Aspects of Scale invariance in Physics and Biology

Abstract

We study three systems that have scale invariance. The first system is a conformal field theory in d > 3 dimensions. We prove that if there is a unique stress-energy tensor and at least one higher-spin conserved current in the theory, then the correlation functions of the stress-energy tensors and the conserved currents of higher-spin must coincide with one of the following possibilities: a) a theory of n free bosons, b) a theory of n free fermions or c) a theory of n (d-2)/2-forms.

The second system is the primordial gravitational wave background in a theory with inflation. We show that the scale invariant spectrum of primordial gravitational waves is isotropic only in the zero-order approximation, and it gets a small correction due to the primordial scalar fluctuations. When anisotropy is measured experimentally, our result will allow us to distinguish between different inflationary models.

The third system is a biological system. The question we are asking is whether there is some simplicity or universality underlying the complexities of natural animal behavior. We use the walking fruit fly (Drosophila melanogaster) as a model system. Based on the result that unsupervised flies’ behaviors can be categorized into one hundred twenty- two discrete states (stereotyped movements), which all individuals from a single species visit repeatedly, we demonstrated that the sequences of states are strongly non-Markovian. In particular, correlations persist for an order of magnitude longer than expected from a model of random state- to-state transitions. The correlation function has a power-law decay, which is a hint of some kind of criticality in the system. We develop a generalization of the information bottleneck method that allows us to cluster these states into a small number of clusters. This more compact description preserves a lot of temporal correlation. We found that it is enough to use a two-cluster representation of the data to capture long- range correlations, which opens a way for a more quantitative description of the system. Usage of the maximal entropy method allowed us to find a description that closely resembles a famous inverse-square Ising model in 1d in a small magnetic field.