Fragility of stealth solutions in mimetic gravity

arXiv:2606.31274 · gr-qc, hep-th · Submitted 2026-06-30 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: General Relativity and Quantum Cosmology. The day in general relativity and quantum cosmology.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Fragility of stealth solutions in mimetic gravity".

Mira: Here's the gist of the paper. We study a broad class of constrained mimetic-type extensions of general relativity with action S = R d4x sqrt(-g/2) + λ C

g, Ψ: + Lm where R is the Ricci scalar, λ is a Lagrange multiplier, C

g, Ψ: is a scalar functional of the metric and generic field content Ψ (possibly involving Ψ and its covariant derivatives) and Lm is the matter Lagrangian.

Kai: On the exactly stealth branch λ¯ = 0, any general relativity geometry can arise as a local solution provided the remaining fields Ψ¯ can be chosen to satisfy C¯ = 0 <ref:2606.31274#pg6> General relativity provides an exceptionally successful description of gravitational phenomena across a wide range of scales, from laboratory and solar system tests to strong field observations around compact objects <ref:2606.31274#pg2> At the same time, a number of open problems in gravitational physics and cosmology motivate exploring extensions of GR, including scenarios in which additional degrees of freedom are present but their effects can be suppressed in regimes where GR is well tested

1–14: <ref:2606.31274#pg3> Mimetic-type constructions provide a distinct route to such a scale-dependent behavior <ref:2606.31274#pg4> Originally introduced as a singular disformal transformation of the metric

16–18: , mimetic models can be formulated more generally as constrained extensions of GR in which a Lagrange multiplier enforces a condition on the kinetic structure of additional field(s)

19–46: <ref:2606.31274#pg5> In cosmological settings, these models are often invoked to address large-scale phenomena ranging from dark matter-like behavior

16: to late-time acceleration

34, 49: and other effective modifications

33, 35: <ref:2606.31274#pg6> The Stealth Branch and Background Solutions On the exactly stealth branch λ¯ = 0, the constrained sector drops out of the background dynamics, so, on domains where a background profile Ψ¯ satisfying C¯ = 0 exists, the theory admits the corresponding general relativity geometries as stealth solutions <ref:2606.31274#pg7> These are stealth configurations: the fields Ψ¯ may have nontrivial profiles, but they do not backreact on the metric <ref:2606.31274#pg8> The central result of this paper is that this apparently benign local branch is perturbatively subtle <ref:2606.31274#pg9> Perturbative Degeneracy and Pathological Limits The central result of this paper is that the stealth branch λ¯ = 0 is perturbatively degenerate with general relativity: the constrained sector contributes to the dynamics only through terms weighted by λ¯, which vanish on the stealth branch, while still imposing an infinite hierarchy of constraints on the fluctuations <ref:2606.31274#pg10> Consequently, the limit λ¯ → 0 is generically non-uniform, making the would-be screening perturbatively pathological If λ¯ = 0, enforcing (4.9), the constraint sector contributes no quadratic operator, S(2)c = 0 This explicitly realizes the non-uniform limit λ¯ → 0: the perturbative spectrum on the stealth branch is not continuously connected to that at arbitrarily small λ¯ ≠ 0 Explicit Realization in Scalar Fields In the scalar realization, every GR metric (including black hole geometries) admits a family of stealth configurations labeled by scalar profiles ϕ¯ solving Hamilton-Jacobi equation (3.6) For static and spherically symmetric backgrounds, the equation is separable, while for rotating black holes Carter separability allows one to construct explicit stealth Kerr profiles labeled by the conserved quantities of geodesic motion The main result of the paper is that this apparently harmless branch is perturbatively singular For nonzero λ¯, the constrained sector contributes genuine quadratic operators and interactions for the additional modes On the exactly stealth branch, however, these contributions vanish, and the constrained sector acts only through the constraint equations on the fluctuations Thus in this minimal setup, which contains no nonlinear terms in the action and therefore avoids the strong-coupling issue that typically arises in such configurations, the singular behavior clearly highlights the problem of the non-uniform limit The naive local screening limit is perturbatively pathological A Screening Radius Estimate In Eq.

Mira: (1.2), we estimated the relation between the screening radius rscreen and the background value of the Lagrange multiplier λ¯ for a compact object of mass M, using the Kretschmann scalar to set the curvature scale This gives a reasonable order-of-magnitude estimate For λ¯ = 0, the Schwarzschild metric is an exact solution of the system A nonzero λ¯ induces deviations from the Schwarzschild case Using this expression in the screening condition (A.2), we find rscreen ∼ M λ¯!1/3 Thus, in the scalar example, evaluating λ¯(r) at r ≃ rscreen gives the same scaling as the dimensional estimate (1.2) Curing Perturbative Degeneracy There are at least three qualitatively different possibilities to cure the perturbative degeneracy of the stealth branch (i) One may add operators to the term λC such as ∇µC∇µC, or more general functions F(C, ∇C,...), which preserves the stealth GR family (ii) A second possibility is to add invariants to λC that depend explicitly on ∇Ψ and/or ∇∇Ψ, which cannot in general be rewritten as functions of C therefore can modify the perturbative dynamics (iii) Finally, one may endow the multiplier with dynamics, e.g.

Kai: by adding (∇λ)2(or kinetic mixing), which can remove the stealth-branch degeneracy by producing a non-singular quadratic operator but it also introduces new degrees of freedom In this paper we studied a broad class of constrained mimetic-type extensions of GR defined by the action (1.

Paper summary: Kai: So we're looking at this paper called "Fragility of stealth solutions in mimetic gravity". Basically, they're studying a big family of modified gravity theories that try to explain why general relativity looks like general relativity in our everyday life, but maybe not on the largest scales.

Mira: Right. The core idea here is about these constrained mimetic extensions of general relativity with action S = R d4x sqrt(-g/two) R over two plus lambda C of g and psi plus Lm, where lambda is a Lagrange multiplier enforcing a condition on the kinetic structure of the extra field psi <ref:2606.31274#pg1,extensions of general relativity with action S = R>.

Kai: They focus specifically on the branch where lambda bar is zero, which they call the exactly stealth branch. On this branch, they claim that any general relativity geometry can arise as a local solution if you can just pick some background profile for psi bar that satisfies a constraint C bar equals zero.

Lev: But what do we get from this setup? The paper suggests that while this looks simple on the surface because the constrained sector drops out of the background dynamics, it's actually quite fragile when you look at small fluctuations around those backgrounds.

Mira: Exactly. The central result is that this branch where lambda bar equals zero is perturbatively subtle with general relativity. They argue that even though the constrained sector doesn't show up in the equations for background dynamics, it still imposes an infinite hierarchy of constraints on the small fluctuations.

Kai: That sounds like a problem because if you take a limit where lambda bar goes to zero, it's not smooth; it becomes non-uniform. They call this behavior perturbatively pathological because the spectrum of fluctuations doesn't connect continuously to what you get when lambda is actually very small but not zero.

Lev: And that non-uniform limit means that the naive idea of screening just working smoothly as you dial down a parameter doesn't hold up in this setup. They show that when lambda bar equals zero, enforcing a condition like equation four point nine, the constraint sector contributes no quadratic operator to the fluctuations at all.

Mira: That is because it’s weighted by lambda bar, which vanishes on the stealth branch, so there is nothing left for those extra modes to do in terms of quadratic interactions. It seems like a decoupling that's too perfect.

Kai: The paper does give an estimate for how far away from zero lambda bar you have to be before you see things happen. They estimate the screening radius, rscreen, and it scales with the mass M of an object as rscreen is proportional to M lambda bar raised to the power of minus one-third.

Paper summary: Lev: So if we look at a compact object, say a black hole of mass M, and we want to see deviations from general relativity due to this sector, that deviation scales like lambda bar raised to the power of negative one-third times M.

Mira: And then they connect that scaling back to the physical curvature scale using the Kretschmann scalar, which is proportional to one over the cube of your radius. This gives us a reasonable order of magnitude estimate for when things start deviating from pure general relativity.

Kai: So for you listening, what this means is that if you're trying to model gravity with these mimetic ideas and you set the background parameter lambda bar to zero, you might get a geometry that looks like general relativity locally, but perturbatively it's not as smooth as we hope.

Lev: It’s about how the constraints handle the extra degrees of freedom when that multiplier goes to zero. The paper points out that this singularity in the limit is a big hint about where things are going wrong with this simple setup.

Mira: They suggest three ways to fix this perturbative degeneracy, first by adding terms like the gradient of C squared to the action, which keeps general relativity geometries as stealth solutions.

Kai: Or they could add invariants that depend on the gradients of psi or its higher derivatives, which aren't just functions of C and can actually change how things behave perturbatively.

Lev: And a third way is to give the multiplier itself dynamics, for example, by adding something like the square of its gradient. That removes the stealth-branch problem by creating a non-singular quadratic operator but you end up with new degrees of freedom that you have to manage on your hardware.

Mira: So they’re basically saying that this simple screening limit is problematic and we need to add more structure to the action itself to make it stable perturbatively.

Kai: It puts pressure on the authors, Alireza Allahyari and his team, who are building these models in theory, to figure out how to introduce these necessary terms without just creating a mess of new physics.

Lev: And from an engineering standpoint, if we wanted to build a real system testing this idea, we’d have to worry about those new degrees of freedom introduced by adding dynamics to the multiplier.

Mira: It’s interesting because it shows that just setting the background parameter lambda bar to zero isn't enough; you need the full structure of the theory, including how those constraints interact with fluctuations.

Kai: So this paper on "Fragility of stealth solutions in mimetic gravity" tells us that our simple picture of a clean screening limit for modified gravity might be too optimistic without proper dynamical terms.

Conclusion: Kai: So this paper is about these constrained mimetic extensions of general relativity, specifically looking at the branch where lambda bar is zero, or what they call the exactly stealth branch.

Mira: Right, so it's checking how stable those solutions are when you take that limit to zero for the multiplier.

Kai: It turns out that while a geometry can look like general relativity locally on this stealth branch, it's actually quite fragile when you look at small fluctuations around it.

Lev: That fragility comes from the fact that even though the background dynamics don't show anything, there's still an infinite hierarchy of constraints on those fluctuations.

Mira: So to sum up, they are showing that this seemingly benign local branch is perturbatively subtle with general relativity because of how the constraints behave when you set that multiplier to zero.

Kai: It means if you try to just set the background parameter lambda bar to zero for a screening radius estimate, it doesn't work smoothly; it becomes non-uniform.

Lev: That non-uniform limit is what makes the naive local screening idea problematic because the perturbative spectrum doesn't connect continuously to what happens when lambda is really small but not zero.

Mira: So the main point here is that this simple setup, without nonlinear terms in the action, clearly shows this problem with how it handles constraints in a singular limit.

Kai: And they suggest that to fix this degeneracy, you need to add specific operators to the action or give the multiplier its own dynamics so things behave better perturbatively.

Lev: I agree, adding dynamics to the multiplier introduces new degrees of freedom, which is something we have to deal with when we try running these kinds of models on real hardware.

Mira: So this paper really highlights that just setting a background parameter to zero isn't enough; you need more structure in the action itself for the theory to be stable perturbatively.

Kai: It puts pressure on the authors, Alireza Allahyari and his team, to figure out how to introduce these necessary terms without creating a messy situation.

Lev: And from an engineering standpoint, if we wanted to build a real system testing this idea, we’d have to worry about those new degrees of freedom introduced by adding dynamics to the multiplier.

Alireza Allahyari, Clarisse Donio, Mohammad Ali Gorji, Zahra Gorji, Javad T. Firouzjaee

Department of Astronomy and High Energy Physics, Kharazmi University · Magistère de Physique Fondamentale d’ Orsay, University Paris-Saclay · Cosmology, Gravity, and Astroparticle Physics Group, Center for Theoretical Physics of the Universe Institute for Basic Science · Department of Physics, K.N. Toosi University of Technology

gr-qc, hep-th

Submitted: 2026-06-30

Updated: 2026-11-16

Comments: 11+8 pages, version to appear in JCAP

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 77/100

The gist: We study a broad class of constrained mimetic-type extensions of general relativity with action S = R d4x sqrt(-g/2) + λ C[g, Ψ] + Lm where R is the Ricci scalar, λ is a Lagrange multiplier…

Key concepts

Mimetic Gravity
This involves extending general relativity by adding fields ($\Psi$) and a constraint term involving a Lagrange multiplier ($\lambda$). This structure is used to explore modified gravity theories where extra degrees of freedom might be present but suppressed in certain regimes, often aiming for scale-dependent effects.
Stealth Branch
This is the specific case where the Lagrange multiplier $\lambda$ is exactly zero ($\bar{\lambda}=0$). In this branch, the dynamics of the extra fields ($\Psi$) do not affect the metric (the geometry), meaning GR solutions can appear as 'stealth' configurations even if $\Psi$ has non-trivial profiles.
Perturbative Degeneracy
This means that when $\lambda$ is infinitesimally small, the dynamics of the system on this stealth branch do not smoothly connect to standard general relativity. The constraint sector vanishes in a way that creates an infinite hierarchy of constraints on fluctuations, making the limit $\lambda \to 0$ non-uniform and perturbatively pathological.
Screening Radius Estimate
This calculation relates the characteristic size of a screening effect ($r_{screen}$) to the background value of the Lagrange multiplier ($\bar{\lambda}$). For $\bar{\lambda}=0$, this estimate shows that $r_{screen}$ scales as $M |\lambda|!^{1/3}$. This demonstrates how deviations from pure GR are related to the magnitude of the multiplier.

Terminology

Summary

We study a broad class of constrained mimetic-type extensions of general relativity with action S = R d4x sqrt(-g/2) + λ C[g, Ψ] + Lm where R is the Ricci scalar, λ is a Lagrange multiplier, C[g, Ψ] is a scalar functional of the metric and generic field content Ψ (possibly involving Ψ and its covariant derivatives) and Lm is the matter Lagrangian. p. 0 On the exactly stealth branch λ¯ = 0, any general relativity geometry can arise as a local solution provided the remaining fields Ψ¯ can be chosen to satisfy C¯ = 0 <ref:2606. p. 231274#pg6>

Introduction to Mimetic Gravity and Screening

General relativity provides an exceptionally successful description of gravitational phenomena across a wide range of scales, from laboratory and solar system tests to strong field observations around compact objects <ref:2606.31274#pg2> At the same time, a number of open problems in gravitational physics and cosmology motivate exploring extensions of GR, including scenarios in which additional degrees of freedom are present but their effects can be suppressed in regimes where GR is well tested [1–14] <ref:2606. p. 131274#pg3> Mimetic-type constructions provide a distinct route to such a scale-dependent behavior <ref:2606. p. 131274#pg4> Originally introduced as a singular disformal transformation of the metric [16–18], mimetic models can be formulated more generally as constrained extensions of GR in which a Lagrange multiplier enforces a condition on the kinetic structure of additional field(s) [19–46] <ref:2606. p. 131274#pg5> In cosmological settings, these models are often invoked to address large-scale phenomena ranging from dark matter-like behavior [16] to late-time acceleration [34, 49] and other effective modifications [33, 35] <ref:2606. p. 131274#pg6>

The Stealth Branch and Background Solutions

On the exactly stealth branch λ¯ = 0, the constrained sector drops out of the background dynamics, so, on domains where a background profile Ψ¯ satisfying C¯ = 0 exists, the theory admits the corresponding general relativity geometries as stealth solutions <ref:2606.31274#pg7> These are stealth configurations: the fields Ψ¯ may have nontrivial profiles, but they do not backreact on the metric <ref:2606. p. 231274#pg8> The central result of this paper is that this apparently benign local branch is perturbatively subtle <ref:2606. p. 231274#pg9>

Perturbative Degeneracy and Pathological Limits

The central result of this paper is that the stealth branch λ¯ = 0 is perturbatively degenerate with general relativity: the constrained sector contributes to the dynamics only through terms weighted by λ¯, which vanish on the stealth branch, while still imposing an infinite hierarchy of constraints on the fluctuations <ref:2606.31274#pg10> Consequently, the limit λ¯ → 0 is generically non-uniform, making the would-be screening perturbatively pathological If λ¯ = 0, enforcing (4. p. 09), the constraint sector contributes no quadratic operator, S(2)c = 0 This explicitly realizes the non-uniform limit λ¯ → 0: the perturbative spectrum on the stealth branch is not continuously connected to that at arbitrarily small λ¯ ≠ 0

Explicit Realization in Scalar Fields

In the scalar realization, every GR metric (including black hole geometries) admits a family of stealth configurations labeled by scalar profiles ϕ¯ solving Hamilton-Jacobi equation (3.6) For static and spherically symmetric backgrounds, the equation is separable, while for rotating black holes Carter separability allows one to construct explicit stealth Kerr profiles labeled by the conserved quantities of geodesic motion

Conclusion on Perturbative Behavior

The main result of the paper is that this apparently harmless branch is perturbatively singular For nonzero λ¯, the constrained sector contributes genuine quadratic operators and interactions for the additional modes On the exactly stealth branch, however, these contributions vanish, and the constrained sector acts only through the constraint equations on the fluctuations Thus in this minimal setup, which contains no nonlinear terms in the action and therefore avoids the strong-coupling issue that typically arises in such configurations, the singular behavior clearly highlights the problem of the non-uniform limit The naive local screening limit is perturbatively pathological

A Screening Radius Estimate

In Eq. (1.2), we estimated the relation between the screening radius rscreen and the background value of the Lagrange multiplier λ¯ for a compact object of mass M, using the Kretschmann scalar to set the curvature scale This gives a reasonable order-of-magnitude estimate For λ¯ = 0, the Schwarzschild metric is an exact solution of the system A nonzero λ¯ induces deviations from the Schwarzschild case Using this expression in the screening condition (A. p. 22), we find rscreen ∼ M λ¯! p. 21/3 Thus, in the scalar example, evaluating λ¯(r) at r ≃ rscreen gives the same scaling as the dimensional estimate (1. p. 22)

Curing Perturbative Degeneracy

There are at least three qualitatively different possibilities to cure the perturbative degeneracy of the stealth branch (i) One may add operators to the term λC such as ∇µC∇µC, or more general functions F(C, ∇C,...), which preserves the stealth GR family (ii) A second possibility is to add invariants to λC that depend explicitly on ∇Ψ and/or ∇∇Ψ, which cannot in general be rewritten as functions of C therefore can modify the perturbative dynamics (iii) Finally, one may endow the multiplier with dynamics, e. p. 8g. by adding (∇λ)2(or kinetic mixing), which can remove the stealth-branch degeneracy by producing a non-singular quadratic operator but it also introduces new degrees of freedom

Summary and Conclusions

In this paper we studied a broad class of constrained mimetic-type extensions of GR defined by the action (1.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can perform non-uniform limit analysis by explicitly distinguishing between exactly zero background multiplier lambda bar=0 and arbitrarily small but nonzero lambda bar not equal to 0. This allows the AI to predict qualitatively different perturbation spectra when approaching the stealth branch from slightly perturbed non-stealth configurations.

  2. Improved AI can identify structural limitations of current theoretical extensions by recognizing that adding operators solely built from C and its covariant derivatives does not cure the degeneracy because these terms do not generate a genuine quadratic operator for the would-be extra mode on the stealth branch.

  3. Improved AI can evaluate the viability of modified gravity theories by assessing whether proposed modifications preserve the stealth GR family, noting that C-only deformations keep GR solutions but remain degenerate while curing pathology requires more drastic modifications.

Sources

Related papers