-
The Autodidactic Universe
Authors:
Stephon Alexander,
William J. Cunningham,
Jaron Lanier,
Lee Smolin,
Stefan Stanojevic,
Michael W. Toomey,
Dave Wecker
Abstract:
We present an approach to cosmology in which the Universe learns its own physical laws. It does so by exploring a landscape of possible laws, which we express as a certain class of matrix models. We discover maps that put each of these matrix models in correspondence with both a gauge/gravity theory and a mathematical model of a learning machine, such as a deep recurrent, cyclic neural network. Th…
▽ More
We present an approach to cosmology in which the Universe learns its own physical laws. It does so by exploring a landscape of possible laws, which we express as a certain class of matrix models. We discover maps that put each of these matrix models in correspondence with both a gauge/gravity theory and a mathematical model of a learning machine, such as a deep recurrent, cyclic neural network. This establishes a correspondence between each solution of the physical theory and a run of a neural network. This correspondence is not an equivalence, partly because gauge theories emerge from $N \rightarrow \infty $ limits of the matrix models, whereas the same limits of the neural networks used here are not well-defined. We discuss in detail what it means to say that learning takes place in autodidactic systems, where there is no supervision. We propose that if the neural network model can be said to learn without supervision, the same can be said for the corresponding physical theory. We consider other protocols for autodidactic physical systems, such as optimization of graph variety, subset-replication using self-attention and look-ahead, geometrogenesis guided by reinforcement learning, structural learning using renormalization group techniques, and extensions. These protocols together provide a number of directions in which to explore the origin of physical laws based on putting machine learning architectures in correspondence with physical theories.
△ Less
Submitted 2 September, 2021; v1 submitted 28 March, 2021;
originally announced April 2021.
-
Ollivier-Ricci curvature convergence in random geometric graphs
Authors:
Pim van der Hoorn,
William J. Cunningham,
Gabor Lippner,
Carlo Trugenberger,
Dmitri Krioukov
Abstract:
Connections between continuous and discrete worlds tend to be elusive. One example is curvature. Even though there exist numerous nonequivalent definitions of graph curvature, none is known to converge in any limit to any traditional definition of curvature of a Riemannian manifold. Here we show that Ollivier curvature of random geometric graphs in any Riemannian manifold converges in the continuu…
▽ More
Connections between continuous and discrete worlds tend to be elusive. One example is curvature. Even though there exist numerous nonequivalent definitions of graph curvature, none is known to converge in any limit to any traditional definition of curvature of a Riemannian manifold. Here we show that Ollivier curvature of random geometric graphs in any Riemannian manifold converges in the continuum limit to Ricci curvature of the underlying manifold, but only if the definition of Ollivier graph curvature is properly generalized to apply to mesoscopic graph neighborhoods. This result establishes the first rigorous link between a definition of curvature applicable to networks and a traditional definition of curvature of smooth spaces.
△ Less
Submitted 9 March, 2021; v1 submitted 3 August, 2020;
originally announced August 2020.
-
Tensor network renormalization with fusion charges: applications to 3d lattice gauge theory
Authors:
William J. Cunningham,
Bianca Dittrich,
Sebastian Steinhaus
Abstract:
Tensor network methods are powerful and efficient tools to study the properties and dynamics of statistical and quantum systems, in particular in one and two dimensions. In recent years, these methods were applied to lattice gauge theories, yet these theories remain a challenge in $(2+1)$ dimensions. In this article, we present a new (decorated) tensor network algorithm, in which the tensors encod…
▽ More
Tensor network methods are powerful and efficient tools to study the properties and dynamics of statistical and quantum systems, in particular in one and two dimensions. In recent years, these methods were applied to lattice gauge theories, yet these theories remain a challenge in $(2+1)$ dimensions. In this article, we present a new (decorated) tensor network algorithm, in which the tensors encode the lattice gauge amplitude expressed in the fusion basis.
This has several advantages: Firstly, the fusion basis does diagonalize operators measuring the magnetic fluxes and electric charges associated to a hierarchical set of regions. The algorithm allows therefore a direct access to these observables. Secondly the fusion basis is, as opposed to the previously employed spin network basis, stable under coarse graining. Thirdly, due to the hierarchical structure of the fusion basis, the algorithm does implement predefined disentangles, that remove short-scale entanglement.
We apply this new algorithm to lattice gauge theories defined for the quantum group $\text{SU}(2)_{\rm k}$ and identify a weak and a strong coupling phase for various levels $\rm k$. As we increase the level $\rm k$, the critical coupling $g_c$ decreases linearly, suggesting the absence of a deconfining phase for the continuous group $\text{SU}(2)$. Moreover, we illustrate the scaling behaviour of the Wilson loops in the two phases.
△ Less
Submitted 24 February, 2020;
originally announced February 2020.
-
Dimensionally Restricted Causal Set Quantum Gravity: Examples in Two and Three Dimensions
Authors:
William J. Cunningham,
Sumati Surya
Abstract:
We study dimensionally restricted non-perturbative causal set quantum dynamics in $2$ and $3$ spacetime
dimensions with non-trivial global spatial topology. The causal set sample space is generated from causal
embeddings into spacetime lattices with global spatial topology $S^1$ and $T^2$ in $2$ and $3$
dimensions, respectively. The quantum gravity partition function over these sample spaces…
▽ More
We study dimensionally restricted non-perturbative causal set quantum dynamics in $2$ and $3$ spacetime
dimensions with non-trivial global spatial topology. The causal set sample space is generated from causal
embeddings into spacetime lattices with global spatial topology $S^1$ and $T^2$ in $2$ and $3$
dimensions, respectively. The quantum gravity partition function over these sample spaces is studied using Markov
Chain Monte Carlo (MCMC) simulations after analytic
continuation. In both $2$ and $3$ dimensions we find a
phase transition that separates the dominance of the action from that of the entropy. The action dominated phase is
characterised by ``layered'' posets with a high degree of connectivity, while the causal sets in the entropy
dominated phase are manifold-like. This phase transition is similar in character to that seen for the sample space of $2$-orders,
which are topologically trivial, hence suggesting that this is a generic feature of dimensionally restricted sample
spaces. The simulations use a newly developed framework for causal set MCMC calculations. Ours is
the first implementation of a causal set dynamics restricted to $3$ dimensions.
△ Less
Submitted 3 December, 2019; v1 submitted 30 August, 2019;
originally announced August 2019.
-
High Performance Algorithms for Quantum Gravity and Cosmology
Authors:
William J. Cunningham
Abstract:
Large scale numerical experiments are commonplace today in theoretical physics. The high performance algorithms described herein are the most compact, efficient methods known for representing and analyzing systems modeled well by sets or graphs. After studying how these implementations maximize instruction throughput and optimize memory access patterns, we apply them to causal set quantum gravity,…
▽ More
Large scale numerical experiments are commonplace today in theoretical physics. The high performance algorithms described herein are the most compact, efficient methods known for representing and analyzing systems modeled well by sets or graphs. After studying how these implementations maximize instruction throughput and optimize memory access patterns, we apply them to causal set quantum gravity, in which spacetime is represented by a partially ordered set. We build upon the low-level set and graph algorithms to optimize the calculation of the causal set action, and then discuss how to measure boundaries of a discrete spacetime. We then examine the broader applicability of these algorithms to greedy information routing in random geometric graphs embedded in Lorentzian manifolds, which requires us to find new closed-form solutions to the geodesic differential equations in Friedmann-Lemaitre-Robertson-Walker spacetimes. Finally, we consider the vacuum selection problem in string theory, where we show a network-centered approach yields a dynamical mechanism for vacuum selection in the context of multiverse cosmology. These algorithms have broad applicability to many physical systems, and they improve existing methods by reducing simulation runtimes by orders of magnitude.
△ Less
Submitted 11 May, 2018;
originally announced May 2018.
-
Vacuum Selection from Cosmology on Networks of String Geometries
Authors:
Jonathan Carifio,
William J. Cunningham,
James Halverson,
Dmitri Krioukov,
Cody Long,
Brent D. Nelson
Abstract:
We introduce network science as a framework for studying the string landscape. Two large networks of string geometries are constructed, where nodes are extra-dimensional six-manifolds and edges represent topological transitions between them. We show that a standard bubble cosmology model on the networks has late-time behavior determined by the largest eigenvector of $-(\mathbf{L} + \mathbf{D})$, w…
▽ More
We introduce network science as a framework for studying the string landscape. Two large networks of string geometries are constructed, where nodes are extra-dimensional six-manifolds and edges represent topological transitions between them. We show that a standard bubble cosmology model on the networks has late-time behavior determined by the largest eigenvector of $-(\mathbf{L} + \mathbf{D})$, where $\mathbf{L}$ and $\mathbf{D}$ are the Laplacian and degree matrices of the networks, which provides a dynamical mechanism for vacuum selection in the string landscape.
△ Less
Submitted 5 September, 2018; v1 submitted 17 November, 2017;
originally announced November 2017.
-
Inference of Boundaries in Causal Sets
Authors:
William J. Cunningham
Abstract:
We investigate the extrinsic geometry of causal sets in $(1+1)$-dimensional Minkowski spacetime. The properties of boundaries in an embedding space can be used not only to measure observables, but also to supplement the discrete action in the partition function via discretized Gibbons-Hawking-York boundary terms. We define several ways to represent a causal set using overlap** subsets, which the…
▽ More
We investigate the extrinsic geometry of causal sets in $(1+1)$-dimensional Minkowski spacetime. The properties of boundaries in an embedding space can be used not only to measure observables, but also to supplement the discrete action in the partition function via discretized Gibbons-Hawking-York boundary terms. We define several ways to represent a causal set using overlap** subsets, which then allows us to distinguish between null and non-null bounding hypersurfaces in an embedding space. We discuss algorithms to differentiate between different types of regions, consider when these distinctions are possible, and then apply the algorithms to several spacetime regions. Numerical results indicate the volumes of timelike boundaries can be measured to within $0.5\%$ accuracy for flat boundaries and within $10\%$ accuracy for highly curved boundaries for medium-sized causal sets with $N=2^{14}$ spacetime elements.
△ Less
Submitted 26 June, 2018; v1 submitted 26 October, 2017;
originally announced October 2017.
-
Causal Set Generator and Action Computer
Authors:
William J. Cunningham,
Dmitri Krioukov
Abstract:
The causal set approach to quantum gravity has gained traction over the past three decades, but numerical experiments involving causal sets have been limited to relatively small scales. The software suite presented here provides a new framework for the generation and study of causal sets. Its efficiency surpasses previous implementations by several orders of magnitude. We highlight several importa…
▽ More
The causal set approach to quantum gravity has gained traction over the past three decades, but numerical experiments involving causal sets have been limited to relatively small scales. The software suite presented here provides a new framework for the generation and study of causal sets. Its efficiency surpasses previous implementations by several orders of magnitude. We highlight several important features of the code, including the compact data structures, the $O(N^2)$ causal set generation process, and several implementations of the $O(N^3)$ algorithm to compute the Benincasa-Dowker action of compact regions of spacetime. We show that by tailoring the data structures and algorithms to take advantage of low-level CPU and GPU architecture designs, we are able to increase the efficiency and reduce the amount of required memory significantly. The presented algorithms and their implementations rely on methods that use CUDA, OpenMP, x86 Assembly, SSE/AVX, Pthreads, and MPI. We also analyze the scaling of the algorithms' running times with respect to the problem size and available resources, with suggestions on how to modify the code for future hardware architectures.
△ Less
Submitted 13 September, 2018; v1 submitted 9 September, 2017;
originally announced September 2017.
-
Exact Geodesic Distances in FLRW Spacetimes
Authors:
William J. Cunningham,
David Rideout,
James Halverson,
Dmitri Krioukov
Abstract:
Geodesics are used in a wide array of applications in cosmology and astrophysics. However, it is not a trivial task to efficiently calculate exact geodesic distances in an arbitrary spacetime. We show that in spatially flat (3+1)-dimensional Friedmann-Lemaitre-Robertson-Walker (FLRW) spacetimes, it is possible to integrate the second-order geodesic differential equations, and derive a general meth…
▽ More
Geodesics are used in a wide array of applications in cosmology and astrophysics. However, it is not a trivial task to efficiently calculate exact geodesic distances in an arbitrary spacetime. We show that in spatially flat (3+1)-dimensional Friedmann-Lemaitre-Robertson-Walker (FLRW) spacetimes, it is possible to integrate the second-order geodesic differential equations, and derive a general method for finding both timelike and spacelike distances given initial-value or boundary-value constraints. In flat spacetimes with either dark energy or matter, whether dust, radiation, or a stiff fluid, we find an exact closed-form solution for geodesic distances. In spacetimes with a mixture of dark energy and matter, including spacetimes used to model our physical universe, there exists no closed-form solution, but we provide a fast numerical method to compute geodesics. A general method is also described for determining the geodesic connectedness of an FLRW manifold, provided only its scale factor.
△ Less
Submitted 27 November, 2017; v1 submitted 1 May, 2017;
originally announced May 2017.
