Cosmological collider signals of modular spontaneous CP breaking

The following summary extracts and quotes relevant sections of the paper, providing a detailed account of its findings without external commentary or information.

---

Theoretical Framework and Motivation

The authors consider a modular-invariant extension of the Standard Model. The motivation stems from scenarios where CP violation arises dynamically, which is naturally realized in string constructions through target-space modular invariance associated with the modulus tau.

The authors specifically investigate a scenario where this modulus tau acts as the inflaton or evolves dynamically during inflation. In this setup, the CP-violating phases of the Yukawa couplings evolve during inflation. This evolution leads to several key physical consequences:

1. Chemical Potentials: Standard Model fermions effectively develop chemical potentials.

2. Higgs Condensate: This dynamics favours a Higgs condensate, resulting in a measurable effect on the cosmological collider signal.

The authors state that their goal is to compute the resulting oscillatory contribution to the bispectrum, which they find is enhanced by these chemical potentials, but ultimately remains small.

The Modular-Invariant Standard Model

The model adds a complex scalar, the modulus tau, to the SM particle content. The extended Standard Model is invariant under transformations defined by SL(2, Z), where the transformation of matter fields is given by:

psi to (c tau + d)-k psi, H to (c tau + d)-kH

The the effective Lagrangian is described as:

L eff = L kin + L Yuk + L anom - V(H, tau)

Physical Mechanisms During Inflation

During inflation, the modulus tau 0(t) is time-dependent. This dynamics leads to two primary physical descriptions of the resulting chemical potentials:

1. The K Basis: Particles acquire a mass with a time-dependent phase.

2. The Y Basis: Massive particles P acquire... a chemical potential mu P about k P' 0 / f.

In this modular theory, the U(1)-breaking Yukawa interactions are cubic in the matter fields.

Consequences for Particles and Observables

• Higgs Condensate: The Higgs field forms a Bose–Einstein condensate, acquiring a large vacuum expectation value v = sqrt H squared 1/2 = mu H / 2 lambda H.

• Fermion Mass and Chemical Potential: "As a result of the large inflationary Higgs vacuum expectation value v about mu H, the SM fermions acquire inflationary Dirac mass terms m +h.c where = (psi L, R) and m = yv. The time dependence of the modulus contributes to chemical potentials mu L,R = k psi' L, R 0 / f for the left-handed and right-handed components of each SM fermion."

• Fermion Density: The enhanced fermion number density is given by n about 4 pi m mu A squared e-pi m/H mu A.

The Computational Approach

The authors utilize the Schwinger-Keldysh formalism to compute the loop correction. The calculation involves analyzing Schwinger-Keldysh fermion propagators and focusing on the non-local part of the propagators, which encodes on-shell propagation and is proportional to the fermion density.

The Result: Oscillatory Non-Gaussianities

The contribution to the three-point function of delta tau (the bispectrum) is calculated using a triangle diagram. The result for the non-Gaussianity parameter f NL is expressed as:

sqrt osc f NL about pi N c P zeta 2-5+4i lambda m squared (2 + 2i lambda)e pi lambda c 2+(lambda)B chard

In the limit of large mu A, the asymptotic expression is:

f NL N c m 4 P zeta

Conclusion and Comparison with Previous Work

The authors conclude that, based on their rigorous evaluation of the loop integral, We find that f NL is too small to be observed in near-future experiments.

This result differs significantly from previous estimates (e.g [17–21]), which claimed a strong enhancement due to chemical potentials. The authors highlight several reasons for the suppression:

• They use 4-component Dirac fermions, ensuring a unified treatment of the propagators, whereas previous analyses used 2-component Weyl spinors.

• The loop integral calculation shows that contributions from different loop momenta interfere destructively, leading to a substantial suppression of the signal.

• They include the vector chemical potential, which is similar to previous works.

The final result is quantified by Figure 3, showing the amplitude of the oscillatory cosmo-collider signal across various parameters mu A/H and m/H.

The integration of this highly specialized theoretical framework allows for the development of an AI system that moves beyond simple data correlation and into high-fidelity, predictive quantum field theory simulation. This improved AI is not merely a search engine; it is a sophisticated computational physics engine.

Here are the specific improvements and resulting capabilities:

---

1. Implementation of the Schwinger-Keldysh (SK) Formalism Solver:

• Improvement: The AI incorporates a specialized module dedicated to solving the closed-time-contour path integral, replacing generalized heuristics with exact analytical methods derived from sections 4.2 and 4.3.

• Functionality: It can accurately calculate the non-local components of the fermion propagators (D plus or minus plus or minus) in de Sitter space, accounting for time evolution (eta to 0) while maintaining rigorous control over the local terms that are physically irrelevant to particle production.

2. Automated Derivation and Integration of Factorized Loop Amplitudes:

• Improvement: The AI is trained on the factorization techniques used in Section 4.4, specifically for handling the complex structure of the triangle diagram (Eq. 46). This includes a high-precision numerical solver for the resulting integral forms (Eq. 52–56).

• Functionality: It can automatically perform and optimize integrals over conformal time (eta) and momentum (q), determining the exact contribution of helicity states (h= plus or minus 1) without relying on the simplified approximations used in previous literature.

3. Dynamic Model Comparison Module (Parameter Mapping):

• Improvement: The AI is equipped with a comprehensive parameter mapping engine that correlates physical inputs (e.g., modular weight k, decay constant f, and chemical potential mu A) to specific observable outputs (f NL).

• Functionality: It can instantly compare the results of this minimal modular scenario against alternative models (e.g., curvaton scenarios) and quantify the suppression factor (e,g., 1/N or 1/sqrt mu A) in real-time, allowing researchers to determine if a specific parameter set yields an observable signal.

4. Integration of Chemical Potential as a Physical Observable:

• Improvement: The AI treats the effective chemical potentials (mu A, mu V) not as ad-hoc parameters (as some previous works did), but as physically derived consequences of time-dependent Yukawa phases (f P tau(d mu tau)).

• Functionality: It can simulate how a change in the inflationary trajectory (0) instantaneously induces a corresponding shift in the particle distribution, allowing for dynamic prediction of the resulting chemical potential enhancement.

The improved AI system is capable of executing high-level scientific tasks that were previously intractable or prone to human error:

• High-Fidelity Predictive Modeling: It predicts the amplitude of the oscillatory cosmic collider signal (f NL) with precision, accurately accounting for destructive interference and phase mismatch between different loop momenta—a crucial distinction from previous estimates.

• Constraint Testing: Given observational constraints from current or next-generation experiments (e.g., limits on f NL), the AI can instantly reverse-engineer the required parameter space (mu A/H, m/H) within the modular framework to determine if any valid physical solution exists.

• Anomaly Assessment: It can evaluate the impact of loop-suppressed anomalous couplings (Section 4.6) and quantify how these effects scale against the primary fermionic signal, determining when a secondary effect becomes relevant under extreme conditions.

• Automated Theory Comparison: By processing input parameters, it generates comparative reports detailing why a specific mechanism (e.g, the minimal modular setup) provides a more constrained prediction than competing models involving additional degrees of freedom.