Papers
Topics
Authors
Recent
Search
2000 character limit reached

Logarithmic Graviton in Critical Gravity

Updated 5 July 2026
  • Logarithmic gravitons are generalized spin‑2 modes exhibiting logarithmic radial behavior arising at criticality in AdS3 gravity.
  • They emerge when massive and massless graviton modes degenerate, forming a rank‑2 Jordan cell that underpins LCFT structures.
  • Similar logarithmic features appear in four-dimensional soft graviton amplitudes and inflationary models, expanding their role in critical gravity.

Searching arXiv for relevant papers on logarithmic gravitons in chiral TMG, LCFT structure, and related critical gravity context. The term logarithmic graviton most commonly denotes a special linearized spin‑2 mode that appears when the graviton spectrum of a critical gravity theory becomes non‑diagonalisable, so that a new solution acquires logarithmic dependence in the radial coordinate and is paired with an ordinary graviton in a rank‑2 Jordan cell. In three-dimensional anti‑de Sitter gravity, this usage is standard at the chiral point of Topologically Massive Gravity (TMG), where the logarithmic graviton is the bulk counterpart of a logarithmic partner in a conjectured boundary logarithmic conformal field theory (LCFT) (Mvondo-She, 8 May 2026). In distinct four-dimensional literatures, the same phrase is also used for soft-graviton logarithms in scattering and for graviton-loop–induced secular logarithms during inflation; these uses concern logarithmic terms in amplitudes or effective fields rather than a new bulk spin‑2 mode (Laddha et al., 2018).

1. Definition and terminological scope

In the AdS3_3 and critical-gravity literature, the logarithmic graviton is a gravitational fluctuation whose falloff near the AdS boundary contains logarithms, and which, in the dual CFT, corresponds to a logarithmic partner of the stress tensor, forming a Jordan block for L0L_0 (Gaberdiel et al., 2010). At a critical point, the usual massive and massless graviton branches degenerate, the linearized operator ceases to be diagonalisable, and a generalized eigenmode appears. In this sense the logarithmic graviton is not an independent particle species at generic couplings, but a critical mode tied to spectral degeneracy.

The term is not fully uniform across subfields. In one four-dimensional usage, “logarithmic graviton” refers to logarithmic terms in the soft-graviton expansion that produce a late-time tail in the gravitational waveform. In another, it refers to secular logarithms generated by graviton loops on de Sitter or inflationary backgrounds. These usages share the presence of logarithms but not the representation-theoretic meaning of the AdS3_3 mode (Agrawal et al., 2023).

2. Emergence in chiral Topologically Massive Gravity

Cosmological TMG in three dimensions supplements the Einstein–Hilbert action with a gravitational Chern–Simons term. Linearized excitations around AdS3\mathrm{AdS}_3 organize into

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.

At the chiral point

μ=1,\mu \ell = 1,

the left-moving central charge vanishes,

cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},

and the left-moving massive graviton degenerates with a left-moving boundary graviton. The logarithmic graviton then appears as the generalized eigenmode produced by this degeneracy (Mvondo-She, 8 May 2026).

For metric perturbations

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},

the logarithmic graviton can be constructed as the derivative of the massive branch at the point of degeneracy,

ψlog=ψM(μ)(μ)μ=1.\psi^{\log}=\left.\frac{\partial \psi^M(\mu\ell)}{\partial(\mu\ell)}\right|_{\mu\ell=1}.

In Fefferman–Graham coordinates with r=eρr=e^\rho, the ordinary left-moving graviton behaves as

L0L_00

whereas the logarithmic graviton behaves as

L0L_01

This logarithmic falloff is the defining bulk signature of the mode.

A closely related description uses the logarithmic graviton of Grumiller and Johansson at the chiral point in global AdSL0L_02. There the mode takes the form

L0L_03

and near the AdSL0L_04 boundary,

L0L_05

This makes explicit that the logarithmic coefficient contains both radial and temporal linear pieces (Mvondo-She, 15 Jun 2026).

3. Jordan cells, Virasoro flow, and radial monodromy

On the boundary, the left-moving graviton L0L_06 is a highest-weight state satisfying

L0L_07

Its logarithmic partner is not an eigenstate of L0L_08; instead,

L0L_09

Equivalently,

3_30

On the span of 3_31, 3_32 takes Jordan form,

3_33

This is the standard LCFT structure: 3_34 is the logarithmic partner of 3_35 (Mvondo-She, 8 May 2026).

Exponentiating 3_36 gives a one-parameter dilatation flow

3_37

so that

3_38

The extra linear term in 3_39 is the algebraic origin of logarithmic growth, with AdS3\mathrm{AdS}_30 in the bulk radial picture.

A central geometric identification is that the LCFT Jordan structure is equivalent to unipotent radial monodromy. Because

AdS3\mathrm{AdS}_31

analytic continuation

AdS3\mathrm{AdS}_32

induces

AdS3\mathrm{AdS}_33

while AdS3\mathrm{AdS}_34 remains single-valued. In the basis AdS3\mathrm{AdS}_35, the monodromy is

AdS3\mathrm{AdS}_36

This identifies the nilpotent part of AdS3\mathrm{AdS}_37 with a bulk monodromy operator (Mvondo-She, 8 May 2026).

The same structure can be formulated as a complexified Virasoro flow. With

AdS3\mathrm{AdS}_38

one has

AdS3\mathrm{AdS}_39

and therefore

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.0

Under

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.1

the logarithmic partner transforms as

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.2

This presents radial evolution, temporal evolution, Jordan mixing, and logarithmic monodromy as aspects of a single complex flow parameter SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.3 (Mvondo-She, 15 Jun 2026).

4. Descendants, modules, and LCFT evidence

The logarithmic structure is not confined to the primary level. For the global left-moving conformal algebra generated by SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.4, descendants are defined by

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.5

Using

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.6

one finds

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.7

and

SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.8

Thus each level SL(2,R)L×SL(2,R)R.SL(2,\mathbb{R})_L \times SL(2,\mathbb{R})_R.9 furnishes a rank‑two Jordan cell, and the whole μ=1,\mu \ell = 1,0 descendant tower forms a reducible but indecomposable logarithmic module. The same unipotent monodromy acts uniformly at every level (Mvondo-She, 8 May 2026).

This module structure is stated to coincide with the logarithmic graviton module previously obtained in linearized analyses of chiral TMG by Grumiller & Johansson and by Skenderis et al. The identification is at the level of primary relations, descendants, asymptotics, and indecomposable module structure. A principal claim is that monodromy-compatible Virasoro flow uniquely reconstructs the full indecomposable logarithmic module, including all descendant levels (Mvondo-She, 8 May 2026).

Independent evidence for the LCFT interpretation comes from the one-loop partition function of Euclidean TMG. At the chiral point μ=1,\mu \ell = 1,1, the graviton one-loop partition function does not factorize holomorphically and takes the form

μ=1,\mu \ell = 1,2

This has the structure expected from a logarithmic CFT, with the massive determinant supplying the sector whose single-particle content matches a field of weights μ=1,\mu \ell = 1,3 and its descendants. In this interpretation, the ordinary left-moving stress tensor μ=1,\mu \ell = 1,4 and its logarithmic partner μ=1,\mu \ell = 1,5 satisfy

μ=1,\mu \ell = 1,6

which is precisely the boundary Jordan structure associated with the bulk logarithmic graviton (Gaberdiel et al., 2010).

5. Generalizations in critical gravities

The appearance of logarithmic gravitons is not confined to chiral TMG. In higher-dimensional curvature-squared gravities at critical points, the massive and massless spin‑2 roots can also coincide, and new logarithmic solutions arise. In four dimensions, for AdS wave profiles μ=1,\mu \ell = 1,7, the critical theory develops multiple roots and the general solution acquires logarithmic terms,

μ=1,\mu \ell = 1,8

In general μ=1,\mu \ell = 1,9,

cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},0

These modes suggest a gravity description of logarithmic CFTs in cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},1 dimensions (Alishahiha et al., 2011).

A detailed mode analysis in critical gravity linearized about AdS classifies logarithmic solutions into spin 2 and Proca sectors. Spin‑2 log modes are those for which the linearized Einstein tensor is a non-gauge solution of the linearized Einstein equations, whereas for Proca modes the linearized Einstein tensor takes the form of a linearized general coordinate transformation. In AdScL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},2, a representative logarithmic factor is

cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},3

The corresponding states form Jordan cells for the AdS energy generator, which suggests a holographically dual logarithmic conformal field theory (Bergshoeff et al., 2011).

In three-dimensional flat spacetime, generalized massive gravity has a critical point where two massive graviton modes degenerate. The resulting massive logarithmic graviton is

cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},4

and the pair cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},5 forms a Jordan block under the Galilean conformal generator cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},6. This is the flat-space analogue of the AdScL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},7 logarithmic graviton (Kim et al., 2013).

Zwei-Dreibein Gravity (ZDG) exhibits critical points in AdScL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},8 where massive graviton modes coincide with pure gauge modes and new logarithmic modes appear, both in linearized analysis and in exact AdS wave solutions of the full nonlinear theory. For the AdS wave profile cL=0,cR=3G,c_L = 0,\qquad c_R = \frac{3\ell}{G},9, the critical point gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},0 yields

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},1

and the conjectured dual LCFT has vanishing central charges but nonzero new anomalies

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},2

This generalizes the New Massive Gravity critical point and embeds logarithmic AdS waves into a bigravity framework (Bergshoeff et al., 2014).

6. Distinct four-dimensional usages

In a separate line of research on soft theorems, “logarithmic graviton” refers to soft gravitons whose amplitude acquires logarithmic dependence on frequency in four dimensions. The logarithmic term in the soft expansion produces a late-time waveform tail,

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},3

so that logarithms in frequency correspond to inverse-time tails in the classical waveform. This tail supplements but is distinct from linear memory. For purely massless final states, the coefficient gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},4 vanishes at this order (Laddha et al., 2018). A symmetry-based formulation derives the same gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},5 soft corrections from superrotation Ward identities, using an enlarged radiative phase space that includes gravitational tails (Agrawal et al., 2023).

In inflationary and de Sitter settings, “logarithmic graviton effects” denotes secular logarithms generated by graviton loops. Examples include the photon field strength correction

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},6

and the Coulomb potential

gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},7

In this usage, the logarithms are attributed to inflationary gravitons and are discussed in terms of curvature-dependent field-strength renormalization, renormalization-group flow, or stochastic resummation (Glavan et al., 2023). Related work studies large temporal and spatial logarithms for massless, minimally coupled scalars, and later argues that the large logarithm in the exchange potential is tied to a gauge-independent coefficient once source and observer correlations are included (Glavan et al., 2021, Glavan et al., 2024).

A further contrast is provided by a massless, conformally coupled scalar on de Sitter. There, with the “obvious” choice for the finite part of the gμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},8 counterterm, one-loop graviton corrections yield neither large infrared logarithms in the mode functions nor large-distance logarithmic growth in the point-source response; instead one finds a decaying logarithmic correction to the mode function and short-distance logarithmic running of the potential (Glavan et al., 2020).

In this broader usage, therefore, “logarithmic graviton” no longer denotes a Jordan-cell spin‑2 mode. It denotes logarithmic soft terms, logarithmic tails, or graviton-loop–induced secular logarithms. The AdSgμν=gˉμν+hμν,g_{\mu\nu}=\bar g_{\mu\nu}+h_{\mu\nu},9 usage remains the representation-theoretic one: a generalized eigenmode of the graviton sector at a critical point, with logarithmic radial behavior and an indecomposable LCFT interpretation (Mvondo-She, 8 May 2026).

Topic to Video (Beta)

No one has generated a video about this topic yet.

Whiteboard

No one has generated a whiteboard explanation for this topic yet.

Follow Topic

Get notified by email when new papers are published related to Logarithmic Graviton.