The Chameleon Universe: How Gravity Learns to Hide ✛ ✎ ✦

Original: "Exploring Hu-Sawicki-like modified gravity with Genetic Algorithms"
arXiv:2607.12745 · 2026-07-14 · arXiv non-exclusive · independent analysis · 3 min
DESI and Pantheon+ data allow only hundred-millionth deviations from Einstein’s cosmological constant: modified gravity is almost indistinguishable from the standard model.

In the depths of space, gravity is mastering the art of disguise. For a quarter century, astrophysicists have been trying to decipher the nature of dark energy, which makes the Universe expand at an accelerating rate. Candidates vying for the role include Einstein’s cosmological constant, scalar fields, and modifications to gravity. Among the latter are f(R) theories: here the force of gravity changes depending on the curvature of space, as if gravity itself decides how noticeable it wants to be. But to survive the onslaught of observational tests, a theory must know how to hide, and the best camouflagers pass through natural selection.

It is precisely this kind of selection that the authors of the new study arranged. They applied genetic algorithms — an evolutionary method in the world of mathematics. Random mutations and crossovers spawn generations of functions; ruthless selection keeps only those that are stable and consistent with observations.

Genetic algorithms in cosmology are like a breeder developing a new breed. But instead of dogs or roses, it’s variants of the geometry of the Universe. Functions crossbreed, mutate; in each generation, only those survive that not only agree with the data but also avoid mathematical catastrophes, such as tachyonic instabilities.

The testing ground included two arenas. The first is supernovae from the Pantheon+ survey. The second is baryon acoustic oscillations imprinted in the galaxy distribution by the DESI collaboration at various redshifts. For each function, the Friedmann–Lemaître equations, the heirs to Einstein’s equations on the cosmological stage, were solved numerically, and the evolution of the Universe’s expansion was traced. The models were asked to do the nearly impossible: a dynamic dark energy that remains indistinguishable from ΛCDM at laboratory and astrophysical scales. That’s where the chameleon screening mechanism comes into play, suppressing any fifth force.

The result of the numerical simulations is elegant and disconcerting. The winning function F(R) deviates from the Einsteinian one by hundred-millionths. The effective dark energy equation of state $w_{eff}$ trembles at -1 with an amplitude of about 10⁻⁷ — almost a perfect cosmological constant. The Om(z) parameter, immutable in ΛCDM, varies by no more than 0.1% even at small redshifts. The statistical advantage over the standard model is negligible: χ² decreases by a few units out of 1418.9. The cosmos whispers: 'No tricks.'

The chameleon mechanism explains this elusiveness. In dense environments, be it a laboratory or a galaxy cluster, the extra 'fifth force' switches off, and the theory reduces to General Relativity. To notice deviations, you need cosmological voids and the latest epochs. That is where next-generation telescopes are aimed.

What’s next? So far we have studied the background — the overall rate of the Universe’s expansion. The real drama unfolds in perturbations: how galaxy clusters are born, how light rays are bent. There, modified gravity promises to be more noticeable. Upcoming missions Euclid and the Nancy Grace Roman Space Telescope will attempt to catch the slightest cracks in the disguise. If dark energy is dynamic, it will be betrayed not by the overall expansion rate, but by the fine structure of the cosmic fabric. For now, the chameleon remains almost invisible, and that is its perfection. But this near-invisibility is an invitation for quantum theories that could fill the microscopic gaps.

Genetic algorithms mimic biological evolution: mutations and crossovers spawn generations of functions; only those best adapted to the data and physical constraints survive. In this work, the winners turned out almost indistinguishable from the cosmological constant.

F(R) = -\frac{2\Lambda}{1+\left(\frac{b\Lambda}{R}\right)^n}
Parameterization of the deviation from GR in the Hu–Sawicki model; when b→0 it recovers ΛCDM.
w_{\rm eff} = -1 - \frac{1}{3}\frac{y_H'}{y_H}
Effective equation of state for dark energy; w_eff=-1 corresponds to the cosmological constant.
Scientists
Fritz ZwickyVera RubinJacob BekensteinKarl SchwarzschildJoseph WeberAdam Riess
Concepts
fr gravity genetic algorithm dark matter modified gravity Generative Adversarial Networks modified gravity vs dark matter