PHYSICAL REVIEW LETTERS 129, 201604 (2022)

Root-TT̄ Deformations in Two-Dimensional Quantum Field Theories

*
Christian Ferko
Center for Quantum Mathematics and Physics (QMAP), Department of Physics and Astronomy,
University of California, Davis, California 95616, USA
†
Alessandro Sfondrini
Dipartimento di Fisica e Astronomia, Università degli Studi di Padova, Istituto Nazionale di Fisica Nucleare,
Sezione di Padova, via Marzolo 8, 35131 Padova, Italy
and Institute for Advanced Study Einstein Drive, Princeton, New Jersey, 08540 USA
‡ §
Liam Smith and Gabriele Tartaglino-Mazzucchelli
School of Mathematics and Physics, University of Queensland, St Lucia, Brisbane, Queensland 4072, Australia
(Received 27 June 2022; accepted 3 October 2022; published 10 November 2022)

In this Letter, we introduce a one-parameter deformation of two-dimensional quantum field theories

generated by a nonanalytic operator that we call Root-T T̄. For a conformal field theory, the operator
coincides with the square root of the T T̄ operator. More generally, the operator is defined so that classically
it is marginal and generates a flow that commutes with the T T̄ flow. Intriguingly, the Root-T T̄ flow is
closely related to the ModMax theory recently constructed by Bandos, Lechner, Sorokin, and Townsend.

DOI: 10.1103/PhysRevLett.129.201604

Introduction and summary.—The so-called “T T̄ oper- where one typically considers relevant or marginal (i.e.,
ator” is a universal composite operator defined from the conformal) perturbations. Nonetheless, the T T̄ flow has
stress-energy tensor T μν of any two-dimensional quantum many remarkable properties. Applied to a CFT, it breaks
field theory (QFT) as [1] scale invariance but preserves infinitely many commuting
Z charges, and it results in a family of integrable QFTs. If S 0
is not conformal, but integrable, the flow still preserves
O¼ d2 zOðz; z̄Þ; ð1aÞ
integrability. In fact, the flow is so well-behaved that it is
possible to express the finite-volume spectrum of the
1 deformed theory S λ in closed form starting from the
Oðz; z̄Þ ¼ −Det½T μν  ¼ ðT μν T μν − T μμ T νν Þ: ð1bÞ
2 spectrum of the seed theory S 0. These properties allowed
for a detailed study of T T̄-deformed theories that has
In a conformal field theory (CFT) the trace of the stress- revealed and is revealing many surprising connections to
energy tensor vanishes and we have, in light-cone coor- string theory [3–5], holography [6–8], two-dimensional
dinates, Oðz; z̄Þ ¼ TðzÞT̄ðz̄Þ. This is, however, not the case gravity [9], geometry [10], random geometry [11], and
for a generic QFT, making “T T̄” somewhat of a misnomer. more—see Refs. [12,13] for reviews and references on T T̄.
Given a two-dimensional QFT defined by a “seed” action In this Letter we propose a new “square-root,” T T̄-like
S 0 , the operator O can be used to define a one-parameter deformation. By this we mean a “marginal” deformation
family of theories S λ [2,3]: defined by ∂γ S γ ¼ Rγ, which, when applied to a two-
dimensional conformal field theory (CFT2 ) S 0 , amounts to
∂λ S λ ¼ Oλ ; S λ¼0 ¼ S 0 : ð2Þ deforming by a nonanalytic operator:
Z
Because O has scaling dimension four, the resulting flow is
R ¼ d2 z Rðz; z̄Þ; ð3aÞ
irrelevant in the sense of the renormalization group. This is
counterintuitive in the usual renormalization group picture, qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Rðz; z̄Þ ¼ TðzÞT̄ðz̄Þ ðfor a CFTÞ: ð3bÞ

Published by the American Physical Society under the terms of Despite the square root, this promises to be a well-behaved
the Creative Commons Attribution 4.0 International license.
Further distribution of this work must maintain attribution to expression in a CFT2 because the operator-product expan-
the author(s) and the published article’s title, journal citation, sion between a chiral and an antichiral field (nonanalytic as
and DOI. Funded by SCOAP3. they may be) is regular. This deformation was recently

0031-9007=22=129(20)=201604(6) 201604-1 Published by the American Physical Society

 PHYSICAL REVIEW LETTERS 129, 201604 (2022)

where gμν is the two-dimensional inverse metric (which we

take to be Minkowski), εμν is the Levi-Civita tensor, and
i ¼ 1; …; N. This ansatz encompasses many QFTs of
interest including vector models, nonlinear sigma models
(where Gij and Bij represent the target-space metric and
Kalb-Ramond fields, respectively), and Wess-Zumino-
FIG. 1. We require that first deforming S 0 by T T̄ with Oλ and
Witten models, but it is not closed under Root-T T̄ defor-
then deforming the result by Root-T T̄ with Rγ is the same as
mations. In fact, terms with a higher number of derivatives
deforming S 0 by Rγ and then by Oλ.
will appear along the flow, as it can be seen recalling that
the (Hilbert) stress-energy tensor is
considered in [14,15] to study the Bondi-Metzner-Sachs
symmetry in three dimensions of ultrarelativistic and non- ∂L
T μν ¼ −2 þ gμν L: ð7Þ
relativistic limits of a CFT2 . It is, however, not obvious how ∂gμν
such an expression should be generalized when S 0 is not
conformal. We do this by requiring that the Root-T T̄ flow To write the general ansatz for a flow driven by the operator
commutes with the ordinary T T̄ flow, i.e., that we can R, Eq. (4), we list the tensor structures that may arise from
construct a “two-parameter” action S ðλ;γÞ so that the the stress-energy tensor associated to Eq. (6) and from its
diagram in Fig. 1 commutes. As we shall explain in the powers. As gμν only appears in the first term of Eq. (6), we
conclusions, this choice will make it easier to study a vaster consider
class of theories. We will show that, classically, this can be
ðX1 Þνμ ¼ Gij ∂μ ϕi ∂ν ϕj ; ðX2 Þνμ ¼ ðX1 Þρμ ðX1 Þνρ : ð8Þ
obtained by setting
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi In principle, we could define ðX3 Þνμ , ðX4 Þνμ , etc., in an
1 μν 1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Rðz; z̄Þ ¼ T T μν − T μμ T νν ¼ −Det½tμν ; ð4Þ obvious way. However, in the Lagrangian only the trace of
2 4 these tensors may appear, and for any 2 × 2 matrix the
only independent trace invariants are x1 ¼ ðX1 Þμμ and
where tμν ¼ T μν − 12 gμν T ρρ is the traceless part of the stress-
x2 ¼ ðX2 Þμμ , due to identities such as
energy tensor T μν. We will construct the explicit two-
parameter action S ðλ;γÞ for a theory of N scalar fields. This
2x3 ¼ 3x1 x2 − x31 ; 2x4 ¼ 2x21 x2 − x41 þ x22 ; ð9Þ
turns out to be related to the action for the unique conformal
and electromagnetic duality-invariant deformation of with xn ¼ ðXn Þμμ . Hence, the most general Lagrangian will
Maxwell theory, the “ModMax” theory [16,17] (see also
depend on x1 , x2 , and possibly on a term x0 that does not
Refs. [18,19]). ModMax is a four-dimensional model, and couple to gμν , Lγ ¼ Lγ ðx0 ; x1 ; x2 Þ. For such a Lagrangian,
it depends on a single marginal parameter γ:
Z  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ∂Lγ ∂Lγ
4 chγ 2 shγ T μν ¼ −2ðX1 Þμν − 4ðX2 Þμν þ gμν Lγ : ð10Þ
S MM
γ ¼ dx − F þ ðF2 Þ2 þ ðFF̃Þ2 ; ð5Þ ∂x1 ∂x2
4 4
Computing the contractions T μν T μν and ðT μμ Þ2 , and simpli-
where F indicates the electromagnetic field strength and F̃ fying the resulting expression with the help of the trace
its dual. Equation (5) admits a T T̄-like deformation in terms identities, Eq. (9), we can evaluate the square of the
of a Born-Infeld action with parameter λ. The resulting two- operator R of Eq. (4) as
parameter, four-dimensional action S ðλ;γÞ is the solution of  
two commuting flow equations [20–22]. The flows we ∂Lγ ∂Lγ 2
R2 ¼ ð2x2 − x21 Þ þ 2x1 : ð11Þ
introduce here precisely generate, and extend, the reduction ∂x1 ∂x2
to two dimensions of the ModMax-Born-Infeld action
recently obtained in Ref. [23]. After discussing the gen- This expression is surprisingly nice in more than one way.
eralization of our construction to actions involving fer- First, the partial derivatives appear in a perfect square so
mions, we will comment on the implications of this relation that, away from ∂x1 Lγ ¼ −2x1 ∂x2 Lγ , the flow equation for
and on several immediate directions of study. the Root-T T̄ operator, Eq. (4), will be linear,
Root-T T̄ flow for N scalars.—We start by considering a
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi∂L ∂Lγ

rather general Lagrangian density for N real scalars ϕi of ∂ 2 γ
L ¼ Rγ ¼  2x2 − x1 þ 2x1 : ð12Þ
the form ∂γ γ ∂x1 ∂x2

1 Both choices of the branch can be recovered by flipping the

L ¼ ðGij ðϕÞgμν þ Bij ðϕÞεμν Þ∂μ ϕi ∂ν ϕj − VðϕÞ; ð6Þ
2 sign of γ. As 2x2 ¼ x21 gives a fixed point of Eq. (12), the

201604-2
 PHYSICAL REVIEW LETTERS 129, 201604 (2022)
 2  2 
branch cannot change along the flow. Additionally, the flow 2 ∂L ∂L ∂L ∂L 2
O ¼ −L − 2 þ2 x þ2 ðx1 − x2 Þ
equation will not depend explicitly on L, so that the ∂x1 ∂x1 ∂x2 1 ∂x2
dependence on x0 will affect the flow only through its  
2 ∂L ∂L
initial conditions. Finally, Eq. (12) can be integrated, × ðx1 − x2 Þ þ 2L x þ2 x : ð19Þ
∂x1 1 ∂x2 2
 
ðγÞ ðγÞ
Lγ ðx0 ; x1 ; x2 Þ ¼ Lγ¼0 x0 ; x1 ; x2 ; ð13Þ Now, consider a T T̄ deformation of Lγ , Eq. (13), where the
ðγÞ
dependence on x1 and x2 appears exclusively through x1
ðγÞ
where and x2 . A change of variables and some algebra, along with
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the observation contained in Eqs. (15) and (16) observations,
ðγÞ allows us to reformulate the operator O in terms of L and its
x1 ¼ chðγÞx1  shðγÞ 2x2 − x21 ; ð14aÞ ðγÞ ðγÞ
partial derivatives with respect to x1 and x2 . In fact, we find
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi that O is given precisely by the same expression, Eq. (19), up
ðγÞ
x2 ¼ chð2γÞx2  shð2γÞx1 2x2 − x21 ; ð14bÞ ðγÞ ðγÞ
to promoting x1 and x2 to x1 and x2 , respectively.
Therefore, to take a theory defined by Lγ and T T̄ deform
and x0 is not deformed. Note that the combination [obtaining Lðλ;γÞ ], we may as well first T T̄ deform L0
ðγÞ ðγÞ
(thereby obtaining Lλ ), and then replace everywhere x1 and
½x1 2 − x2 ¼ x21 − x2 ð15Þ ðγÞ ðγÞ
x2 with x1 and x2 . This last step is precisely tantamount
to a Root-T T̄ deformation, which goes to show that the
is constant along the flow. In fact, an action that depends on
diagram of Fig. 1 indeed commutes.
only this combination and through x0 is a fixed point of this
The two-parameter Lagrangian can then be found by
flow. We also note that Eq. (14) has the property
taking any T T̄-deformed Lagrangian, such as the ones in
∂Lγ ∂Lγ [24], and replacing x1 , x2 with the expressions in Eq. (14).
ðγÞ ∂Lγ ðγÞ ∂Lγ
x1 þ 2x2 ¼ x1 þ 2x2 ; ð16Þ For the special case of N free scalars we find
∂x1 ∂x2 ∂x
ðγÞ
∂x
ðγÞ
1 2 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðγÞ ðγÞ ðγÞ
and that the trace of the stress-energy tensor for Lγ , 1 þ 2λx1 þ 2λ2 f½x1 2 − x2 g − 1
Lðλ;γÞ ¼ ; ð20Þ
Eq. (10), vanishes if and only if 2λ
  which is precisely the result obtained in Ref. [23] from the
1 μ ∂Lγ ∂Lγ
T μ ¼ Lγ − x1 þ 2x2 ¼ 0: ð17Þ dimensional reduction of the four-dimensional ModMax-
2 ∂x1 ∂x2
Born-Infeld theory [16,17,21,22].
Fermions.—It is natural to wonder whether our discus-
Equation (16) therefore implies that, if the stress-
sion may be extended to models with fermions. It is unclear
energy tensor for a seed theory L0 is traceless, then the
how to treat a nonanalytic function of Grassmann variables
same is true for the deformed theory Lγ. Importantly, this
only, such as the one that would arise in a Root-T T̄-
confirms that the operator Rγ defines a classically marginal deformed model involving only fermions. It is more
deformation. natural to first look at a theory of bosons and fermions,
Note that, in the special case of N free bosons, it holds which admits a regular expansion in Grassmann vari-
that ables (with nonanalytic coefficients in the boson fields).
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Let us consider a free model of N B bosons and N F Dirac
chγ μ shγ
Lγ ¼ Φμ  2Φνμ Φμν − ðΦμμ Þ2 ; ð18Þ fermions,
2 2
1 i i
with Φμν ¼ ∂μ ϕi ∂ν ϕi . This is the result found in Ref. [23] L ¼ ∂μ ϕi ∂μ ϕi þ ψ̄ I γ μ ∂μ ψ I − ð∂μ ψ̄ I Þγ μ ψ I ; ð21Þ
2 2 2
from dimensional reduction of ModMax, Eq. (5). Notice
also that this flow is nontrivial only for N ≥ 2, for which x1 where i ¼ 1; …N B , and I ¼ 1; …N F .
and x2 are algebraically independent; otherwise, it is merely To build an ansatz for the γ-flowed action, we need to
a rescaling of the Lagrangian, Lγ ¼ eγ L0 . consider the stress-energy tensor of this model. Here, we
Two-parameter flow.—Having determined the Root-T T̄ will consider the Hilbert, rather than Noether, stress-energy
flow, we will now verify that it commutes with the ordinary tensor. Much like for T T̄ flows [25], we expect that the
T T̄ flow. The latter is sourced by the operator O of Eq. (1). deformed Lagrangian differs in the two choices by
We can express O in terms of the Lagrangian and its terms that vanish on shell. It is possible to show that,
derivatives following what we did for R around Eq. (11), for theories that depend on the fermions only via the
and we get the (rather cumbersome) formula symmetric tensor ψ̄γ ðμ ∂νÞ ψ − ð∂ðμ ψ̄Þγ νÞ ψ, the Hilbert tensor

201604-3
 PHYSICAL REVIEW LETTERS 129, 201604 (2022)

is the symmetrized Noether tensor. This ensures that the deformation (of a CFT) and only later break conformal
deformed Lagrangian will be a function of invariance by T T̄. This will clarify the general quantum
properties of Root-T T̄.
ðX1 Þμν ¼ ∂μ ϕi ∂ν ϕi þ iψ̄ I γ ðμ ∂νÞ ψ I − ið∂ðμ ψ̄ I Þγ νÞ ψ I ; ð22Þ Again with quantization in mind, it would also be useful
to study Root-T T̄ in Hamiltonian terms [14,15], as was
and of ðX2 Þμν ¼ ðX1 Þμρ ðX1 Þρν —more precisely of their recently done [27] for ModMax, and to investigate whether
traces x1 and x2 . Given such a dependence, the flow the Root-T T̄ deformation acts in a simple manner on the S
equation can be determined by truncating all expressions matrix of (integrable and nonintegrable) theories, like T T̄
to the bosons, which immediately reproduces Eq. (12). does [9,31]. (See Ref. [32] for recent work on the
Hence, given an action that depends on x1 and x2 as in integrability of Root-T T̄.) In turn, understanding the
Eq. (22), the Root-T T̄ flow is found by replacing the γ- quantum properties of Root-T T̄ would also extend our
dependent expressions, Eq. (14). The same argument of the understanding of ModMax, where quantization remains a
previous section confirms this flow also commutes with the major stumbling block. This is expected for a D ¼ 4 model
usual T T̄ flow. Note that in Eq. (21) we started from a for which no perturbative quantization scheme is available.
system of free Dirac fermions. However, the same argu- Approaching this issue from two dimensions, where CFTs
ment concerning the integration of the flows holds if are under especially good control, may shed new light on
chirality conditions are imposed. Interestingly, in the the issue.
presence of at least one real chiral fermion the deformed Recalling that T T̄ is only one of many current-current
Lagrangian is nontrivial even with N B ¼ 1. deformations [2], all obeying a flow equation [33], it would
Conclusions and outlook.—We have discussed Root-T T̄ be interesting to see if the higher-spin analogs of T T̄ also
deformations, which commute with the T T̄ flow and are commute with Root-T T̄. Moreover, one could also consider
defined in terms of a classically marginal nonanalytic higher roots of higher-spin T T̄-like operators. They would
operator. One might have expected the deformed theories also be naively marginal (nonanalytic) operators. Their
to be pathological and unnatural. Instead, they are appeal- study would probably be difficult, as is the case for higher-
ing, satisfying a simple flow equation, and encompass the spin T T̄ deformations [33], but might be instructive already
models arising from dimensional reduction of ModMax in the case of simple theories; see Ref. [34] for T T̄. By
[16,17]. ModMax is also a nonanalytic classically con- allowing for the breaking of Lorentz symmetry, more
formal theory, with many intriguing properties: it admits general scale-invariant roots of current-current deforma-
various physical solutions such as plane waves [16] and
tions could also be considered—for example, a Root-JT̄
dyons [26], its Hamiltonian is bounded from below [27], it
deformation based on the one of [35].
can be supersymmetrized [28–30], and it can be related to
In this Letter, we have merely initiated the study of
Maxwell theory coupled to an axion dilaton [26]—see also
models involving fermions, considering a relatively small
the recent review [19]. While many questions about
class of models. It would be especially interesting to
ModMax remain open, it is certainly a physically interest-
perform a systematic study of the Root-T T̄ deformations
ing theory and it is inspiring that the two-dimensional Root-
of superconformal theories. It is well-established that a
T T̄ deformation falls in the same class of models.
supersymmetric model remains supersymmetric under T T̄
The main open question is how to quantize such
flows [25,36–39]. The fact that ModMax may be super-
nonanalytic theories. This will be subtle but may be
symmetrized is also encouraging. However, it is not clear
possible for Root-T T̄ flows of CFTs by demanding the
that the deformation of a free superconformal action of the
deformed theory to be conformal and bootstrapping its
Eq. (21) type [for example, the N ¼ ð1; 1Þ supersymmetric
properties. If successful, one could construct new marginal
system when N B ¼ N F ] gives a superconformal model—
deformations of generic CFTs. In general, the Root-T T̄
this certainly requires the supersymmetry transformations
operator will not be part of spectrum of local operators of
the theory (with the intriguing exception of the single free to be modified. In the case of the standard T T̄ deformation
boson). Hence, the deformation should result in one- it was shown that the operator O of Eq. (1) is (on shell)
parameter families of theories that are quite different from equivalent to a manifestly supersymmetric supercurrent-
the familiar local unitary two-dimensional CFTs—and it squared operator defined in superspace [25,36]. It is
would be very interesting to understand such differences. A straightforward to propose manifestly supersymmetric
first step is understanding how N ≥ 2 free bosons get extensions of the Root-T T̄ operator, Eq. (3), for super-
coupled by the deformation and extending the analysis to conformal field theories. In the simplest case of N ¼ ð0; 1Þ
Wess-Zumino-Witten models and minimal models where supersymmetry the following superspace Lagrangian is an
algebraic techniques are very well developed. The quan- example
tization of Root-T T̄-deformed QFTs (as opposed to CFTs) Z pffiffiffiffiffiffiffi
S þþþ T −−−−
is likely to be harder. By imposing that T T̄ and Root-T T̄ dθþ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ T T̄ þ fermions: ð23Þ
deformations commute, one can first perform the Root-T T̄ T þþþþ T −−−−

201604-4
 PHYSICAL REVIEW LETTERS 129, 201604 (2022)

Here, S þþþ and T −−−− define the N ¼ ð0; 1Þ super- holographic duals, and we plan to return to some of these
currents, and T þþþþ ¼ Dþ S þþþ ; see Refs. [25,36]. For in the near future.
a single N ¼ ð0; 1Þ real scalar multiplet, the superspace
We thank Gleb Arutyunov, Grisha Korchemsky,
Lagrangian A− ðγÞ ¼ eγ ðDþ ΦÞð∂−− ΦÞ satisfies a classical
Dima Sorokin, Marika Taylor, Roberto Volpato, and
flow driven by the Eq. (23) operator. This is a direct analog
Aron Wall for helpful discussions, and Dima Sorokin for
of the bosonic Root-T T̄ deformation of a single real free
useful comments on a draft of this Letter. C. F. is supported
scalar field discussed above. Similar superspace Root-T T̄ by U.S. Department of Energy Grant No. DE-SC0009999
operators can be defined for models with more supersym- and by funds from the University of California.
metry. Aside from supersymmetry, it would also be A. S. acknowledges support from the European Union—
interesting to consider the bosonization and fermionization NextGenerationEU, and from the program STARS@UNIPD,
of simple two-dimensional CFTs and their Root-T T̄ flows. under project “Exact-Holography,” A new exact approach to
Finally, a major application of T T̄ deformations has been holography: harnessing the power of string theory, con-
to holography, mapping them to gravitational theories in formal field theory, and integrable models. L. S. is sup-
three-dimensional anti–De Sitter space-time [6,7]; see also ported by a postgraduate scholarship at the University of
Refs. [40–42]. The holographic interpretation of such a Queensland. The work of G. T.-M. is supported by the
marginal but nonanalytic deformation of the boundary Australian Research Council (ARC) Future Fellowship
CFT2 should be understood more clearly. An immediate FT180100353, and by the Capacity Building Package of
application of this Letter is to recast the Root-T T̄ defor- the University of Queensland. L. S. and G. T.-M. also thank
mation in terms of the boundary conditions of the AdS3 the mathematical research institute MATRIX in Australia
metric much like in [43], which will provide a holographic where part of this research was performed.
formulation of the deformation and yield a new class of
holographic backgrounds [44].
Holography provided a way to define new irrelevant Note added.—After the completion of this work we became
deformations in D > 2 dimensions [45,46], in particular in aware of the second version of Ref. [23], where it was
D ¼ 4 for Maxwell [20], its supersymmetrization [47], and independently shown that the operator (4) generates the
ModMax theories [21,22]. Both our D ¼ 2 operator R γ-flow of the Lagrangian (20) in the particular case of
[Eq. (4)], and the D ¼ 4 operator generating the γ flow of N Bosons with seed Lagrangian (6), when Gij ðϕÞ ¼ δij and
ModMax-Born-Infeld [21] can be written as Bij ðϕÞ ¼ 0. Further, Ref. [24], which appeared soon after
our work, explores the relation between Root-T T̄ and
pffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi 1 ModMax theories.
−tμν tμν ; tμν ¼ T μν − g T ρρ ; ð24Þ
D μν
where tμν is the traceless part of T μν . It is natural to wonder *
Corresponding author.
whether this operator plays a universal role in generating [email protected]
ModMax-like deformations in any dimension, and to †
Corresponding author.
investigate generic Root-T T̄-type deformations in D > 2 [email protected]
‡
models, including holographic ones. One piece of evidence Corresponding author.
that nonanalytic deformations may be relevant for holog- [email protected]
§
raphy in D > 2 is that operators involving square roots of Corresponding author.
field gradients have been studied in the context of the fluid- [email protected]
gravity correspondence [48] in general dimension D. [1] A. B. Zamolodchikov, arXiv:hep-th/0401146.
[2] F. A. Smirnov and A. B. Zamolodchikov, Nucl. Phys. B915,
A different, but very interesting, direction is the
363 (2017).
“generalized holographic principle” of [49], which also [3] A. Cavaglià, S. Negro, I. M. Szécsényi, and R. Tateo,
pffiffiffiffiffiffiffideformations in D > 2; one might
involves stress-tensor J. High Energy Phys. 10 (2016) 112.
wonder whether TT -type operators may be needed to [4] M. Baggio and A. Sfondrini, Phys. Rev. D 98, 021902(R)
UV complete the deformations used in that proposal. (2018).
Furthermore, the interplay between the ultrarelativistic [5] S. Frolov, Proc Steklov Inst Math / Trudy Matematiche-
and nonrelativistic limits of a CFT2 and the Bondi- skogo instituta imeni VA Steklova 309, 120 (2020).
Metzner-Sachs symmetry in three dimensions found in [6] L. McGough, M. Mezei, and H. Verlinde, J. High Energy
Phys. 04 (2018) 010.
[14,15] also suggests possible relations between operators
[7] A. Giveon, N. Itzhaki, and D. Kutasov, J. High Energy Phys.
like Eq. (24) and deformations of space-time symmetries in 07 (2017) 122.
dimensions higher than two. [8] S. Dubovsky, V. Gorbenko, and M. Mirbabayi, J. High
We believe that addressing all these questions may Energy Phys. 09 (2017) 136.
result in exciting new insights on the structure of conformal [9] S. Dubovsky, V. Gorbenko, and G. Hernández-Chifflet,
and nonconformal quantum field theories and of their J. High Energy Phys. 09 (2018) 158.

201604-5
 PHYSICAL REVIEW LETTERS 129, 201604 (2022)

[10] R. Conti, S. Negro, and R. Tateo, J. High Energy Phys. 02 [31] G. Mussardo and P. Simon, Nucl. Phys. B578, 527 (2000).
(2019) 085. [32] R. Borsato, C. Ferko, and A. Sfondrini, arXiv:2209.14274.
[11] J. Cardy, J. High Energy Phys. 10 (2018) 186. [33] G. Hernández-Chifflet, S. Negro, and A. Sfondrini, Phys.
[12] Y. Jiang, Commun. Theor. Phys. 73, 057201 (2021). Rev. Lett. 124, 200601 (2020).
[13] C. Ferko, Supersymmetry and irrelevant deformations, [34] G. Camilo, T. Fleury, M. Lencsés, S. Negro, and A.
Ph.D. thesis, Chicago University, 2021. Zamolodchikov, J. High Energy Phys. 10 (2021) 062.
[14] P. Rodríguez, D. Tempo, and R. Troncoso, J. High Energy [35] M. Guica, SciPost Phys. 5, 048 (2018).
Phys. 11 (2021) 133. [36] C.-K. Chang, C. Ferko, and S. Sethi, J. High Energy Phys.
[15] A. Bagchi, A. Banerjee, and H. Muraki, arXiv:2205.05094. 04 (2019) 131.
[16] I. Bandos, K. Lechner, D. Sorokin, and P. K. Townsend, [37] H. Jiang, A. Sfondrini, and G. Tartaglino-Mazzucchelli,
Phys. Rev. D 102, 121703(R) (2020). Phys. Rev. D 100, 046017 (2019).
[17] I. Bandos, K. Lechner, D. Sorokin, and P. K. Townsend, [38] C.-K. Chang, C. Ferko, S. Sethi, A. Sfondrini, and G.
J. High Energy Phys. 03 (2021) 022. Tartaglino-Mazzucchelli, Phys. Rev. D 101, 026008 (2020).
[18] B. P. Kosyakov, Phys. Lett. B 810, 135840 (2020). [39] E. A. Coleman, J. Aguilera-Damia, D. Z. Freedman, and
[19] D. P. Sorokin, Fortschr. Phys. 70, 2200092 (2022). R. M. Soni, J. High Energy Phys. 10 (2019) 080.
[20] R. Conti, L. Iannella, S. Negro, and R. Tateo, J. High Energy [40] L. Apolo, S. Detournay, and W. Song, J. High Energy Phys.
Phys. 11 (2018) 007. 06 (2020) 109.
[21] H. Babaei-Aghbolagh, K. B. Velni, D. M. Yekta, and H. [41] S. Chakraborty, A. Giveon, and D. Kutasov, J. High Energy
Mohammadzadeh, Phys. Lett. B 829, 137079 (2022). Phys. 11 (2020) 057.
[22] C. Ferko, L. Smith, and G. Tartaglino-Mazzucchelli, SciPost [42] E. Coleman, E. A. Mazenc, V. Shyam, E. Silverstein, R. M.
Phys. 13, 012 (2022). Soni, G. Torroba, and S. Yang, J. High Energy Phys. 07
[23] R. Conti, J. Romano, and R. Tateo, J. High Energy Phys. 09 (2022) 140.
(2022) 085. [43] A. Bzowski and M. Guica, J. High Energy Phys. 01
[24] G. Bonelli, N. Doroud, and M. Zhu, J. High Energy Phys. (2019) 198.
06 (2018) 149. [44] S. Ebert, C. Ferko, H.-Y. Sun, and Z. Sun (to be published).
[25] M. Baggio, A. Sfondrini, G. Tartaglino-Mazzucchelli, and [45] M. Taylor, arXiv:1805.10287.
H. Walsh, J. High Energy Phys. 06 (2019) 063. [46] T. Hartman, J. Kruthoff, E. Shaghoulian, and A. Tajdini,
[26] K. Lechner, P. Marchetti, A. Sainaghi, and D. P. Sorokin, J. High Energy Phys. 03 (2019) 004.
Phys. Rev. D 106, 016009 (2022). [47] C. Ferko, H. Jiang, S. Sethi, and G. Tartaglino-
[27] C. A. Escobar, R. Linares, and B. Tlatelpa-Mascote, Int. J. Mazzucchelli, J. High Energy Phys. 02 (2020) 016.
Mod. Phys. A 37, 2250011 (2022). [48] G. Compere, P. McFadden, K. Skenderis, and M. Taylor,
[28] I. Bandos, K. Lechner, D. Sorokin, and P. K. Townsend, J. High Energy Phys. 07 (2011) 050.
J. High Energy Phys. 10 (2021) 031. [49] G. Araujo-Regado, R. Khan, and A. C. Wall, arXiv:
[29] S. M. Kuzenko, J. High Energy Phys. 09 (2021) 180. 2204.00591.
[30] S. M. Kuzenko and E. S. N. Raptakis, Phys. Rev. D 104, [50] H. Babaei-Aghbolagh, K. Babaei Velni, D. M. Yekta, and H.
125003 (2021). Mohammadzadeh, arXiv:2206.12677.

201604-6