scispace - formally typeset
Search or ask a question
Institution

École Normale Supérieure

OtherParis, Île-de-France, France
About: École Normale Supérieure is a other organization based out in Paris, Île-de-France, France. It is known for research contribution in the topics: Population & Catalysis. The organization has 68439 authors who have published 99414 publications receiving 3092008 citations.


Papers
More filters
Journal ArticleDOI
06 Nov 2008-Nature
TL;DR: It is shown that winter weather and snow conditions, together with density dependence in the net population growth rate, account for the observed population dynamics of the rodent community dominated by lemmings in an alpine Norwegian core habitat between 1970 and 1997, and predict the observed absence of rodent peak years after 1994.
Abstract: Norwegian lemmings (Lemmus lemmus) are well known for their population cycles, which are thought, at their peak, to influence other ecosystem components. In fact the role of the physical environment — climate included — in determining rodent cycle dynamics has remained largely a matter of conjecture. Now from a combination of long-term (1970–2007) data on rodent density, bird densities and field estimates of snow pack conditions together with meteorological data, a clearer picture of the lemming cycle has been obtained. What emerges is a marked shift away from the familiar 3–5-year rodent cycles to an aperiodic, mostly low-amplitude state, which can be explained and predicted by the between-year variations in winter climate. There is strong evidence for the hypothesis that climate effects on rodent dynamics are transmitted to other parts of the ecosystem. The population cycles of rodents at northern latitudes have puzzled people for centuries1,2, and their impact is manifest throughout the alpine ecosystem2,3. Climate change is known to be able to drive animal population dynamics between stable and cyclic phases4,5, and has been suggested to cause the recent changes in cyclic dynamics of rodents and their predators3,6,7,8,9. But although predator–rodent interactions are commonly argued to be the cause of the Fennoscandian rodent cycles1,10,11,12,13, the role of the environment in the modulation of such dynamics is often poorly understood in natural systems8,9,14. Hence, quantitative links between climate-driven processes and rodent dynamics have so far been lacking. Here we show that winter weather and snow conditions, together with density dependence in the net population growth rate, account for the observed population dynamics of the rodent community dominated by lemmings (Lemmus lemmus) in an alpine Norwegian core habitat between 1970 and 1997, and predict the observed absence of rodent peak years after 1994. These local rodent dynamics are coherent with alpine bird dynamics both locally and over all of southern Norway, consistent with the influence of large-scale fluctuations in winter conditions. The relationship between commonly available meteorological data and snow conditions indicates that changes in temperature and humidity, and thus conditions in the subnivean space, seem to markedly affect the dynamics of alpine rodents and their linked groups. The pattern of less regular rodent peaks, and corresponding changes in the overall dynamics of the alpine ecosystem, thus seems likely to prevail over a growing area under projected climate change.

441 citations

Proceedings Article
08 Dec 2019
TL;DR: This work shows that this "lazy training" phenomenon is not specific to over-parameterized neural networks, and is due to a choice of scaling that makes the model behave as its linearization around the initialization, thus yielding a model equivalent to learning with positive-definite kernels.
Abstract: In a series of recent theoretical works, it was shown that strongly over-parameterized neural networks trained with gradient-based methods could converge exponentially fast to zero training loss, with their parameters hardly varying. In this work, we show that this "lazy training" phenomenon is not specific to over-parameterized neural networks, and is due to a choice of scaling, often implicit, that makes the model behave as its linearization around the initialization, thus yielding a model equivalent to learning with positive-definite kernels. Through a theoretical analysis, we exhibit various situations where this phenomenon arises in non-convex optimization and we provide bounds on the distance between the lazy and linearized optimization paths. Our numerical experiments bring a critical note, as we observe that the performance of commonly used non-linear deep convolutional neural networks in computer vision degrades when trained in the lazy regime. This makes it unlikely that "lazy training" is behind the many successes of neural networks in difficult high dimensional tasks.

441 citations

Journal ArticleDOI
16 May 2019-Cell
TL;DR: An ∼12-fold expanded global ocean DNA virome dataset is established of 195,728 viral populations, now including the Arctic Ocean, and it is validated that these populations form discrete genotypic clusters.

441 citations

Journal ArticleDOI
TL;DR: It is shown, using theory and numerical simulation, that the landscape is much rougher than is classically assumed and undergoes a 'roughness transition' to fractal basins, which brings about isostaticity and marginal stability on approaching jamming.
Abstract: Glasses are amorphous solids whose constituent particles are caged by their neighbors and thus cannot flow. This sluggishness is often ascribed to the free energy landscape containing multiple minima (basins) separated by high barriers. Here we show, using theory and numerical simulation, that the landscape is much rougher than is classically assumed. Deep in the glass, it undergoes a "roughness transition" to fractal basins. This brings about isostaticity at jamming and marginality of glassy states near jamming. Critical exponents for the basin width, the weak force distribution, and the spatial spread of quasi-contacts at jamming can be analytically determined. Their value is found to be compatible with numerical observations. This advance therefore incorporates the jamming transition of granular materials into the framework of glass theory. Because temperature and pressure control which features of the landscape are experienced, glass mechanics and transport are expected to reflect the features of the topology we discuss here. Hitherto mysterious properties of low-temperature glasses could be explained by this approach.

441 citations

Journal ArticleDOI
15 Jun 2006-Neuron
TL;DR: It is shown using magnetic resonance imaging in humans that the posterior portion of the prefrontal cortex, including Broca's area and its homolog in the right hemisphere, contains a system of executive processes that control start and end states and the nesting of functional segments that combine in hierarchically organized action plans.

441 citations


Authors

Showing all 68584 results

NameH-indexPapersCitations
Didier Raoult1733267153016
Simon Baron-Cohen172773118071
Andrew Zisserman167808261717
Edward T. Bullmore165746112463
H. Eugene Stanley1541190122321
Pierre Bourdieu153592194586
Gerald M. Rubin152382115248
Stanislas Dehaene14945686539
Melody A. Swartz1481304103753
J. Fraser Stoddart147123996083
Jean-François Cardoso145373115144
Richard S. J. Frackowiak142309100726
Cordelia Schmid135464103925
Jean Tirole134439103279
Ion Stoica13349394937
Network Information
Related Institutions (5)
Centre national de la recherche scientifique
382.4K papers, 13.6M citations

96% related

École Polytechnique Fédérale de Lausanne
98.2K papers, 4.3M citations

95% related

University of Paris
174.1K papers, 5M citations

94% related

ETH Zurich
122.4K papers, 5.1M citations

94% related

Pierre-and-Marie-Curie University
56.1K papers, 2.3M citations

93% related

Performance
Metrics
No. of papers from the Institution in previous years
YearPapers
202340
2022382
20213,853
20204,300
20194,313
20184,336