Every Wrinkle Carries A Memory: An Integro-differential Bootstrap for Features in Cosmological Correlators

This paper extends the cosmological bootstrap program to scale-breaking scenarios by deriving integro-differential boundary equations with memory kernels for heavy fields with time-dependent masses, enabling analytical and numerical solutions that reveal exponentially enhanced primordial features in the squeezed bispectrum.

The Gist

Imagine the universe as a giant, expanding balloon. For a long time, physicists have been trying to figure out what happened when that balloon was first being blown up (a period called "inflation"). To do this, they look at the "fossils" left behind: tiny ripples in the cosmic microwave background and the distribution of galaxies today. These ripples are like the wrinkles on the surface of the balloon.

Usually, physicists assume the balloon was expanding perfectly smoothly and evenly. This makes the math easy, like a song with a steady, unchanging beat. But what if the balloon wasn't perfectly smooth? What if it had a rhythmic, wobbly pulse as it expanded? Maybe the "fabric" of space itself was vibrating like a guitar string.

This paper is about learning how to read the music of a wobbly universe.

The Big Idea: The "Memory" of the Universe

The authors introduce a new way of doing physics called the "Cosmological Bootstrap."

Think of the universe like a movie.

• The Old Way (The "Bulk" View): To understand the movie, you usually have to watch every single frame from start to finish, calculating exactly what every character does at every moment. This is incredibly hard, especially if the characters are heavy and the scenery is changing.
• The New Way (The "Boundary" View): Instead of watching the whole movie, you only look at the very last frame (the "boundary" of the universe today). The Bootstrap method says: "If we know the rules of the game (like cause-and-effect and energy conservation), we can figure out the whole movie just by looking at the final picture."

In a smooth, perfect universe, this "final picture" follows simple rules (differential equations). But in this paper, the authors tackle a universe that is breaking the rules of smoothness. The mass of heavy particles is oscillating (wobbling) over time.

The "Integro-Differential" Equation: A Recipe with a Memory

When the universe wobbles, the simple rules break down. The authors found that the equations describing these wrinkles become "Integro-Differential Equations."

That's a fancy name for a recipe that has a memory.

• Differential: "How fast is the cake rising right now?"
• Integral: "But wait, the cake's rise right now depends on how much it rose yesterday, and the day before, and the day before that."

In this cosmic recipe, the "memory" is a kernel. It's like a ghostly echo of the universe's past evolution. Every time you try to calculate a feature of the universe today, you have to sum up the effects of everything that happened in the past, weighted by how the universe was wobbling back then.

The "Cosmological Collider": Listening for Heavy Particles

The paper focuses on "heavy" particles that existed during inflation but are too heavy to be created in our current particle accelerators (like the Large Hadron Collider). We can't build a collider big enough to make them, so we use the entire universe as our collider.

When these heavy particles were created and then decayed, they left a specific "sound" or pattern in the wrinkles of the universe.

• The Standard Signal: Usually, these heavy particles are so heavy that they are "Boltzmann suppressed." Imagine trying to hear a whisper in a hurricane; the signal is so faint it's almost impossible to detect.
• The New Discovery: The authors found that if the universe is wobbling (oscillating mass) at just the right frequency, it acts like a resonance chamber. It's like pushing a swing at the exact right moment. This "parametric resonance" amplifies the signal massively.

They found that these heavy particles can be "pumped" into existence by the wobbly mass, making their signal exponentially louder. It's like turning a whisper into a shout just by tuning the frequency of the wind.

The "Numerical Bootstrap": Solving the Puzzle with a Computer

Solving these "memory" equations by hand is a nightmare. The authors did two things:

1. Analytical Solution: They solved a simplified version of the puzzle using advanced math, finding that the signal has a specific, rhythmic pattern (oscillations) that breaks the usual symmetry.
2. Numerical Solution: They built a computer program to solve the full, messy equation. This is the first time anyone has successfully "bootstrapped" a cosmological correlator using a computer in this way. They treated the equation like a grid of numbers and solved it step-by-step, confirming their math was right.

The "Wrinkle" Metaphor

The title says, "Every Wrinkle Carries A Memory."

Imagine the surface of the universe is a piece of fabric.

• In a smooth universe, the wrinkles are simple and predictable.
• In this wobbling universe, the fabric is being shaken. Every time a wrinkle forms, it "remembers" the specific shake that caused it.
• The authors developed a new tool to look at a wrinkle and say, "Ah, this specific shape tells me that the fabric was shaken at a frequency of X, and it created a heavy particle that we can't see directly, but we can hear its echo."

Why Does This Matter?

1. New Physics: It opens a window to see particles that are far too heavy for our current technology to create.
2. Better Tools: It gives us a new "bootstrap" method that works even when the universe isn't perfectly symmetrical.
3. Future Observations: The authors predict that upcoming telescopes (like the Euclid satellite or the DESI survey) might be able to detect these "resonant" signals. If we find them, it would be direct proof that the early universe was a dynamic, wobbly place, and it would reveal the existence of a hidden "heavy sector" of particles.

In short: The authors built a new mathematical telescope. Instead of just looking at the static picture of the universe, they learned how to listen to the "echoes" of its past wobbles, allowing us to hear the heavy, invisible particles that shaped our cosmos.

Technical Summary

1. Problem Statement

The Cosmological Bootstrap program aims to reconstruct inflationary correlators (such as the power spectrum and bispectrum) using only fundamental principles like unitarity, locality, and causality, without explicitly solving bulk time integrals.

• The Limitation: Previous bootstrap successes relied heavily on exact de Sitter (dS) invariance (specifically scale invariance). In exact dS, bulk locality translates into ordinary differential equations (ODEs) for boundary correlators.
• The Challenge: Realistic inflation models often break scale invariance (e.g., via time-dependent masses or sound speeds, as seen in axion monodromy). When scale invariance is broken, the correspondence between bulk locality and boundary constraints becomes non-local. The standard ODE bootstrap fails, and the resulting equations are notoriously difficult to solve because they involve nested time integrals that cannot be easily reduced to local kinematic constraints.
• Goal: The authors seek to extend the bootstrap program to scale-breaking scenarios, specifically those involving heavy fields with time-dependent masses, to derive and solve the corresponding boundary constraints.

2. Methodology

The paper introduces a novel framework where bulk locality in a scale-breaking background manifests as Integro-Differential Equations (IDEs) on the boundary.

A. Derivation of Boundary IDEs

• Model Setup: The authors consider a heavy scalar field $\sigma$ with a time-dependent mass $m^2(t) = m_0^2 + \Delta m^2(t)$ coupled to the inflaton (Goldstone boson $\pi$).
• From Bulk to Boundary: By utilizing the Schwinger-Keldysh formalism and the bulk equation of motion for the propagator, they derive a relation where the differential operator acting on the correlator (derived from the Klein-Gordon operator) is equated to a source term plus an integral over the memory of the system.
• The Memory Kernel: The breaking of scale invariance introduces a memory kernel $K(q)$ in momentum space. The resulting equation for the exchange diagram $F$ takes the schematic form:
  $$(\hat{\Delta}_u + m_0^2) F(u; s) = \text{Source} - \int_0^u du' \, K(\dots) F(u'; s)$$
  This equation relates a correlator at a specific momentum configuration to an integral over "more squeezed" configurations ($u' < u$), effectively encoding the universe's evolution history.

B. Solving the IDEs

The authors employ a multi-pronged strategy to solve these IDEs:

1. Perturbative Analytic Solution:
   • They focus on monochromatic mass oscillations ($m^2 \propto \cos(\omega t)$), motivated by axion monodromy.
   • They expand the solution in powers of the oscillation amplitude $g^2$.
   • Key Constraints: To fix the integration constants and solve the IDEs without knowing the bulk mode functions explicitly, they impose:
     • Microcausality: The commutator of the heavy field vanishes outside the light cone. This enforces a factorization property where the exchange diagram splits into products of three-point functions (plus analytic terms) in the soft limit.
     • Analyticity: The correlators must be regular in the folded limits ($k_{12} = s$) and satisfy specific behaviors near total-energy singularities, dictated by the Bunch-Davies vacuum.
2. Non-Perturbative Resummation (Parametric Resonance):
   • The perturbative expansion reveals apparent divergences (poles) when the oscillation frequency $\omega$ approaches $2\mu$ (where $\mu$ is the mass index).
   • The authors identify these as Parametric Resonances (IR resonances). They perform a non-perturbative analysis by resumming the infinite series of mass insertions, leading to anomalous scaling exponents in the squeezed limit.
3. Numerical Bootstrap:
   • They implement the first numerical bootstrap for cosmological correlators beyond scale invariance.
   • Using the Finite Difference Method (FDM), they discretize the IDEs into a linear algebra problem ($DF = S + QF$).
   • They verify that the numerical solutions converge to the perturbative results and correctly capture the non-perturbative scaling exponents near resonances, independent of specific boundary condition choices.

3. Key Contributions

1. The Integro-Differential Bootstrap

The paper establishes that for scale-breaking models, the bootstrap equations are IDEs rather than ODEs. This provides a rigorous boundary formulation for non-scale-invariant inflation, encapsulating the "memory" of the bulk evolution in a kernel $K$.

2. Analytical Solution for Oscillatory Masses

They provide the first closed-form analytical solution for the exchange diagram with time-dependent masses at leading order ($O(g^2)$). This is achieved purely from boundary constraints (microcausality and analyticity) without performing the difficult bulk time integrals.

3. Discovery of Enhanced Cosmological Collider Signals

• Standard Signals: In standard dS, heavy field exchange signals are exponentially suppressed by the Boltzmann factor $e^{-\pi \mu}$.
• New Signals: The authors show that high-frequency mass oscillations ($\omega \gtrsim m_0$) can trigger parametric resonances.
• Result: This leads to an exponential enhancement of the bispectrum amplitude in the squeezed limit, overcoming the Boltzmann suppression. The signal takes the form of oscillatory features with frequencies determined by both the mass $\mu$ and the oscillation frequency $\omega$.

4. Non-Perturbative Resonance and Anomalous Scaling

• They demonstrate that at specific frequencies ($\omega \approx 2\mu/n$), the perturbative expansion breaks down due to IR resonances.
• By resumming these effects, they derive anomalous scaling exponents for the correlators: $F \sim u^{-1/2 + \lambda \pm i\mu}$, where $\lambda$ is a growth rate dependent on the coupling $g$.
• They also identify UV resonances (transient enhancements when the physical momentum crosses the oscillation scale) and IR resonances (persistent growth).

5. First Numerical Bootstrap in Cosmology

The paper presents a successful numerical implementation of the bootstrap program for scale-breaking correlators. This serves as a proof of concept for solving IDEs in cosmology and confirms the universality of the scaling exponents near resonances.

4. Results

• Bispectrum Formula: They derive the squeezed-limit bispectrum $B(k_1, k_2, k_3)$, showing it contains terms like:
  $$B \sim P(k_1)P(k_3) \left(\frac{k_3}{k_1}\right)^{-3/2} \cos\left[ \mu \log\left(\frac{k_1}{k_3}\right) + \omega \log(-k_3 \eta_0) \right]$$
  The amplitude of this signal is significantly enhanced compared to the scale-invariant case.
• Scaling Exponents: Near the primary resonance $\omega \approx 2\mu$, the scaling exponent acquires a real part $\lambda_1 \propto g^2$, indicating exponential growth in the squeezed limit.
• Numerical Validation: The numerical bootstrap results perfectly match the analytical predictions for the scaling exponents near resonances, validating the non-perturbative resummation approach.
• Light Fields: They extend the analysis to light fields, discovering a new anomalous scaling effect at order $O(g^4)$ analogous to the stabilization of a Kapitza pendulum, where rapid oscillations effectively modify the mass of the field.

5. Significance

• Bridging Theory and Observation: The work provides a concrete theoretical framework for interpreting "primordial features" (oscillatory deviations from power laws) in CMB and Large Scale Structure data. It predicts that these features can be much larger than previously thought due to resonance effects.
• New Computational Paradigm: It moves the bootstrap program beyond the "de Sitter lamppost," offering a powerful tool to study complex, realistic inflationary models that were previously intractable analytically.
• Efficiency: The boundary IDE approach is shown to be formally more efficient than traditional bulk methods (in-in formalism) for these specific problems, as it avoids nested time integrals and directly resums scale-breaking effects.
• Phenomenology: The prediction of exponentially enhanced cosmological collider signals suggests that future experiments (like CMB-S4, Simons Observatory, or galaxy surveys like Euclid/DESI) could detect heavy fields with masses far above the Hubble scale, provided they have time-dependent masses.

In summary, this paper successfully generalizes the cosmological bootstrap to scale-breaking scenarios, deriving integro-differential equations that encode bulk memory, solving them analytically and numerically, and uncovering a new class of enhanced, resonant cosmological collider signals.