A probabilistic analysis of the difficulties of unifying quantum mechanics with the theory of relativity

A procedure is given for the transformation of quantum mechanical operator equations into stochastic equations. The stochastic equations reveal a simple correlation between quantum mechanics and classical mechanics: Quantum mechanics operates with “optimal estimations,” classical mechanics is the limit of “complete information.” In this connection, Schrödinger's substitution relationsp _x → -i? ?/?x, etc, reveal themselves as exact mathematical transformation formulas. The stochastic version of quantum mechanical equations provides an explanation for the difficulties in correlating quantum mechanics and the theory of relativity: In physics “time” is always thought of as a numerical parameter; but in the present formalism of physics “time” is described by two formally totally different quantities. One of these two “times” is a numerical parameter and the other a random variable. This last concept of time shows all the properties required by the theory of relativity and is therefore to be considered as the relativistic time.     

Celebrity Physicist: How the Press Sensationalized Einstein’s Search for a Unified Field Theory

In Einstein’s later years, from the late 1920s onward, his reputation in the physics community as an innovator had faded as he pursued increasingly unrealistic unified field theories. Yet from the perspective of the press, his image and ideas were still marketable. We will see how his various attempts to craft a unified field theory generated numerous headlines, despite their lack of experimental evidence or even realistic solutions. We will examine how Einstein’s “latest theory,” became a much sought-after commodity used to generate interest in books, magazines, and newspapers.     

Kink dynamics in the highly discrete sine-Gordon system

We study kink dynamics in a very discrete sine-Gordon system where the kink width is of the order of the lattice spacing. Numerical simulations exhibit new properties of kinks in this case: they lose the memory of their initial velocity and propagate preferentially at well-defined velocities which correspond to quasi-steady states, while a kink moving at other velocities suffers relatively high rates of radiation of small amplitude oscillations. When a small external driving force is applied to the system, the same velocities appear as plateus in the strongly nonlinear mobility of the kink. The energy radiated by the kink is calculated for a simple model that preserves the discrete character of the system, and the preferential velocities for the kink are obtained to good accuracy. Similar results may be expected to be valid for other discrete systems manifesting topological solitons. The numerical simulations reveal also new stable “multiple-kink” excitations which can propagate almost freely in extremely discrete systems where “ordinary” simple kinks are pinned to the lattice by discreteness. The stability of the “multiple-kinks” is discussed.     

Small-angle Coulomb collision model for particle-in-cell simulations

We construct and investigate a set of stochastic differential equations that incorporate the physics of velocity-dependent small-angle Coulomb collisions among the plasma particles in a particle-in-cell simulation. Each particle is scattered stochastically from all the other particles in a simulation cell modeled as one or more Maxwellians. Total energy and momentum are conserved by linear transformation of the velocity increments. In two test simulations the proposed “particle-moment” collision algorithm performs well with time steps as large as 10% of the relaxation time – far larger than a particle-pairing collision algorithm, in which pairs of particles are scattered from one another, requires to achieve the same accuracy.     

Evaluation of dynamical properties of open quantum systems using the driven Liouville-von Neumann approach: methodological considerations

Methodological aspects of using the driven Liouville-von Neumann (DLvN) approach for simulating dynamical properties of molecular junctions are discussed. As a model system we consider a non-interacting resonant level uniformly coupled to a single Fermionic bath. We demonstrate how a finite system can mimic the depopulation dynamics of the dot into an infinite band bath of continuous and uniform density of states. We further show how the effects of spurious energy resolved currents, appearing due to the approximate nature of the equilibrium state obtained in DLvN calculations, can be avoided. Several ways to approach the wide band limit, which is often adopted in analytical treatments, using a finite numerical model system are discussed including brute-force increase of the lead model bandwidth as well as efficient cancellation or direct subtraction of finite-bandwidth effect. These methodological considerations may be relevant also for other numerical schemes that aim to study non-equilibrium thermodynamics via simulations of open quantum systems.     

The Shelter Island Conferences Revisited: “Fundamental” Physics in the Decade 1975–1985

The focus of this broad historical overview of “the steady evolution of theoretical ideas” from Shelter Island I in 1947 to Shelter Island II in 1983 is some of the developments in “fundamental” physics after the establishment of the standard model, in particular, the adoption of the view that all present day field theories are “effective field theories” based on the gauge concept; taking seriously big bang cosmology, grand unified field theories (GUTs), and inflation; and the emergence of a new symbiosis of physics and mathematics.     

Why Does Deep and Cheap Learning Work So Well?

We show how the success of deep learning could depend not only on mathematics but also on physics: although well-known mathematical theorems guarantee that neural networks can approximate arbitrary functions well, the class of functions of practical interest can frequently be approximated through “cheap learning” with exponentially fewer parameters than generic ones. We explore how properties frequently encountered in physics such as symmetry, locality, compositionality, and polynomial log-probability translate into exceptionally simple neural networks. We further argue that when the statistical process generating the data is of a certain hierarchical form prevalent in physics and machine learning, a deep neural network can be more efficient than a shallow one. We formalize these claims using information theory and discuss the relation to the renormalization group. We prove various “no-flattening theorems” showing when efficient linear deep networks cannot be accurately approximated by shallow ones without efficiency loss; for example, we show that n variables cannot be multiplied using fewer than \(2^n\) neurons in a single hidden layer.