### Toward a measurement with CMB-Stage 4 Experiment

A major part of my research has focused on doing simulations and forecasting works to design the best data analysis techniques for probing the curvaton model, a type of inflation model with an extra scalar field. In particular, with Prof. Daniel Grin and Wayne Hu, I developed and tested new techniques to improve the signal-to-noise of the curvaton signatures in the CMB and to remove bias effects from gravitational lensing. Measuring those signatures with CMB-Stage 4 will confirm or eliminate certain scenarios of the curvaton model in which baryon number or dark matter originates from the curvaton.

### Compensated Isocurvature Perturbations in the Curvaton Model

The curvaton model naturally produces compensated isocurvature perturbations (CIPs), a mode of perturbation in which the cold dark matter (CDM) and baryons have exactly opposite fluctuations. These CIPs modulate the sound horizon across the CMB sky and break statistical isotropy. They can therefore be reconstructed using correlations of different CMB multipoles analogous to lensing reconstruction (see Grin et al. 2011).

In He et al. 2015, with Prof. Daniel Grin and Wayne Hu, I improved the signal-to-noise of the CIPs from the curvaton model by a factor of 2-3 by taking advantage of its cross-correlations with the CMB. This uses the fact that CIPs in the curvaton model are correlated with the underlying curvature fluctuations, which dominate the CMB anisotropies. The amount of correlation A differs in different decay-scenarios of the curvaton, so that measuring A provides a way to distinguish between different scenarios. It turns out that the factor of 2-3 improvement is key in enabling CMB-Stage 4 to detect (at 3 sigma) CIPs from the curvaton scenario with the largest CIP case — in which baryon number is a product of curvaton decay and CDM is produced non-thermally before the decay.