⚛️ general relativity

Anti-Ultralocality and Plateau Models of Inflation

This paper demonstrates that plateau-shaped inflaton potentials, often favored for their low tensor-to-scalar ratios, are particularly susceptible to anti-ultralocality effects which, starting from generic initial conditions, either prevent sufficient inflation or trigger quantum runaway, thereby imposing increasingly severe fine-tuning requirements as the inflation energy scale decreases.

Original authors: Joshua Shterenberg, David Garfinkle, Anna I. Rosenzweig, David Shlivko, Paul J. Steinhardt

Published 2026-08-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Joshua Shterenberg, David Garfinkle, Anna I. Rosenzweig, David Shlivko, Paul J. Steinhardt

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The story of our universe begins with a question that has haunted cosmologists for decades: how did the cosmos become so remarkably smooth, flat, and uniform? When we look out at the vast expanse of space, we see a universe that is nearly identical in every direction, with matter spread out evenly on the largest scales. This state of order is puzzling because the standard model of the Big Bang suggests that the universe should have started as a chaotic, lumpy mess, filled with wild fluctuations in density and temperature. To explain this transition from chaos to order, physicists proposed a theory called inflation. This theory suggests that in the very first fraction of a second after the Big Bang, the universe underwent a period of incredibly rapid expansion, stretching out any initial wrinkles and smoothing everything out, much like blowing up a balloon to make its surface appear flat.

For this theory to work, the universe must have been driven by a specific type of energy field, often called the inflaton. The shape of the energy landscape for this field determines how the universe expands. For many years, scientists focused on simple, steep energy landscapes, but observations of the cosmic microwave background—the afterglow of the Big Bang—have ruled those out. Instead, the data points toward a "plateau" model, where the energy landscape is incredibly flat for a long stretch before dropping off. This flatness is necessary to match the specific patterns we see in the sky today. However, a new study suggests that while this plateau model fits the data we observe, it may be fundamentally unable to start the process of smoothing the universe in the first place.

A team of researchers, led by physicists at Princeton University and other institutions, set out to test whether these popular plateau models could actually work under realistic conditions. They did not rely on simplified equations or idealized assumptions. Instead, they used powerful supercomputers to run complex numerical simulations that tracked the evolution of space and time from the moment the universe emerged from the quantum chaos of the Big Bang. Their goal was to see if the universe could naturally evolve from a messy, irregular state into the smooth, flat state required for inflation to take hold. They specifically looked at a phenomenon known as "anti-ultralocality," a counter-intuitive effect where the unevenness of space, rather than being smoothed out immediately, actually grows stronger and faster than the forces trying to expand the universe.

In their simulations, the researchers started with a universe that was intentionally rough and uneven, representing the chaotic state expected right after the Big Bang. They then watched how the universe evolved under the rules of general relativity and the specific plateau energy model. They found that the universe struggled to get started. Because the energy landscape was so flat, the forces driving the expansion were weak compared to the growing unevenness of space. As the universe expanded, the differences in density and curvature between different regions grew larger and larger, overwhelming the gentle push of inflation. This growth of unevenness prevented the universe from ever reaching the smooth, flat state needed to begin the long period of accelerated expansion. In many of their simulations, the universe simply could not smooth itself out enough to last for the required duration, which is about sixty cycles of doubling in size, a number needed to explain the universe we see today.

The researchers discovered a second, even more problematic failure mode. To avoid the first problem, they tried starting the simulations with a universe that was already very smooth and with the inflaton field positioned high up on the plateau. While this allowed the universe to expand for long enough, it triggered a different disaster. In these flat regions, the field became so sensitive that tiny quantum jitters took over, causing the field to fluctuate wildly rather than rolling smoothly down the hill. This led to a state of "quantum runaway," where the universe would split into an infinite number of different outcomes, a scenario often called a multiverse. In such a state, the laws of physics as we know them break down, and it becomes impossible to make any specific predictions about what the universe should look like. Essentially, the very feature that makes the plateau model fit current observations—the extreme flatness of the energy landscape—is the same feature that makes it impossible for the universe to start inflating from a realistic, messy beginning without falling into chaos or becoming unpredictable.

The team concluded that for the plateau model to work, the initial conditions of the universe would have to be impossibly perfect. The universe would need to start out already incredibly smooth and flat, far more so than what we expect from a chaotic Big Bang beginning. In fact, the level of smoothness required is so extreme that it defeats the purpose of inflation, which was originally proposed to explain how the universe became smooth in the first place. If the universe had to be this smooth to begin with, then inflation is not doing the work of smoothing it out; it is merely preserving a state that was already there. The study suggests that the current favorite models of inflation, which rely on these flat plateau potentials, may be unable to solve the fundamental problem of how our universe got its start. Instead, the researchers propose that the energy scale of inflation might need to be higher, or a completely different mechanism might be required to explain the uniformity of the cosmos, challenging the prevailing wisdom in cosmology.

Technical Summary: Anti-Ultralocality and Plateau Models of Inflation

Problem Statement
The paper addresses the "homogeneity problem" in cosmology: whether inflation can naturally emerge from generic initial conditions expected after a quantum gravity-dominated Big Bang. While inflation was proposed to smooth and flatten the universe, previous numerical relativity studies demonstrated that simple power-law inflaton potentials (e.g., $V \propto \phi^2$ or $\lambda\phi^4$ ) fail to do so. This failure is attributed to "anti-ultralocality," a phenomenon in decelerated expansion where spatial gradient terms in the Einstein-scalar field equations grow faster than velocity terms, driving the universe away from a homogeneous, flat Friedmann-Robertson-Walker (FRW) state.

The authors investigate whether "plateau" inflation models—specifically the Starobinsky and Higgs inflation forms, $V(\phi) = V_0(1 - e^{-\gamma\phi})^2$ —are more robust. These models are currently favored because they predict a tensor-to-scalar ratio ( $r$ ) consistent with observational upper bounds ( $r \approx 0.003$ ). However, the authors hypothesize that the extreme flatness of the plateau and the vast energy density gap ( $\sim 10$ orders of magnitude) between the Planck scale and the inflationary scale make these models uniquely vulnerable to anti-ultralocality effects.

Methodology
The authors employ a validated (3+1) numerical relativity scheme based on orthonormal tetrads and constant-mean-curvature (CMC) time slicing. They evolve the coupled Einstein-scalar field equations in a periodic simulation box, initially allowing spatial variations in one dimension for computational efficiency.

To objectively assess viability, the authors introduce a two-part Testing Protocol:

Initial State Test: Verifies that initial conditions are "non-perturbatively far" from flat FRW, mimicking the chaotic state expected post-Big Bang. This is quantified using gauge-invariant measures of the Weyl scalar ( $\hat{C}$ ) and Chern-Pontryagin invariant ( $\hat{P}$ ), requiring $\hat{C}, \hat{P} \gtrsim O(1)$ . It also ensures spatial variations occur on scales smaller than the initial mean curvature radius.
Successful Smoothing Test: Verifies that the resulting evolution:
- Achieves at least 60 e-folds of accelerated expansion after the universe has smoothed sufficiently (defined as $\hat{C}, \hat{P} < 10^{-10}$ ).
- Avoids "quantum runaway," a regime where quantum fluctuations dominate classical evolution ( $B_{multi} > 1$ ), leading to a multiverse of outcomes and the breakdown of the semi-classical approximation.

Key Results
The numerical simulations of the plateau model reveal two distinct failure modes when starting from generic initial conditions satisfying the Initial State Test:

Failure Mode 1: Insufficient Smoothing. Due to anti-ultralocality, spatial gradients grow rapidly during the initial decelerated phase. This growth delays the onset of inflation and causes the inflaton field ( $\phi$ ) to roll down the potential significantly before inflation begins. Consequently, by the time the universe smooths enough ( $\hat{C}, \hat{P} < 10^{-10}$ ) to support a scale-invariant spectrum, the field value $\phi$ has dropped below the threshold required to sustain 60 e-folds of inflation. The model fails to produce the necessary duration of inflation.
Failure Mode 2: Quantum Runaway and Multiverse. To avoid Failure Mode 1, one might attempt to start with a higher initial field value ( $\phi_0$ ) and lower velocity. However, the combination of anti-ultralocality and the extreme flatness of the plateau leads to a different outcome. As inflation begins, the Hubble parameter $H$ stabilizes while the field velocity $\dot{\phi}$ continues to decrease. This causes the ratio $B_{multi} \propto H^4/\dot{\phi}^2$ to exceed unity, triggering quantum runaway. In this regime, the semi-classical approximation breaks down, and the system enters a self-reproduction phase (eternal inflation) where the outcome is a multiverse of possibilities, rendering specific predictions for our observable universe impossible.

Fine-Tuning Requirements
The authors find that avoiding both failure modes simultaneously requires initial conditions that are exponentially fine-tuned. Specifically, to achieve a tensor-to-scalar ratio consistent with observations ( $V_0 \approx 10^{-10} M_{Pl}^4$ ), the initial inhomogeneity parameters $\hat{C}_0$ and $\hat{P}_0$ must be less than $10^{-8}$ . This level of smoothness and flatness is comparable to the state inflation is supposed to produce, not the state expected to precede it. As the inflationary energy scale is lowered to match observational constraints, the required fine-tuning becomes more severe.

Significance and Claims
The paper concludes that plateau inflation models, despite their observational success in predicting low $r$ , are likely unviable as a dynamical mechanism to solve the homogeneity and flatness problems of the standard Big Bang. The authors argue that:

Generic initial conditions lead to either insufficient inflation or a breakdown of predictivity via quantum runaway.
The success of these models relies on initial conditions that are already homogeneous and isotropic to an extraordinary degree, effectively negating the need for inflation to smooth the universe.
There is a critical tension: lowering the inflationary energy scale to satisfy CMB observations exacerbates anti-ultralocality effects, requiring increasingly non-generic initial conditions.

The authors suggest that if inflation is to succeed, a mechanism other than the standard semi-classical evolution of a scalar field may be required to explain the initial homogeneity of the universe. They do not propose a new model but rather highlight a fundamental tension in current inflationary paradigms when subjected to rigorous numerical relativity tests.

More like this