SearcharxivSearch

arXiv subjects

Trevor Cheung

Publications and source records attributed to Trevor Cheung.

2 recordsLinked to original sources

Massive Spinning Fields During Inflation: Feynman rules and correlator comparison

We consider the dynamics of massive spinning fields during inflation and the resulting signatures in the cosmological correlators of inflaton perturbations computed in the Poincar\'e patch of de Sitter space. There are (at least) two ways to describe the fluctuations of such new spinning degrees of freedom and these are distinguished by the symmetries of the de Sitter group that they linearly realise. The primary question we ask is: do these two set-ups yield distinct signatures in cosmological observables? After systematically deriving the Feynman rules for exchange diagrams consisting of massive spinning fields, where we discover the necessity of effective propagators that augment the naive Schwinger-Keldysh ones by delta functions corresponding to instantaneous propagation, we show that the two set-ups are indistinguishable at the level of the inflaton bispectrum but distinguishable at the level of the trispectrum and other higher-point correlation functions. The bispectrum is special since in the corresponding tree-level Feynman diagrams, only the helicity-zero modes of the spinning fields can propagate. The bispectrum correspondence holds up to the addition of contact diagrams arising from the self-interactions of the inflaton, and is consistent with the symmetries of the effective field theory of inflation. Our results suggest that the cosmological collider signals in the bispectrum are universal and do not depend on the detailed description of the massive spinning field.

hep-th

The Multi-modality Cell Segmentation Challenge: Towards Universal Solutions

Cell segmentation is a critical step for quantitative single-cell analysis in microscopy images. Existing cell segmentation methods are often tailored to specific modalities or require manual interventions to specify hyper-parameters in different experimental settings. Here, we present a multi-modality cell segmentation benchmark, comprising over 1500 labeled images derived from more than 50 diverse biological experiments. The top participants developed a Transformer-based deep-learning algorithm that not only exceeds existing methods but can also be applied to diverse microscopy images across imaging platforms and tissue types without manual parameter adjustments. This benchmark and the improved algorithm offer promising avenues for more accurate and versatile cell analysis in microscopy imaging.

eess.IV