arXiv subjects

Mudit Jain

Publications and source records attributed to Mudit Jain.

30 records · Page 2Linked to original sources

arXiv2022

Polarized Vector Oscillons

Oscillons are spatially localized, time-periodic and long-lived configurations that were primarily proposed in scalar field theories with attractive self-interactions. In this letter, we demonstrate that oscillons also exist in the low-energy effective theory of an interacting massive (real) vector field. We provide two types of vector oscillons with vanishing orbital angular momentum, and approximately spherically symmetric energy density, but not field configurations. These are: (1) "directional" oscillons (linearly polarized), with vanishing total intrinsic spin, and (2) "spinning" oscillons (circularly polarized) with a macroscopic instrinsic spin equal to $\hbar\times$ number of particles in the oscillon. In contrast to the case with only gravitational interactions, the two oscillons have different energy at a fixed particle number even in the nonrelativistic limit. By carrying out relativistic $3+1$d simulations, we show that these oscillons can be long-lived (compared to the oscillation time for the fields), and can arise from a range of Gaussian initial spatial profiles. These considerations make vector oscillons potentially relevant during the early universe and in dark photon dark matter, with novel phenomenology related to their polarization.

astro-ph.CO↗

arXiv2021

Polarized solitons in higher-spin wave dark matter

We first show that the effective non-relativistic theory of gravitationally interacting, massive integer-spin fields (spin-$0$, $1$, and $2$ in particular) is described by a $2s+1$ component Schrödinger-Poisson action, where $s$ is the spin of the field. We then construct $s+1$ distinct, gravitationally supported solitons in this non-relativistic theory from identically polarized plane waves. Such solitons are extremally polarized, with macroscopically large spin, but no orbital angular momentum. These $s+1$ solitons form a basis set, out of which partially polarized solitons can be constructed. All such solitons are ground states, have a spherically symmetric energy density but not field configurations. We discuss how solitons in higher-spin fields can be distinguished from scalar solitons, and potential gravitational and non-gravitational probes of them.

hep-th↗

arXiv2021

CMB birefringence from ultra-light axion string networks

The polarization of Cosmic Microwave Background (CMB) photons is rotated as they pass through (ultralight-) axion string loops. Studying this birefringence can reveal valuable information about the axion-photon coupling and the structure of the string network. We develop an approximate analytic formalism and identify a kernel function that can be used to calculate the two-point correlation function for CMB birefringence induced by an arbitrary axion string network. Using this formalism, we evaluate the birefringence signal for some simple loop distributions (including scaling and network collapse). We find that the angular correlation function has a characteristic angular scale set by $θ_\mathrm{min}$, which corresponds to the angular extent of the loops at the time of recombination. This results in a peak in the birefringence power spectrum around $\ell_p \sim 1/θ_\mathrm{min}$. An additional scale, controlled by the axion's mass, is introduced if the network collapses before today.

astro-ph.CO↗

arXiv2021

VisualCheXbert: Addressing the Discrepancy Between Radiology Report Labels and Image Labels

Automatic extraction of medical conditions from free-text radiology reports is critical for supervising computer vision models to interpret medical images. In this work, we show that radiologists labeling reports significantly disagree with radiologists labeling corresponding chest X-ray images, which reduces the quality of report labels as proxies for image labels. We develop and evaluate methods to produce labels from radiology reports that have better agreement with radiologists labeling images. Our best performing method, called VisualCheXbert, uses a biomedically-pretrained BERT model to directly map from a radiology report to the image labels, with a supervisory signal determined by a computer vision model trained to detect medical conditions from chest X-ray images. We find that VisualCheXbert outperforms an approach using an existing radiology report labeler by an average F1 score of 0.14 (95% CI 0.12, 0.17). We also find that VisualCheXbert better agrees with radiologists labeling chest X-ray images than do radiologists labeling the corresponding radiology reports by an average F1 score across several medical conditions of between 0.12 (95% CI 0.09, 0.15) and 0.21 (95% CI 0.18, 0.24).

eess.IV↗

arXiv2020

CheXphoto: 10,000+ Photos and Transformations of Chest X-rays for Benchmarking Deep Learning Robustness

Clinical deployment of deep learning algorithms for chest x-ray interpretation requires a solution that can integrate into the vast spectrum of clinical workflows across the world. An appealing approach to scaled deployment is to leverage the ubiquity of smartphones by capturing photos of x-rays to share with clinicians using messaging services like WhatsApp. However, the application of chest x-ray algorithms to photos of chest x-rays requires reliable classification in the presence of artifacts not typically encountered in digital x-rays used to train machine learning models. We introduce CheXphoto, a dataset of smartphone photos and synthetic photographic transformations of chest x-rays sampled from the CheXpert dataset. To generate CheXphoto we (1) automatically and manually captured photos of digital x-rays under different settings, and (2) generated synthetic transformations of digital x-rays targeted to make them look like photos of digital x-rays and x-ray films. We release this dataset as a resource for testing and improving the robustness of deep learning algorithms for automated chest x-ray interpretation on smartphone photos of chest x-rays.

eess.IV↗

arXiv2020

Explanation for why the Early Universe was Stable and Dominated by the Standard Model

The Standard Model (SM) possesses an instability at high scales that would be catastrophic during or just after inflation, and yet no new physics has been seen to alter this. Furthermore, modern developments in quantum gravity suggest that the SM degrees of freedom are not unique; that a typical low energy effective theory should include a large assortment of hidden sector degrees of freedom. It is therefore puzzling that cosmological constraints from BBN and CMB reveal that the early universe was almost entirely dominated by the SM, when the inflaton $ϕ$ could have decayed into many sectors. In this work we propose the following explanation for all of this: we allow the lowest dimension operators with natural coefficients between the inflaton and both the Higgs and hidden sectors. Such hidden sectors are assumed to be entirely natural; this means all unprotected masses are pushed up to high scales and project out of the spectrum, while only massless (or protected) degrees of freedom remain, and so the inflaton can only reheat these sectors through higher dimension (and suppressed) operators. On the other hand, the SM possesses a special feature: it includes a light Higgs $H$, presumably for life to exist, and hence it allows a super-renormalizable coupling to the inflaton $ϕH^\dagger H$, which allows rapid decay into the SM. We show that this naturally (i) removes the instability in the Higgs potential both during and after inflation due to a tree-level effect that increases the value of the Higgs self-coupling from the IR to the UV when one passes the inflaton mass, (ii) explains why the SM is dominant in the early universe, (iii) allows dark matter to form in hidden sector/s through subsequent dynamics (or axions, etc), (iv) allows for high reheating and baryogenesis, and (v) accounts for why there so far has been no direct detection of dark matter or new physics beyond the SM.

hep-ph↗

arXiv2020

Ground State Wave Function Overlap in Superconductors and Superfluids

In order to elucidate the quantum ground state structure of non-relativistic condensates, we explicitly construct the ground state wave function for multiple species of bosons, describing either superconductivity or superfluidity. Since each field $Ψ_j$ carries a phase $θ_j$ and the Lagrangian is invariant under rotations $θ_j \to θ_j + α_j$ for independent $α_j$, one can investigate the corresponding wave function overlap between a pair of ground states $\langle G|G'\rangle$ differing by these phases. We operate in the infinite volume limit and use a particular prescription to define these states by utilizing the position space kernel and regulating the UV modes. We show that this overlap vanishes for most pairs of rotations, including $θ_j \to θ_j + m_j ε$, where $m_j$ is the mass of each species, while it is unchanged under the transformation $θ_j \to θ_j + q_j ε$, where $q_j$ is the charge of each species. We explain that this is consistent with the distinction between a superfluid, in which there is a non-trivial conserved number, and the superconductor, in which the electric field and conserved charge is screened, while it is compatible with a non-zero order parameter in both cases. Moreover, we find that this bulk ground state wave function overlap directly reflects the Goldstone boson structure of the effective theory and provides a useful diagnostic of its physical phase.

cond-mat.supr-con↗

arXiv2020

Clustering of cosmic string loops

Observational effects of cosmic string loops depend on how loops are distributed in space. Chernoff \cite{Chernoff} has argued that loops can be gravitationally captured in galaxies and that for sufficiently small values of $Gμ$ their distribution follows that of dark matter, independently of the loop's length. We re-analyze this issue using the spherical model of galaxy formation with full account taken of the gravitational rocket effect -- loop accelerated motion due to asymmetric emission of gravitational waves. We find that only loops greater than a certain size are captured and that the number of captured loops is orders of magnitude smaller than estimated by Chernoff.

astro-ph.CO↗

arXiv2020

Eternal Inflation and Reheating in the Presence of the Standard Model Higgs Field

We study the details of eternal inflation in the presence of a spectator Higgs field within the framework of the minimal Standard Model. We have recently shown that in the presence of scalar field(s) which allow inflation only within a finite domain of field values, the universe reaches a steady state where the normalized distribution for the field(s) converges to a steady state distribution [1]. In this paper, we analyze this eternal inflation scenario with the renormalized Standard Model Higgs potential, since it also allows inflation in a finite domain, but turns over at high scales due to the running of the self-coupling, marking an exit from inflation. We compute the full steady state distribution for the Higgs using an integral evolution technique that we formulated in [1] and the fractal dimension of the universe. We then obtain a bound on the inflationary Hubble scale in order to have a large observable universe contained within the instability scale of $H\lesssim\mathcal{O}(10^9)$GeV depending on the top mass. Upon reheating of the universe, thermal fluctuations in the Higgs field could potentially pose another problem; however, we compute the rate of thermal bubble production and find that the probability of tunneling in the post-inflationary era is negligibly small even for very high reheat temperatures.

hep-ph↗

arXiv2019

Statistics of Inflating Regions in Eternal Inflation

We compute the distribution of sizes of inflating and non-inflating regions in an eternally inflating Universe. As a first illustrative problem, we study a simple scenario of an eternally inflating Universe in the presence of a massless scalar field $φ$ whose field values lie within some finite domain $φ\in(-φ_{cr},φ_{cr})$, with $\pmφ_{cr}$ marking the onset of thermalization/crunching. We compute many important quantities, including the fractal dimension, distribution of field values among inflating regions, and the number of inflating and non-inflating Hubble regions. With the aid of simulations in 1 spatial dimension, we show this eternally inflating Universe reaches a steady state in which average sizes of inflating regions grow only as a power law in the field's crunch value $\sim φ_{cr}^2$ (extension to higher dimensions is $\simφ^{2/D}$), contrary to a naive expectation of an exponential dependence. Furthermore, the distribution in sizes exhibits an exponential fall off for large distances (with an initial power law for inflating regions). We leave other interesting cases of more realistic potentials and time varying Hubble parameter for future work, with a possible application to the SM Higgs in the early Universe.

astro-ph.CO↗

arXiv2019

Counting of States in Higgs Theories

We enumerate the micro-states in Higgs theories, addressing (i) the number of vacuum states and (ii) the appropriate measure in the quantum path integral. To address (i) we explicitly construct the set of ground state wave-functionals in the field basis focussing on scalar modes $θ(x)$. Firstly, we show that in the limit in which the gauge coupling is zero, we obtain an infinite set of degenerate ground states at large volume distinguished by $θ(x)\toθ(x)+θ_0$, spontaneously breaking the global symmetry, as is well known. We then show that at finite gauge coupling there is a unique ground state at large volume since the wave-functional only depends on $\nablaθ$ in the IR, and we explain this at the level of the Lagrangian. Since gauge fields fall off exponentially from sources there are no conserved charges or symmetries in the Higgs phase; so the Higgs mechanism is the removal of symmetry from the theory. We show how physical features of defects, such as cosmic strings in the abelian Higgs model, are best understood in this context. Since there is a unique ground state, we address (ii) whether the path integral is a volume measure for the radial Higgs field $\mathcal{D}ρ\,ρ^{N-1}$ from the $N$ components of the Higgs multiplet, or a line measure $\mathcal{D}ρ$ as the $N-1$ would-be Goldstones can be removed in unitary gauge. We prove that the need to avoid quartic divergences demands a tower of counter terms that resum to exactly give the volume measure. So the size of the Hilbert space in the zero gauge coupling case and finite gauge coupling case are in one-to-one correspondence, despite the degeneracy of the ground state being lifted in the latter. As a cosmological application, we point out that the volume measure can make it exponentially more unlikely in $N(=4)$ for the Standard Model Higgs to relax to the electroweak vacuum in the early universe.

hep-th↗

arXiv2016

Generating Functionals for Quantum Field Theories with Random Potentials

We consider generating functionals for computing correlators in quantum field theories with random potentials. Examples of such theories include condensed matter systems with quenched disorder (e.g. spin glass) or cosmological systems in context of the string theory landscape (e.g. cosmic inflation). We use the so-called replica trick to define two different generating functionals for calculating correlators of the quantum fields averaged over a given distribution of random potentials. The first generating functional is appropriate for calculating averaged (in-out) amplitudes and involves a single replica of fields, but the replica limit is taken to an (unphysical) negative one number of fields outside of the path integral. When the number of replicas is doubled the generating functional can also be used for calculating averaged probabilities (squared amplitudes) using the in-in construction. The second generating functional involves an infinite number of replicas, but can be used for calculating both in-out and in-in correlators and the replica limits are taken to only a zero number of fields. We discuss the formalism in details for a single real scalar field, but the generalization to more fields or to different types of fields is straightforward. We work out three examples: one where the mass of scalar field is treated as a random variable and two where the functional form of interactions is random, one described by a Gaussian random field and the other by a Euclidean action in the field configuration space.

hep-th↗