We explore the application of machine learning based on mixture density neural networks (MDNs) to the interior characterization of low-mass exoplanets up to 25 Earth masses constrained by mass, radius, and fluid Love number $k_2$. We create a dataset of 900$\:$000 synthetic planets, consisting of an iron-rich core, a silicate mantle, a high-pressure ice shell, and a gaseous H/He envelope, to train a MDN using planetary mass and radius as inputs to the network. For this layered structure, we show that the MDN is able to infer the distribution of possible thicknesses of each planetary layer from mass and radius of the planet. This approach obviates the time-consuming task of calculating such distributions with a dedicated set of forward models for each individual planet. While gas-rich planets may be characterized by compositional gradients rather than distinct layers, the method presented here can be easily extended to any interior structure model. The fluid Love number $k_2$ bears constraints on the mass distribution in the planets' interior and will be measured for an increasing number of exoplanets in the future. Adding $k_2$as an input to the MDN significantly decreases the degeneracy of the possible interior structures.

We report the discovery of OGLE-2016-BLG-1190Lb, which is likely to be the first Spitzer microlensing planet in the Galactic bulge/bar, an assignation that can be confirmed by two epochs of high-resolution imaging of the combined source-lens baseline object. The planet's mass M_p= 13.4+-0.9 M_J places it right at the deuterium burning limit, i.e., the conventional boundary between "planets" and "brown dwarfs". Its existence raises the question of whether such objects are really "planets" (formed within the disks of their hosts) or "failed stars" (low mass objects formed by gas fragmentation). This question may ultimately be addressed by comparing disk and bulge/bar planets, which is a goal of the Spitzer microlens program. The host is a G dwarf M_host = 0.89+-0.07 M_sun and the planet has a semi-major axis a~2.0 AU. We use Kepler K2 Campaign 9 microlensing data to break the lens-mass degeneracy that generically impacts parallax solutions from Earth-Spitzer observations alone, which is the first successful application of this approach. The microlensing data, derived primarily from near-continuous, ultra-dense survey observations from OGLE, MOA, and three KMTNet telescopes, contain more orbital information than for any previous microlensing planet, but not quite enough to accurately specify the full orbit. However, these data do permit the first rigorous test of microlensing orbital-motion measurements, which are typically derived from data taken over <1% of an orbital period.

The formation and dynamical history of hot Jupiters is currently debated, with wide stellar binaries having been suggested as a potential formation pathway. Additionally, contaminating light from both binary companions and unassociated stars can significantly bias the results of planet characterisation studies, but can be corrected for if the properties of the contaminating star are known. We search for binary companions to known transiting exoplanet host stars, in order to determine the multiplicity properties of hot Jupiter host stars. We also characterise unassociated stars along the line of sight, allowing photometric and spectroscopic observations of the planetary system to be corrected for contaminating light. We analyse lucky imaging observations of 97 Southern hemisphere exoplanet host stars, using the Two Colour Instrument on the Danish 1.54m telescope. For each detected companion star, we determine flux ratios relative to the planet host star in two passbands, and measure the relative position of the companion. The probability of each companion being physically associated was determined using our two-colour photometry. A catalogue of close companion stars is presented, including flux ratios, position measurements, and estimated companion star temperature. For companions that are potential binary companions, we review archival and catalogue data for further evidence. For WASP-77AB and WASP-85AB, we combine our data with historical measurements to determine the binary orbits, showing them to be moderately eccentric and inclined to the line of sight and planetary orbital axis. Combining our survey with the similar Friends of Hot Jupiters survey, we conclude that known hot Jupiter host stars show a deficit of high mass stellar companions compared to the field star population; however, this may be a result of the biases in detection and target selection by ground-based surveys.