesis into mechanical work.

Wave Functions and Effective Hamiltonian for Bloch Electrons in an Electric Field

Abstract: Following up an earlier communication, wave functions are constructed which satisfy the Schr\"odinger equation for a potential which is a sum of a periodic and a uniform field term. The wave functions are Houston modifications of Bloch type functions; the Bloch functions form an orthogonal set whose members are fully determined except for phase. The theory exhibits them in the form of power series in the field strength; the unmodified Bloch band functions form the zero order term of that series. The solutions themselves do not allow for a Zener effect, but the fact that they are only given as power series in $E$ may imply that there is a remainder term causing interband transitions; it would have to be asymptotically smaller than any power of $E$. Instead of constructing time dependent solutions of the Schr\"odinger equation one can take the time independent functions to construct an effective Hamiltonian for electrons in one band; it has the form (16). Certain indeterminacies are attached to this form of representation; it is shown, however, that final physical answers are unique. The study furnishes an incidental proof that k-space is a finite space consisting in its entirety of what is customarily called the first Brillouin zone. An appendix treats the case of degenerate bands; such bands have singularities in k-space even in the absence of a field. The difficulty is circumvented by working with a set which is not yet diagonalized but free of singularities; these intermediate functions can be continued as power series in $E$ in the same way as nondegenerate band functions.

One sense per discourse

Abstract: It is well-known that there are polysemous words like sentence whose "meaning" or "sense" depends on the context of use. We have recently reported on two new word-sense disambiguation systems, one trained on bilingual material (the Canadian Hansards) and the other trained on monolingual material (Roget's Thesaurus and Grolier's Encyclopedia). As this work was nearing completion, we observed a very strong discourse effect. That is, if a polysemous word such as sentence appears two or more times in a well-written discourse, it is extremely likely that they will all share the same sense. This paper describes an experiment which confirmed this hypothesis and found that the tendency to share sense in the same discourse is extremely strong (98%). This result can be used as an additional source of constraint for improving the performance of the word-sense disambiguation algorithm. In addition, it could also be used to help evaluate disambiguation algorithms that did not make use of the discourse constraint.

Structure in aqueous solutions of nonpolar solutes from the standpoint of scaled-particle theory

Abstract: Underlying assumptions have been examined in scaled-particle theory for the case of a rigid-sphere solute in liquid water. As a result, it has been possible to improve upon Pierotti’s corresponding analysis in a way that explicitly incorporates measured surface tensions and radial-distribution functions for pure water. It is pointed out along the way that potential energy nonadditivity should create an orientational bias for molecules in the liquid-vapor interface that is peculiar to water. Some specific conclusions have been drawn about the solvation mode for the nonpolar rigid-sphere solute.

Stochastic Gradient Learning in Neural Networks

Abstract: Many connectionist learning algorithms consists of minimizing a cost of the form C(w) = E(J(z,w)) = J(z,w)dP(z) where dP is an unknown probability distribution that characterizes the problem to learn, and J, the loss function, defines the learning system itself. This popular statistical formulation has led to many theoretical results. The minimization of such a cost may be achieved with a stochastic gradient descent algorithm, e.g.: wt+1 = wt − ɛt∇wJ(z,wt) With some restrictions on J and C, this algorithm converges, even if J is non differentiable on a set of measure 0. Links with simulated annealing are depicted.