Gravitational waves provide a new probe into the strong-field regime of gravity. It is thus essential to identify the predictions of General Relativity on the nature of the two-body problem, and to contrast them to alternative theories. In this talk, I will review how Effective Field Theory formalisms allow to place model-independent constraints on gravitational theories. I will first introduce some of the tools used in the General Relativistic two-body problem, then show how one can adapt them to Scalar-Tensor theories. Finally, I will comment on black holes properties in Scalar-Tensor theories (the so-called "no-hair theorem") and introduce a new Effective Field Theory designed to test these properties with gravitational waveforms.

The growth rate of cosmic structures is a key quantity to probe gravitational interactions and dark energy. In the late-time Universe the growth rate becomes non-linear, making it more difficult to probe, but also makes it in principle a source of additional cosmological information than its linear counterpart.
In this talk, I will discuss why it becomes interesting to probe the growth rate in regions of different density, as well as recent developments in measuring the growth rate using RSD around voids.

Observations of the large-scale structure (LSS) of the Universe offer an increasingly competitive probe of dark energy, as it is shaped by the opposing effects of gravity and the accelerated expansion. Even though it is not directly observable, it can be traced by galaxies and reconstructed by the gravitational lensing signal it imprints on galaxy survey and cosmic microwave background (CMB) data. In this talk, I will present how these probes can be jointly analyzed, including their cross-correlations, to tighten constraints on cosmological parameters. I will first focus on multiprobe analyses with current galaxy and CMB surveys, in particular from the ongoing Dark Energy Survey. I will then show how to test cosmological models with multiple, correlated observations of the LSS and how multiprobe analysis will thus enable upcoming surveys, such as Rubin/LSST, Euclid and Roman/WFIRST, to track hints of new physics from massive data sets.

In this talk we will show the cosmological implications of a varying absolute magnitude of Type Ia supernovae using the Pantheon compilation, and we test them by reconstructing different phenomenological approaches that could justify a varying absolute magnitude, but also approaches based on cosmic voids, modified gravity models and a binning scheme. Anticipating the results, we find good agreement with the expected values of the standard LCDM model and no evidence of new physics.

Information about the Universe is encoded in the large scale structure, which can be extracted using statistical tools. In particular, the three-point correlation function (bispectrum) of the matter overdensity provides complementary information to its two-point counterpart (power spectrum). In the first part of the talk, I will present my results on the large-scale matter bispectrum using perturbative methods up to three loops, showing the scales up to which they are valid by comparing to numerical simulations and I will describe phenomenological techniques to model the bispectrum on smaller scales. In the second part of the talk, I will explain how, by adding the bias and RSDs, the galaxy bispectrum can be used to tighten constraints on the amplitude of primordial non-Gaussianity with respect to the CMB values from Planck.

The dark sector of our universe might very well be constituted of particles with masses as large as the Exa-electronvolt (10^9 GeV). I will discuss in this talk how such a heavy dark-matter candidate might be populated in the early universe and how one could hope to detect it nowadays in astroparticle searches experiments. In particular, I will discuss the possible imprint of such models on the dynamics of cosmic inflation, but also how such a heavy candidate might be metastable and decay in the present universe, leading to detectable signals in the form of ultra-high energy cosmic rays.

There is mounting evidence that the stellar initial mass function (IMF) could extend much beyond the canonical Mi ~100, Msun limit, but the impact of such hypothesis on the chemical enrichment of galaxies still remains to be clarified.

We aim to address this question by analysing the observed abundances of thin- and thick-disc stars in the Milky Way with chemical evolution models that account for the contribution of very massive stars dying as pair-instability supernovae.

We built new sets of chemical yields from massive and very massive stars up to Mi ~ 350 Msun, by combining the wind ejecta extracted from our hydrostatic stellar evolution models with explosion ejecta from the literature. Using a simple chemical evolution code we analyse the effects of adopting different yield tables by comparing predictions against observations of stars in the solar vicinity.

After several tests, we focus on the [O/Fe] ratio which best separates the chemical patterns of the two Milky Way components. We find that with a standard IMF, truncated at Mi ~ 100 Msun, we can reproduce various observational constraints for thin-disc stars, but the same IMF fails to account for the [O/Fe] ratios of thick-disc stars. The best results are obtained by extending the IMF up to Mi = 350 Msun and including the chemical ejecta of very massive stars, in the form of winds and pair-instability supernova explosions.

Our study indicates that PISN played a significant role in shaping the chemical evolution of the Milky Way thick disc. By including their chemical yields it is easier to reproduce not only the level of the alpha-enhancement but also the observed slope of thick-disc stars in the [O/Fe] vs [Fe/H] diagram. The bottom line is that the contribution of very massive stars to the chemical enrichment of galaxies is potentially quite important and should not be neglected in chemical evolution models.

In this talk, I will discuss some aspects of non-linear structure formation.
Calculations of correlation functions usually involve linear (relativistic) physics on large scales and non-linear (but Newtonian) physics on small scales. For the standard case of the 2-point correlation function, the relativistic effects remain small. However, for non-Gaussian signals, one needs to consider, the 3-point correlation function (the bispectrum) which couples large scales with small scales in the so-called squeezed limit. I will present the calculation of the non-linear relativistic corrections to the dynamics of dark matter and galaxy perturbations. The relativistic corrections are of the same order of magnitude, in the squeezed limit, than a primordial local non-Gaussianity signal. Such signal is expected to be within the accuracy of the next generation of LSS surveys. I will finally complement my study with some preliminary results of N-body simulations.

General Relativity (GR) provides so far our best effective field theory of the gravitational field and its interaction with matter. However, provided some energy conditions, it is well known that the theory predicts that pathological singularities develop at sufficiently high energy, breaking the predictability of cosmological and gravitational collapse solutions of GR beyond these focusing points.

Bouncing cosmologies have provided useful models where the initial Big Bang singularity is removed and where the universe transits smoothly from a contraction phase to the current expanding one. Over the last decades, numerous scenarios motivated either by modified gravity theories or quantum gravity proposals have been investigated to resolve the singularity and the phenomenology of these bouncing cosmologies has been intensively studied to confront them to the data available for early cosmology (See: https://arxiv.org/abs/1406.2790).

More recently, the idea to import the techniques from bouncing cosmologies to build model of non-singular gravitational collapse has attracted a lot of efforts. This idea appears indeed a natural extension of bouncing cosmologies to astrophysical objects. The resulting models of bouncing compact objects provide a new platform for phenomenology and suggest a new potential astrophysical window to test quantum gravity proposals.

In this talk, I will review this line of research and present a model of bouncing compact object based which provides an effective quantum extension of the seminal Oppenheimer-Snyder collapse model. I will discuss the main challenge and argue that crucial information on the causal structure of such bouncing compact object can be extracted in a model independent manner. While I will focus on the theoretical construction of the model, I shall also discuss the potential observational signatures it suggests to further investigate the physics of such exotic astrophysical object.

This talk is based on the recent work:
1) Towards consistent black-to-white hole bounces from matter collapse, JCAP, 2020
J. Ben Achour, S. Brahma, S. Mukohyama, J-P. Uzan
(https://arxiv.org/abs/2004.12977)

See also:
1) Bouncing compact objects. II. Effective theory of a pulsating Planck star, PRD, 2020
J. Ben Achour, J-P. Uzan,
(https://arxiv.org/abs/2001.06153)
2) Bouncing compact objects. I. Quantum extension of the Oppenheimer-Snyder collapse, 2020, JCAP
J. Ben Achour, S. Brahma, J-P. Uzan
(https://arxiv.org/abs/2001.06148)