@article{63, abstract = {African cichlids display a remarkable assortment of jaw morphologies, pigmentation patterns, and mating behaviors. In addition to this previously documented diversity, recent studies have documented a rich diversity of sex chromosomes within these fishes. Here we review the known sex-determination network within vertebrates, and the extraordinary number of sex chromosomes systems segregating in African cichlids. We also propose a model for understanding the unusual number of sex chromosome systems within this clade.}, author = {Gammerdinger, William J and Kocher, Thomas}, journal = {Genes}, number = {10}, publisher = {MDPI AG}, title = {{Unusual diversity of sex chromosomes in African cichlid fishes}}, doi = {10.3390/genes9100480}, volume = {9}, year = {2018}, } @article{296, abstract = {The thermodynamic description of many-particle systems rests on the assumption of ergodicity, the ability of a system to explore all allowed configurations in the phase space. Recent studies on many-body localization have revealed the existence of systems that strongly violate ergodicity in the presence of quenched disorder. Here, we demonstrate that ergodicity can be weakly broken by a different mechanism, arising from the presence of special eigenstates in the many-body spectrum that are reminiscent of quantum scars in chaotic non-interacting systems. In the single-particle case, quantum scars correspond to wavefunctions that concentrate in the vicinity of unstable periodic classical trajectories. We show that many-body scars appear in the Fibonacci chain, a model with a constrained local Hilbert space that has recently been experimentally realized in a Rydberg-atom quantum simulator. The quantum scarred eigenstates are embedded throughout the otherwise thermalizing many-body spectrum but lead to direct experimental signatures, as we show for periodic recurrences that reproduce those observed in the experiment. Our results suggest that scarred many-body bands give rise to a new universality class of quantum dynamics, opening up opportunities for the creation of novel states with long-lived coherence in systems that are now experimentally realizable.}, author = {Turner, Christopher and Michailidis, Alexios and Abanin, Dmitry and Serbyn, Maksym and Papić, Zlatko}, journal = {Nature Physics}, pages = {745 -- 749}, publisher = {Nature Publishing Group}, title = {{Weak ergodicity breaking from quantum many-body scars}}, doi = {10.1038/s41567-018-0137-5}, volume = {14}, year = {2018}, } @article{607, abstract = {We study the Fokker-Planck equation derived in the large system limit of the Markovian process describing the dynamics of quantitative traits. The Fokker-Planck equation is posed on a bounded domain and its transport and diffusion coefficients vanish on the domain's boundary. We first argue that, despite this degeneracy, the standard no-flux boundary condition is valid. We derive the weak formulation of the problem and prove the existence and uniqueness of its solutions by constructing the corresponding contraction semigroup on a suitable function space. Then, we prove that for the parameter regime with high enough mutation rate the problem exhibits a positive spectral gap, which implies exponential convergence to equilibrium.Next, we provide a simple derivation of the so-called Dynamic Maximum Entropy (DynMaxEnt) method for approximation of observables (moments) of the Fokker-Planck solution, which can be interpreted as a nonlinear Galerkin approximation. The limited applicability of the DynMaxEnt method inspires us to introduce its modified version that is valid for the whole range of admissible parameters. Finally, we present several numerical experiments to demonstrate the performance of both the original and modified DynMaxEnt methods. We observe that in the parameter regimes where both methods are valid, the modified one exhibits slightly better approximation properties compared to the original one.}, author = {Bodova, Katarina and Haskovec, Jan and Markowich, Peter}, journal = {Physica D: Nonlinear Phenomena}, pages = {108--120}, publisher = {Elsevier}, title = {{Well posedness and maximum entropy approximation for the dynamics of quantitative traits}}, doi = {10.1016/j.physd.2017.10.015}, volume = {376-377}, year = {2018}, } @article{294, abstract = {We developed a method to calculate two-photon processes in quantum mechanics that replaces the infinite summation over the intermediate states by a perturbation expansion. This latter consists of a series of commutators that involve position, momentum, and Hamiltonian quantum operators. We analyzed several single- and many-particle cases for which a closed-form solution to the perturbation expansion exists, as well as more complicated cases for which a solution is found by convergence. Throughout the article, Rayleigh and Raman scattering are taken as examples of two-photon processes. The present method provides a clear distinction between the Thomson scattering, regarded as classical scattering, and quantum contributions. Such a distinction lets us derive general results concerning light scattering. Finally, possible extensions to the developed formalism are discussed.}, author = {Fratini, Filippo and Safari, Laleh and Amaro, Pedro and Santos, José}, journal = {Physical Review A - Atomic, Molecular, and Optical Physics}, number = {4}, publisher = {American Physical Society}, title = {{Two-photon processes based on quantum commutators}}, doi = {10.1103/PhysRevA.97.043842}, volume = {97}, year = {2018}, } @article{606, abstract = {We establish the existence of a global solution for a new family of fluid-like equations, which are obtained in certain regimes in as the mean-field evolution of the supercurrent density in a (2D section of a) type-II superconductor with pinning and with imposed electric current. We also consider general vortex-sheet initial data, and investigate the uniqueness and regularity properties of the solution. For some choice of parameters, the equation under investigation coincides with the so-called lake equation from 2D shallow water fluid dynamics, and our analysis then leads to a new existence result for rough initial data.}, author = {Duerinckx, Mitia and Fischer, Julian L}, journal = {Annales de l'Institut Henri Poincare (C) Non Linear Analysis}, number = {5}, pages = {1267--1319}, publisher = {Elsevier}, title = {{Well-posedness for mean-field evolutions arising in superconductivity}}, doi = {10.1016/j.anihpc.2017.11.004}, volume = {35}, year = {2018}, } @inproceedings{5959, abstract = {Formalizing properties of systems with continuous dynamics is a challenging task. In this paper, we propose a formal framework for specifying and monitoring rich temporal properties of real-valued signals. We introduce signal first-order logic (SFO) as a specification language that combines first-order logic with linear-real arithmetic and unary function symbols interpreted as piecewise-linear signals. We first show that while the satisfiability problem for SFO is undecidable, its membership and monitoring problems are decidable. We develop an offline monitoring procedure for SFO that has polynomial complexity in the size of the input trace and the specification, for a fixed number of quantifiers and function symbols. We show that the algorithm has computation time linear in the size of the input trace for the important fragment of bounded-response specifications interpreted over input traces with finite variability. We can use our results to extend signal temporal logic with first-order quantifiers over time and value parameters, while preserving its efficient monitoring. We finally demonstrate the practical appeal of our logic through a case study in the micro-electronics domain.}, author = {Bakhirkin, Alexey and Ferrere, Thomas and Henzinger, Thomas A and Nickovicl, Deian}, booktitle = {2018 International Conference on Embedded Software}, isbn = {9781538655603}, location = {Turin, Italy}, pages = {1--10}, publisher = {IEEE}, title = {{Keynote: The first-order logic of signals}}, doi = {10.1109/emsoft.2018.8537203}, year = {2018}, } @inproceedings{5962, abstract = {Stochastic Gradient Descent (SGD) is a fundamental algorithm in machine learning, representing the optimization backbone for training several classic models, from regression to neural networks. Given the recent practical focus on distributed machine learning, significant work has been dedicated to the convergence properties of this algorithm under the inconsistent and noisy updates arising from execution in a distributed environment. However, surprisingly, the convergence properties of this classic algorithm in the standard shared-memory model are still not well-understood. In this work, we address this gap, and provide new convergence bounds for lock-free concurrent stochastic gradient descent, executing in the classic asynchronous shared memory model, against a strong adaptive adversary. Our results give improved upper and lower bounds on the "price of asynchrony'' when executing the fundamental SGD algorithm in a concurrent setting. They show that this classic optimization tool can converge faster and with a wider range of parameters than previously known under asynchronous iterations. At the same time, we exhibit a fundamental trade-off between the maximum delay in the system and the rate at which SGD can converge, which governs the set of parameters under which this algorithm can still work efficiently.}, author = {Alistarh, Dan-Adrian and De Sa, Christopher and Konstantinov, Nikola H}, booktitle = {Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing - PODC '18}, isbn = {9781450357951}, location = {Egham, United Kingdom}, pages = {169--178}, publisher = {ACM Press}, title = {{The convergence of stochastic gradient descent in asynchronous shared memory}}, doi = {10.1145/3212734.3212763}, year = {2018}, } @article{5860, abstract = {A major problem for evolutionary theory is understanding the so-called open-ended nature of evolutionary change, from its definition to its origins. Open-ended evolution (OEE) refers to the unbounded increase in complexity that seems to characterize evolution on multiple scales. This property seems to be a characteristic feature of biological and technological evolution and is strongly tied to the generative potential associated with combinatorics, which allows the system to grow and expand their available state spaces. Interestingly, many complex systems presumably displaying OEE, from language to proteins, share a common statistical property: the presence of Zipf's Law. Given an inventory of basic items (such as words or protein domains) required to build more complex structures (sentences or proteins) Zipf's Law tells us that most of these elements are rare whereas a few of them are extremely common. Using algorithmic information theory, in this paper we provide a fundamental definition for open-endedness, which can be understood as postulates. Its statistical counterpart, based on standard Shannon information theory, has the structure of a variational problem which is shown to lead to Zipf's Law as the expected consequence of an evolutionary process displaying OEE. We further explore the problem of information conservation through an OEE process and we conclude that statistical information (standard Shannon information) is not conserved, resulting in the paradoxical situation in which the increase of information content has the effect of erasing itself. We prove that this paradox is solved if we consider non-statistical forms of information. This last result implies that standard information theory may not be a suitable theoretical framework to explore the persistence and increase of the information content in OEE systems.}, author = {Corominas-Murtra, Bernat and Seoane, Luís F. and Solé, Ricard}, issn = {17425689}, journal = {Journal of the Royal Society Interface}, number = {149}, publisher = {Royal Society Publishing}, title = {{Zipf's Law, unbounded complexity and open-ended evolution}}, doi = {10.1098/rsif.2018.0395}, volume = {15}, year = {2018}, } @inproceedings{5961, abstract = {The area of machine learning has made considerable progress over the past decade, enabled by the widespread availability of large datasets, as well as by improved algorithms and models. Given the large computational demands of machine learning workloads, parallelism, implemented either through single-node concurrency or through multi-node distribution, has been a third key ingredient to advances in machine learning. The goal of this tutorial is to provide the audience with an overview of standard distribution techniques in machine learning, with an eye towards the intriguing trade-offs between synchronization and communication costs of distributed machine learning algorithms, on the one hand, and their convergence, on the other.The tutorial will focus on parallelization strategies for the fundamental stochastic gradient descent (SGD) algorithm, which is a key tool when training machine learning models, from classical instances such as linear regression, to state-of-the-art neural network architectures. The tutorial will describe the guarantees provided by this algorithm in the sequential case, and then move on to cover both shared-memory and message-passing parallelization strategies, together with the guarantees they provide, and corresponding trade-offs. The presentation will conclude with a broad overview of ongoing research in distributed and concurrent machine learning. The tutorial will assume no prior knowledge beyond familiarity with basic concepts in algebra and analysis. }, author = {Alistarh, Dan-Adrian}, booktitle = {Proceedings of the 2018 ACM Symposium on Principles of Distributed Computing - PODC '18}, isbn = {9781450357951}, location = {Egham, United Kingdom}, pages = {487--488}, publisher = {ACM Press}, title = {{A brief tutorial on distributed and concurrent machine learning}}, doi = {10.1145/3212734.3212798}, year = {2018}, } @article{5960, abstract = {In this paper we present a reliable method to verify the existence of loops along the uncertain trajectory of a robot, based on proprioceptive measurements only, within a bounded-error context. The loop closure detection is one of the key points in simultaneous localization and mapping (SLAM) methods, especially in homogeneous environments with difficult scenes recognitions. The proposed approach is generic and could be coupled with conventional SLAM algorithms to reliably reduce their computing burden, thus improving the localization and mapping processes in the most challenging environments such as unexplored underwater extents. To prove that a robot performed a loop whatever the uncertainties in its evolution, we employ the notion of topological degree that originates in the field of differential topology. We show that a verification tool based on the topological degree is an optimal method for proving robot loops. This is demonstrated both on datasets from real missions involving autonomous underwater vehicles and by a mathematical discussion.}, author = {Rohou, Simon and Franek, Peter and Aubry, Clément and Jaulin, Luc}, issn = {1741-3176}, journal = {The International Journal of Robotics Research}, number = {12}, pages = {1500--1516}, publisher = {SAGE Publications}, title = {{Proving the existence of loops in robot trajectories}}, doi = {10.1177/0278364918808367}, volume = {37}, year = {2018}, }