Arxiv:2012.01838V2 [Gr-Qc] 26 May 2021 Oee,Ece Esrmnsby 2]

arXiv:2012.01838v2 [gr-qc] 26 May 2021 oee,ece esrmnsby 2]. measurements [1, exceed contribution, observed Universe this however, for the our computations for of theoretical account Quantum expansion to to accelerated anticipated gravita- magnitude is late-time the adequate that to energy of attributed vacuum be the be of to effect physics. thought Relativitytional modern generally to General is enigma of tenacious It a Theory pose Einstein’s to continues in Λ constant t n ih 1–1.Acroiyo h omlgclcon- cosmological the of grav- curiosity of A speeds [19–21]. the light between and ity equality challenges measured hard the with modi- confronted from a is from that dynamics gravity self-acceleration of energy cosmic fication dark of on growing concept both the the on 18], and [17, and constraints Λ stronger [3] to ever alternatives fine-tuned puts observations be of to wealth of dark need However, breakdown may [14–16]. a energy scales or large dark at Cosmos a Relativity ac- the as General cosmic such permeating origin with field different extent energy a full to attributed to a celeration gravitating of vacuum from conjecture preventing energy the [5–13] motivates mechanism undetermined also yet it for prescription But energy vacuum missing standard contribution. a the of imply determination may correct This the 4]. [3, nitude ∗ † [email protected] [email protected] h udmna aueudryn h cosmological the underlying nature fundamental The xlrn h eftnn ftecsooia osatfro constant cosmological the of self-tuning the Exploring aumeeg rmfel rvttn n nbiga estim an enabling and cosmological gravitating the freely of from mechanism energy self-tuning vacuum a as proposed was te nvre rsnlsotaseilcs oftr explo future to case special fo a to a out basi frequently number singles is the this number or as same find universes proton the we other whether its Universe of of question our size the For leave the life. to order eventually universe cos the and a choose the co of we throughout multiverse, of size the life w horizon be of of epoch, to analysis emergence typical anthropic matter an the very and compare a constant fro to at cosmological inferred quantity live age, the we cosmic of that of densities conclude function proba We a the as for processes. Universe interpre estimation our multiverse an in a provide life motivate and w to problem inflation used coincidence of conv be unification the a alternatively of in can possibility t embedded the find be explore furthermore and then We we theory cases, scalar-tensor Horndeski certain For a incapac energy. from prevented vacuum be Gauss Model can coupled loops a graviton by me strength, action field self-tuning the sequeste local of the the d to of formulation extension self-tuning first We of Jordan-frame differences proposal. current and this similarities the of aspects to new formalism here explore We value. eety h aito ftePac asi h eea Rel General the in mass Planck the of variation the Recently, .INTRODUCTION I. ´ preetd hsqeTóiu,Uiested Genè Université de Théorique, Physique Département de 4qa retAsre,11 e`v ,Switzerland Genève 4, 1211 Ansermet, Ernest quai 24 ailSba Blanco Sobral Daniel & 0odr fmag- of orders 50 ∗ n ua Lombriser Lucas and u nvre[8.Arlvn ruett ovn both solving to for argument with case relevant universe the A not a [28]. simply of Universe is case our which the particles, massless for GL(4), with- only except possible group not fine-tuning, the is this out by that given states is theorem Weinberg’s that symmetry such remaining translations, under the symmetric is the state assuming the vacuum Then, “absorb” protecting accordingly. sector, curvature to field spacetime introduced matter the the be of energy could vacuum Typ- large field invokes sector. a matter that such the for problem in ically, check field old scalar consistency dynamical the extra important to an solution an candidate here any offers [3] rem er n cl naine Weinberg’s supersym- of invariance. context scale the to in solution and instance, potential metry for a aspect, as old on the early have invoked principles been Symmetry research already reviews). of for amount [24–28] great Refs. a (see motivated has Problem stant expansion. late-time accelerated discov- the the of before arose ery already that gravitation vacuum h bevdcsi ceeainwt ml observed small the and a constant with cosmological acceleration coincident cosmic observed the hsb sflt itnus ewe the between distinguish to useful be thus oniec rbe 4 2 3.T drs h differ- the address To the 23]. of aspects 22, ent [4, problem coincidence rei h at o ili etu nteftr.This future. the in true be it will a nor provokes not in past, circumstance was time This the same existence. in the into true at come we matter be that of to history that cosmic happens to density size energy comparable physical of its a that finding to is hint a understanding, give may perhaps which stant, h eeiyadiprac fteCsooia Con- Cosmological the of importance and severity The igmcaim ete hwhwwt an with how show then We mechanism. ring Bne emadacmainfour-form companion a and term -Bonnet ttn h ervtto fteStandard the of degravitation the itating ration. tradtretilpae formation planet terrestrial and star m t eftnn.Tersligequations resulting The self-tuning. ith fmgiuedffrneo h evolving the of difference magnitude of the revisit we context, this In tation. smdfrteeegneo ieacross life of emergence the for ssumed to ftemgiueo t observed its of magnitude the of ation osat rvnigSadr Model Standard preventing constant, uligboko tm,molecules, atoms, of block building c iiyo h mrec fintelligent of emergence the of bility i itr fdffrn nvre in universes different of history mic prbesz.Fradimensionless a For size. mparable ninllclsl-uigformalism. self-tuning local entional vlpa qiaetEinstein-frame equivalent an evelop a hsetnincnb eatas recast be can extension this hat hns n s hst highlight to this use and chanism mpa tapoiaey4.We 42. approximately at peak rm eew hudepc h energy the expect should we here tvsi isenHletaction Einstein-Hilbert ativistic omlgclCntn Problem Constant Cosmological lnkms variation mass Planck m † ve h Now? Why oudu rthe or conundrum new old No-Go rbe of problem rbe of problem tcan it , Theo- 2 the old and new aspects of the Cosmological Constant ple uniform prior on our location in its evolution yields Problem is furthermore given by Weinberg’s Anthropic an expected value for the current fractional energy den- Principle [22] that only allows for values of Λ that do not sity of the net radiatively stable cosmological constant a priori preclude the formation of structure and stars of ΩΛ = 0.704 [12, 13], in good agreement with current and hence the emergence of life and conscious life in par- measurements [17]. This simple approach therefore offers ticular. a solution to both the old and new aspects of the Cosmo- There is a rich pool of proposed solutions to the Cos- logical Constant Problem. In Ref. [13] it was furthermore mological Constant Problem. Supergravity models, for shown how similarly to sequestering the global self-tuning example, typically involve the appearance of additional mechanism can be cast into a local form, specifically as three-form gauge fields together with the Standard Model a local scalar-tensor theory where the scalar field addi- particles, which may turn the cosmological constant into tionally couples to the flux of a three-form gauge field. a dynamical variable [6, 7, 29–31]. Perhaps the most In this paper, we investigate further yet unexplored as- successful candidate within this family of theories is the pects of this new self-tuning mechanism. Specifically, we Bousso & Polchinski neutralization mechanism [32]. It is develop an alternative Einstein-frame formulation to the a model with J four-form gauge field strengths coupled current Jordan-frame formalism of the mechanism. Fur- to the correspondent J charged membrane species that thermore, we address the question of whether graviton contribute to the effective value of the cosmological con- loops may spoil the degravitation of the Standard Model stant. The quantization of the gauge field fluxes yields vacuum energy contributions. Finally, we explore a uni- a steady-step neutralization of the cosmological constant fication of inflation with the local self-tuning formalism as long as the membrane species are continuously nucle- and a related multiverse interpretation of the resulting ating out of the vacuum. The mechanism may require equations. In this context, we also provide a crude esti- J (100) membranes to operate, which agrees with mate for the probability of the emergence of intelligent the∼ expected O number arising from the compactification life in our Universe as a function of cosmic age, inferred of String/M-theory to four dimensions [27, 32]. Alterna- from star and terrestrial planet formation processes. tively, one may, for instance, adopt a more usual path The paper is organised as follows. Sec. II briefly re- integral approach such as in the proposal of Barrow & views the previously established results on the self-tuning Shaw [8, 9]. Using the partition function of the Universe, mechanism, including the global and local Jordan-frame one can promote Λ to be a variable parameter by sum- formulations, the embedding of the local theory in Horn- mation over a given range of values. Taking a statistical deski scalar-tensor gravity [33], and the estimation of ΩΛ approach one can then construct the probability distribu- from our likely location in cosmic history. In Sec. III tion of observing a universe with a value Λ′ [Λ, Λ+dΛ]. we develop the Einstein-frame formulation of the self- Anthropic selection also plays a role in this model∈ in pro- tuning mechanism and we discuss in detail the similarities viding a window of allowed values of Λ′ with the observed and differences between self-tuning and sequestering. In value argued to be typical by the link of the time scale Sec. IV we explore the unification of inflationary models needed for galaxy formation and the age of the Universe. with the self-tuning mechanism, conduct an expedition Other solutions propose the gravitational decoupling of into the multiverse, and estimate the probability of the the vacuum energy. An example of this is the sequester- emergence of conscious life throughout cosmic history. ing mechanism [10, 11], in which in addition to the met- Sec. V is devoted to a discussion of the effect of graviton ric variation, Λ and a matter coupling are varied in the loops on the self-tuning feature. Finally, we conclude our Einstein-Hilbert action supplied with a global term that analysis in Sec. VI. depends on the new variables. This introduces a global constraint equation that cancels the vacuum energy contributions at the level of the Einstein field equations. A II. THE SELF-TUNING MECHANISM local version can also be formulated, in which case the global term is replaced by two coupled four-form gauge Recently, the simple additional variation of the General field strengths. Relativistic Einstein-Hilbert action with respect to the 2 Recently, another remedy of the Cosmological Con- quadratic Planck mass MP was proposed as a remedy to stant Problem was proposed from the simple extra vari- the Cosmological Constant Problem [12]. This allows the ation of the Einstein-Hilbert action with respect to the interpretation of the gravitational coupling as a Lagrange Planck mass or the gravitational coupling [12]. An inter- multiplier which imposes General Relativistic dynamics. pretation of this is the treatment of the Planck mass in The additional variation introduces a global constraint the action as a global Lagrange multiplier that imposes equation that yields a self-consistent self-tuning mecha- General Relativistic dynamics on the metric prescribing nism in which the large vacuum energy contributions to the spacetime for the matter fields. The resulting addi- the cosmological constant are no longer freely gravitat- tional constraint equation prevents vacuum energy from ing. A local counterpart to the global formalism was then freely gravitating by the self-tuning of the classical Λ. developed in Ref. [13]. We shall commence with a brief Moreover, it was shown that an evaluation of the con- review of the local version in Sec. II A and then show how straint equation for the observable Universe with a sim- the global version emerges from that in Sec. IIB. 3

2 A. Local self-tuning in a simple scalar-tensor model allows this auxiliary field to take the value MP across the entire manifold . Alternatively, by integrating out the M The local self-tuning mechanism is best illustrated three-form sector from the beginning and taking the limit ϕ M 2, we can interpret the gravitational coupling as with the most basic scalar-tensor theory. See, however, → P Sec. IIC for realisations in the most general scalar-tensor a global Lagrange multiplier (Sec. IIB) [12, 13]. theories and Refs. [12, 13] for a discussion of how self- Varying the total action with respect to gµν yields the tuning may be realised in other theoretical frameworks. Einstein field equations Given a general manifold equipped with a Rieman- M ϕ Gµν + (V (ϕ)+Λvac(ϕ)) gµν = τµν , (3) nian metric gµν , the simplest set up for the self-tuning mechanism is the Jordan-frame scalar-tensor theory de- where τµν is the stress-energy tensor attributed to m. scribed by the action L Next, the variation of S +SA with respect to ϕ yields the constraint equation 4 ϕ S = d x√ g R V (ϕ)+ m(gµν , Ψm) (1) M − 2 − L Z h i 4 1 ′ 2 ′ d x√ g R V (ϕ) MP Λvac(ϕ) with scalar field ϕ and potential V (ϕ). Eq. (1) may repre- M − 2 − − Z " sent the low-energy limit of a more fundamental theory, σ′(ϕ) εµνργ for instance, obtained from the four-dimensional com- + Fµνργ =0 , pactification or reduction of a higher-dimensional the- 4! √ g − # ory of gravity [34]. m represents the matter content, (4) L e.g., the Standard Model of Particle Physics with Ψm generically denoting any matter component. We shall where primes denote derivatives with respect to ϕ here separate the vacuum energy contributions into the form and throughout the paper. We stress that in Eqs. (3)–(4) 2 m = m MP Λvac(ϕ), where Λvac(ϕ) may be regarded the terms V (ϕ), Λvac(ϕ), σ(ϕ), and their derivatives must asL an arbitraryL − function of ϕ (see Ref. [13]). Note that be regarded as formal functions of the scalar field but Eq. (1) is a Jordan-Brans-Dicke action with Brans-Dicke that their value is fixed thanks to the condition ∂µ ϕ = 0. parameter ω = 0, which is also the case for f(R) theories. This is why Eq. (4) is not a dynamical equation as would 2 General Relativity is recovered in the limit ϕ MP . be the case for usual scalar-tensor theories but rather a In addition to the scalar-tensor action, we→ introduce constraint equation for each value of ϕ. the topological contribution Taking the trace of Eq. (3) and splitting off an arbitrary classical counterterm Vc(ϕ) from the potential with 1 4 µνρσ ∆ V V Vc, Eq. (4) may be recast into the form SA = d xǫ σ(ϕ) Fµνρσ , (2) ≡ − 4! M Z (2 β)M −2∆V − P where Fµνρσ = ∂[µAνρσ] is the field strength of an auxil- −2 −2 MP iary three-form gauge field Aνρσ to which we couple the + (2 α)(Λvac + MP Vc)= τ + ∆Λ , scalar field ϕ through an arbitrary smooth, non-linear − 2 h i (5) function σ(ϕ). Note that this additional sector may arise for example in supergravity models [6, 7, 29–32]. The where we have defined local self-tuning mechanism is thus described by the to- 2 ′ µνργ ′ 4 µνργ tal action S + SA. The formulation is similar to the local MP σ ε σ d xε Fµνργ ∆Λ Fµνργ = . sequestering framework of Ref. [11] but with some impor- ϕ ≡− 4! √ g − 4! d4x√ g R tant differences that we shall highlight in the following. − − (6) Most importantly, for self-tuning we only need one scalar For notational convenience, we have furthermoreR in- and three-form gauge field rather than two of each, and troduced here the set of parameters α ∂ ln Λvac/∂ ln ϕ we will see that here a classical counterterm is self-tuned and β ∂ ln ∆ V /∂ ln ϕ that characterize≡ the depen- ≡ to prevent vacuum energy from gravitating rather than dency of Λvac and ∆ V on the scalar field ϕ. The coun- obtaining a cancelation of the vacuum contributions. Al- terterm can now be chosen such that the cancelation though we note that effectively the sequestering mecha- of Λvac becomes explicit. Hence, ∂ ln Vc/∂ ln ϕ α. nism can be recovered in a special limit of the self-tuning Note that α and β do not need to be constants≡ but framework. their values are fixed once we set the value of ϕ. In The equations of motion for the total action S + SA Eqs. (5)–(6) we have also introduced the braket nota- 4 4 are obtained from its variations with respect to gµν , ϕ, tion ... = d x√ g (...)/ d x√ g to denote the h i M − M − and Aνρσ. The variation with respect to the three-form spacetime average of a quantity. Note that such averages R R gauge field yields the crucial condition ∂µ ϕ = 0 such that have been considered in the context of the Cosmological the dynamics of the scalar field is completely fixed, in the Constant Problem before such as in the Normalized Gen- sense that it does not have any local propagating degrees eral Relativity framework [35, 36] (see also Refs. [37–39]). of freedom. Note that ϕ shall be regarded as an auxiliary Finally, solving the constraint equation (5) for the clas- field. Hence, the presence of the additional SA sector sical counterterm Vc and plugging the result into Eq. (3), 4 we obtain the effective Einstein equations for the self- stein equations is tuning mechanism, 1 Gµν + (1 α)Λ F 1 −2 2 α " − − Gµν + (β α)MP ∆V − 2 α " − M −2 − + P τ g = M −2 τ , −2 2 µν P µν MP −2 h i # + τ + ∆Λ gµν = MP τµν , 2 h i # (8)

(7) ′ where we have defined F = (σ F)/VM, and VM is the four-volume of the cosmic manifold. Note that the addi- where we have set ϕ M 2. Vacuum energy contribu- ≡ P tional constant contribution F can simply be reabsorbed tions are prevented from freely gravitating in these equa- into the free classical Λ by the redefinition Λ=Λ˜ F. tions of motion, where Λ M −2 ∆ V is a free, radia- − ≡ P This contribution is radiatively stable and must be set by tively stable, classical cosmological constant to be deter- measurement. Furthermore, the same line of reasoning mined by measurement. Note that for α = β = 1 we re- as for ∆Λ holds for F in that it is naturally expected to cover the simplest global self-tuning result of Ref. [12, 13], be a vanishing quantity. Hence, we can simply drop the while the choice α = β = 0 reproduces the effective field global sector in the global formalism. In other words, we equations of the sequestering mechanism of Refs. [10, 11] can safely work with the General Relativistic Einstein- but with a different expression for the ∆Λ term. In local Hilbert action and simply perform the additional varia- sequestering, where one works with two additional three- 2 tion with respect to MP . The global self-tuning action form sectors, the quantity ∆Λ is essentially a ratio be- and the General Relativistic Einstein-Hilbert action are tween the fluxes of both three-forms, whereas in Eq. (6) equivalent. the denominator is instead the four-volume of the cosmic manifold . With the flux of the three-form gauge field M in the numerator being a finite, small, ultraviolet-stable C. Self-tuning in Horndeski gravity quantity, and assuming the Universe grows sufficiently old, it is natural to expect that ∆Λ 0. → As mentioned in Sec. II A the local self-tuning mechanism is not limited to action (1) but can be realised for scalar-tensor theories more generally [13]. This can B. Self-tuning in the global limit be shown by generalizing the procedure, for instance, to Horndeski gravity [33], which describes the most general local scalar-tensor theory in four dimensions that yields The Einstein field equations (7) may also directly be at most second-order equations of motion. The Horn- obtained from varying the General Relativistic Einstein- deski action can be written as [33, 40] Hilbert action both with respect to the metric and 2 the quadratic Planck mass MP and combining the re- 5 4 1 sults [12]. As we have seen in Sec. II A the global self- S = d x√ g i(gµν , ϕ)+ m(gµν , Ψm) , − 2 L L tuning formalism can be directly recovered from the local ZM " i=2 # one after integrating out the three-form sector in the ac- X (9) 2 2 tion and fixing ϕ MP and V (ϕ) MP Λ with Λ being where the sum runs over the generalized Lagrangian den- a free classical cosmological≡ constant.≡ This reproduces sities the General Relativistic Einstein-Hilbert action with an 2 = G (ϕ, X) , (10) additional sector σ(MP )F sitting outside the integral, 2 2 4 µνρσ L where F = (1/4!) M d x√ g ε Fµνρσ is the flux of 3 = G3(ϕ, X)φ , (11) the three-form. The separation− of the Lagrangian den- L = G (ϕ, X)R + G ,X (ϕ, X) R 2 L4 4 4 sity is performed as before, m = m MP Λvac. To 2 ;µν L L − (ϕ) + ϕ µν ϕ , (12) illustrate the global self-tuning we adopt for simplicity × ; ;µν a power-law scaling for the vacuum energy contributions 5 = G5(φ, X)Gµν ϕ 2α L such as Λvac = MP Λvac, where the overbar denotes in- 1 2 3 ν α µ ;µν dependency of M , but we could instead also maintain a G5,X (ϕ, X) (φ) +2ϕ µϕ ν ϕ α 3ϕ;µν ϕ ϕ . P − 6 ; ; ; − fully general behaviour in form of arbitrary functions of (13) 2 MP (see Refs. [12, 13]). The self-tuning is made explicit by splitting off a coun- Eq. (9) may again be seen as the effective low-energy limit 2α terterm Λc = MP Λc, analogously to the local case in of some more fundamental theory, for example, the four- Sec. II A, for which we then solve the constraint equa- dimensional compactification or reduction of a higher- tion obtained from the variation of the total action with dimensional theory of gravity [34]. The Gi are general 2 µ respect to M . The final expression for the effective Ein- functions of the scalar field ϕ with X = (1/2)∂µϕ ∂ ϕ P − 5 denoting the canonical kinetic term. Semicolons indicate framework provides interesting insights into the Why the standard covariant derivatives, ϕ;µν = µ ν ϕ. Note now? problem of why the energy densities of the matter that we recover the action (1) from (9) for∇ ∇ the choices and the cosmological constant happen to coincide today. G2 = 2 V (ϕ), G4 = ϕ, and G3 = G5 = 0. The evolution of τ in the observable Universe may be At inspection− of the terms appearing in action (9), we modeled by the cosmic background and the local spher- see that the local self-tuning mechanism remains opera- ical top-hat overdensity, which can be characterized in tive once we add the additional sector with coupling of terms of its physical radius. Its evolution is governed by the scalar field ϕ to the field strength of a three-form the competition between the self-gravitation inside the gauge field, Eq. (2). This term fixes the dynamics of overdensity and the background expansion. The coin- ϕ to take a constant value across the entire manifold cidence problem may then be phrased in terms of be- . Therefore, this condition makes all the derivative ing located at a particular place in this evolution. By termsM in Eq. (9) vanish, leaving only a free choice of con- then adopting a uniform prior on the dimensionless phys- stants G2(ϕ) and G4(ϕ). The local self-tuning mecha- ical top-hat radius as the simplest ansatz to estimate our nism can further be generalized to Degenerate Higher- likely location (see Refs. [12, 13, 42]), we find a present Order Derivative Scalar-Tensor (DHOST) theories [41] fractional energy density in the cosmological constant of or beyond-DHOST theories [34]. We refer the reader to Ref. [13] for a more detailed discussion on the operation of the self-tuning mechanism in more complex scalar-tensor ΩΛ(t0)=0.704 (14) models.

D. The value of Λ with t0 denoting the current age of the Universe. This value is in very good agreement with observations [17]. We refer the reader to Refs. [12, 13, 42] for a discussion Finally, a very interesting attribute of the self-tuning of more sophisticated priors in this anthropic ansatz. In mechanism is that by the inherent relation of the cosmo- Sec. IV D we shall perform a brief analysis of the emer- logical constant to the global spacetime average of the gence of intelligent life in our Universe. A generalisation matter content, it provides a natural framework to esti- of this computation to arbitrary cosmological and funda- mate the value of the observed cosmological constant Λ. mental parameters may serve to establish an anthropic For illustration, let us consider the case α = β = 1 and theoretical prior in a multiverse approach in future work. work with a vanishing F (or ∆Λ). The effective Einstein equations (8) or(7) then contain a residual cosmological Finally, we note the interesting observation that the −2 −2 constant of the form Λres = M τ /2, which must cor- scenario discussed here with Λ = M τ /2 following P h i res P h i respond to the observed cosmological constant and is a from adopting α = β = 1 and F = ∆Λ = 0 reproduces calculable quantity [12, 13, 42]. the prediction for Λ in the causal universe scenario of A na¨ıve average over the cosmic background manifold Refs. [46–48]. An interesting observational implication of for in τ would suggest that Λres 0 for a universe this scenario could be an inhomogeneous and anisotropic M h i → that grows old. In fact, that way the Universe should cosmological constant [49]. While the same is expected have collapsed one billion years ago for Λres to match the for the ultimate collapsed cells in the sequestering mech- observed finite value of the cosmological constant [12]. anism [42], the scenario discussed here for the self-tuning However, due to cosmic acceleration, in the far future mechanism produces a homogeneous and isotropic Λ (see, the observable Universe is not a simple extrapolation of however, Sec. IVC). our current cosmic background but instead reduces to the maximally gravitationally bound structure beyond which everything is redshifted away in finite time [43–45]. We can model the future Universe by adopting a halo model view of the Cosmos, in which one divides the cosmic manifold into the collection of ultimately isolated, maximally gravitationally bound, disconnected matter cells. Λres III. EINSTEIN-FRAME FORMULATION has been shown to then reproduce the observed value of the cosmological constant Λ in each of the patches separately as well as in the empty space between them, no So far the self-tuning mechanism has only been studied matter the sizes of these cells [12, 13, 42]. Adopting the in the Jordan frame [12, 13], where matter fields follow observable Universe as the manifold over which to per- geodesics of the metric but the metric satisfies a modified form the spacetime averages provides a self-consistent so- Einstein equation. We shall now examine how self-tuning lution to the Cosmological Constant Problem. However, operates in the conformally related and physically equiv- since the properties of the manifold are determined by Λ, alent Einstein frame, where metrics satisfy the Einstein the value of Λ remains a free quantity in this case. field equations but do not describe the geodesics of mat- Nevertheless, the relation of Λ to τ in the self-tuning ter fields. Hence, we consider the conformal transforma- h i 6

6 tiong ˜µν = ϕgµν of action (1), which gives if we redefine Λ˜ λ V (λ) in Eq. (16) and rewrite the argument of the≡ three-form coupling function, we could ˜ 4 R 3 2 −2 now perform variations with respect to λ as well as with S = d x g˜ ( µ ln ϕ) ϕ V (ϕ) respect to Λ˜ as is done for the global sequestering model. M − " 2 − 4 ∇ − Z p However, as we shall see in the following the single ex- −2 µν tra variation with respect to λ in the self-tuning recipe is + ϕ m(ϕ g˜ , Ψm) sufficient for preventing vacuum fluctuations from freely L # gravitating. 1 4 µνρσ + d xε σ(ϕ)Fµνρσ , 4! M Z (15) B. Self-tuning in the Einstein frame where tildes denote quantities in the Einstein frame. We Performing the variation of the self-tuning Einstein- could furthermore redefine the scalar field ϕ in order to frame action (16) with respect tog ˜µν , one finds the Ein- recover a canonical kinetic term by using the relation stein field equations (dϕ/dϕ˜ )2 (3/2)(d ln ϕ/dϕ)2. ≡ ˜ 4 2 Gµν + λ (V (λ)+ MP Λvac(λ))g ˜µν =τ ˜µν , (18) A. Similarities with sequestering mechanism whereτ ˜µν is the stress energy tensor for the matter ex- citations, which is related to its counterpart in the Jor- 2 Before we discuss the self-tuning mechanism in the Ein- dan frame through the relationτ ˜µν = λ τµν . This im- stein frame, we shall briefly discuss some similarities with plies that the traces are related by the formulaτ ˜ = λ4 τ. the sequestering mechanism of Refs. [10, 11]. Here, the Next, we shall perform the variation with respect to λ. local version [11] is written in the Jordan frame whereas This may be interpreted as the Einstein-frame analog of 2 the global version [10] is in the Einstein frame. The sim- the MP variation performed in Sec. II A. Analogously to ilarities between the sequestering and self-tuning mecha- our Jordan-frame calculations, we define the Einstein- nisms become more evident when we work in the global frame parametersα ˜ ∂ ln Λ /∂ ln λ = ∂ ln Vc/∂ ln λ ≡ vac limit of Eq. (15). After integrating out the three-form and β˜ ∂ ln ∆V˜ /∂ ln λ, where we have also introduced sector, we can set the scalar field value to ϕ λ−2, the arbitrary≡ counterterm V , now a function of λ, which ≡ c where λ is a real parameter. We can thus interpret λ shares the same dependency as Λvac. The constraint as the reciprocal of MP in the Einstein frame. Taking equation then reads into account the condition ∂µ ϕ = 0, the global limit of 2 −4 Eq. (15) reads (4+˜α)(Vc + M Λ )+(4+ β˜)∆V˜ = λ τ˜ + F˜ , (19) P vac h i ˜ ˜ F˜ 6 F ˜ 4 R 4 where ∆V V Vc and = (2/λ ) σλ /VM with S = d x g˜ λ V (λ) ≡ − −4 − σλ dσ/dλ and V˜M = λ VM. Again, although the dif- M − " 2 − ≡ Z p ferent contributions are formal functions of λ, they take 4 −2 µν −2 a constant value for any given value of the parameter. + λ m(λ g˜ , Ψm) + σ(λ )F, L # Solving the constraint equation (19) for the counterterm V , and plugging the result into Eq. (18), we imme- (16) c diately see that the vacuum energy contributions cancel where F is again the flux of the three-form on . This out in the effective Einstein equations, is clearly reminiscent of the global sequesteringM action of 1 4 4 Ref. [10], G˜µν + (˜α β˜)λ ∆V˜ + τ˜ + λ F˜ g˜µν =τ ˜µν . 4+˜α − h i h i (20) M 2 S = d4x g˜ P R˜ Λ Similar arguments to those given for the Jordan-frame M − 2 − result in Sec. II also follow here. In particular, the ef- Z " p fective sequestering mechanism is contained within this 4 −2 µν Λ family of solutions forα ˜ = β˜ = 0, which gives the char- + λ m(λ g˜ , Ψm) + σ 4 4 . L # λ µ acteristic fraction of 1/4 of τ˜ as the contribution to the h i (17) residual cosmological constant. Forα ˜ = 2 we obtain a fraction of 1/2. If furthermore β˜ =α ˜, this− corresponds The sequestering action (17) is varied separately with re- to the choice of a potential of the form V (λ) λ−2. µν 2 ∼ spect tog ˜ , Λ, and λ. In contrast, for self-tuning the role Note that we have fixed ϕ MP in the Jordan frame of the cosmological constant is played by the second term whereas here we have used ϕ≡ λ−2. The global param- ≡ in Eq. (16), which in this case depends on the parameter eters MP and λ are thus interpreted as the reciprocal of λ. However, this turns out to be the same parameter each other. Hence, by construction, the dependency pa- that controls the coupling of the matter sector. Thus, rameters in the different frames are also related such that 7

˜ 2 α˜ = 2α and β = 2β. The Einstein tensor is invariant, This fixes the value of the inflaton at the scale MP in −˜ − 2 i.e., Gµν = Gµν once we have fixed the scalar field to the limit ϕ MP . The impact of the modification in the aforementioned constant value. The three-form flux Eq. (21) on→ the equations of motion is minimal. In the ˜ 2 2 contribution is simply F = (2/λ ) F. limit ϕ MP , the topological contribution now takes Finally, with these considerations,− we recover the Ein- the form→ Σ(ϕ) σ(ϕ)+ σ′(ϕ) = 0. This is forced to dif- stein field equations in Jordan frame, Eq. (7), and physics fer from zero for≡ a scalar field with6 frozen dynamics, i.e., is indeed independent of frame choice as expected. for the condition ∂µ ϕ = 0 to hold. This is natural for a reasonable, non-linear differentiable function of the inflaton field, but notice that it removes the possibility that IV. INFLATION AND THE MULTIVERSE σ(ϕ) takes the form of an exponential decay. Then, the constraint equation for the cancelation of vacuum energy It has been shown in Ref. [10] that an early inflationary contributions is completely analogous to that of Eq. (4) era does not affect the residual cosmological constant and but with Σ(ϕ) replacing σ(ϕ). the vacuum energy cancelation in the sequestering mech- From a formal point of view, we have consistently com- anism since the spacetime averages are dominated by the bined inflation with the self-tuning mechanism. The late-time evolution. The same conclusion applies to the idea is essentially that the scalar field evolves follow- self-tuning mechanism [12, 13]. With a scalar field being ing some inflationary model, possibly slow-rolling to the the source of local self-tuning another immediate ques- value ϕ M 2. From this point, the inflaton dynam- → P tion that suggests itself is whether it could play the role of ics freezes, the gravitational coupling is fixed and the the inflaton. This would imply a temporal division of the self-tuning mechanism starts to operate, producing the manifold into an era where the field would be evolv- degravitation of the vacuum energy contributions. This ing and oneM where it is has become fixed. Conceptually, idea is, however, not free of difficulties. Further analysis vacuum energy may still be prevented from gravitating in is needed in order to determine if this scenario is phe- the inflationary era due to a constraint equation imposed nomenologically feasible. Furthermore, it would be in- by the boundary to the post-inflationary era where the teresting to analyse whether this could be further united field is fixed, thereby avoiding Weinberg’s No-Go Theo- within an extended model of the Higgs boson. rem. This scenario also opens another question, which is whether this boundary could also be drawn between causally disconnected regions of spacetime and how self- B. Self-tuning in the inflationary era tuning works in a multiverse picture. We shall briefly inspect these questions in Secs. IV A–IVC. Given that with Eq. (21) the scalar field is dynamical at early times, we need to show that Weinberg’s No-Go Theorem is evaded and vacuum energy contributions are A. Can the scalar be the inflaton? not freely gravitating during the inflationary era. As we will see, this is achieved through a boundary constraint Eq. (1) is the action of a simple scalar-tensor theory, on the inflaton. Formally, let us take Eq. (21) and split embedding for example f(R) gravity, and hence it is per- the spacetime manifold along the timelike direction in fectly suitable for describing an inflationary scenario. In the region after the end of inflation and the inflationary this case, the scalar field ϕ is the so-called inflaton, for U 2 region = . By construction, we have (ϕ MP ) < which one may impose particular conditions such as a 0 in Nsuch thatM\U the scalar field ϕ is dynamical and− drives slow-roll evolution. Could we now conciliate both infla- the expansion.N In general, each region will have different tion and the self-tuning mechanism in one formalism with metric tensors, which shall be determined separately by the inflaton field ϕ driving inflation at early times and solving the correspondent field equations. Let us denote self-tuning through its late-time limit? the metric of by hµν and that of by gµν . N U To address this question, let us modify the self-tuning The region is just the late-time description for the action specified by Eqs. (1) and (2) by including a step Universe specifiedU by the action (1) with self-tuned cos- function in the topological sector, mological constant. We thus recover the results from Sec. II. In the region , however, physics is described by 4 ϕ µν N S = d x√ g R V (ϕ)+ m(g , Ψm) a dynamical scalar-tensor theory, which for our choice of M − 2 − L Z h i scalar-tensor action, Eq. (1), gives the two equations of 1 4 2 µνρσ motion + d x Θ ϕ MP ǫ σ(ϕ) Fµνρσ . 4! M − Z 2 (21) ϕ Gµν + (V (ϕ)+ M Λvac(ϕ)) hµν P (22) = ( µ ν hµν ) ϕ + τµν , Thus, by construction, the inflaton has active dynamics ∇ ∇ − 2 in the regime ϕ MP < 0, where the last term van- − 2 ishes. In the complementary regime of ϕ MP 0, R − ≥ ′ 2 ′ we recover the self-tuning action of Eqs. (1) and (2). V (ϕ) MP Λvac(ϕ)=0 , (23) 2 − − 8 where Gµν and R refer here to the metric hµν . Note that this calculation takes the form by virtue of Weinberg’s No-Go Theorem, the solution of M −2 ϕα these two equations alone will be a universe of massless P Λres N = α τ +3 ϕ particles or one in which we have not solved the fine- | (2 α) ϕ " 2 h i h i − h i tuning problem for Λ [3]. We will see how this is evaded in the following. + (2 α) ϕα ϕ (27) − h i # We first take the trace of the Einstein field equations, β which now involves derivatives of ϕ originating from the − β 2 β α−β ϕ +M 2∆V ϕ 1 − ϕ . additional first term on the right-hand side of Eq. (22). P 0 α " − 2 α ϕ # Since the field is evolving in we can no longer drop − h i them as we did in Sec. II. ButN we can still split off a Taking this expression on its own, we would first have to determine the evolution of ϕ from the equations of mo- counterterm Vc(ϕ) from the potential as well as define the functions α and β. With that the trace takes the tion to know the behaviour of the residual cosmological form constant in the region , which in turn would govern the evolution of ϕ. Hence,N this would not solve the fine- tuning problem in . However, the continuity conditions (2 β) ∆V (ϕ) in Eqs. (25) and (N26) provide a constraint between the − spacetime averages of ϕ. Specifically, continuity at the 2 τ 3 + (2 α)(Vc(ϕ)+ MP Λvac(ϕ)) = ϕ . boundary requires Λ ∂N =Λ U , where − 2 − 2 res| res| (24) −2 MP 1 2 Λ U = (β α)∆V + τ + M ∆Λ . (28) res| (2 α) − 2 h i P − " # Unlike Eq. (5) this is not a constraint equation but rather Therefore, by taking the limit ϕ M 2 in Eq. (27) which a local dynamical equation which relates the contribu- P must hold at the boundary ∂ ,→ and equating the result tions from the matter sector and the scalar field. This is with Eq. (28), we find a constraintN relating ϕα , ϕβ , due to the absence of a coupled three-form sector in the and ϕ . Hence, the boundary conditions self-tuneh i the region , which implies the lack of the crucial constraint solutionh toi prevent vacuum energy from freely gravitating ∂ ϕ =N 0 on the dynamics of the scalar field. µ in . N However, for consistency with the physics in the re- The simplest case for this scenario is the family of mod- gion , we require continuity of ϕ across the boundary els where β = α, to which also the sequestering model betweenU the two regions. More precisely, we impose the belongs. Here, we obtain conditions τ +3 ϕ ϕα = M 2α h iN h iN . (29) h iN P τ +2M 2∆Λ h iU P ∂ϕ ∂N = ∂ϕ U =0 , (25) Thus, the average value for the function that dictates the | | behavior of the different contributions to the cosmologi- Λres ∂N =Λres U . (26) | | cal constant, namely, vacuum energy and the scalar field potential, is given by a ratio between averaged quantities 2 in the two regions and . Only models that satisfy Hence, we ensure that we can take ϕ MP at the U N boundary of the region . As we shall see→ now this con- this constraint, or its more general version, will be suit- tinuity condition providesN a constraint in the inflationary able for a simultaneous description of inflation and self- region that solves the vacuum gravitation problem in . tuning. Specialising further to the simplest self-tuning N toy model with α = 1 and assuming slow-roll conditions For this purpose, let us take the spacetime average for the inflaton field such that ϕ N 0 in Eq. (29), of Eq. (24) over the region . This transforms the one obtains h i ≈ local equation into a relationN between averaged quan- 2 τ N tities. For simplicity, we shall assume power-law be- ϕ N MP h i 2 . (30) h i ≃ τ U +2MP ∆Λ haviours in ϕ for Vc, ∆V , and Λvac. We could, how- h i ever, adopt general functions of ϕ analogously to Sec. II. Then, the averaged value of ϕ in the region is com- α α N More specifically, we take Vc = V0 ϕ ,Λvac = Λ0 ϕ pletely determined by the matter content, the extra β and ∆V = ∆V0 ϕ such that the residual is of the form three-form gauge sector of the region , and the cor- β 2 α U Λres = (∆V0 ϕ + (MP Λ0 + V0) ϕ )+ ϕ, where sub- respondent four-volumes of each of the regions. scripts of zero denote amplitudes. We can now solve the spacetime average of constraint (24) for the amplitude C. Expedition into the multiverse of the counterterm V0, and use the result to show that the vacuum energy contribution cancels at the level of the Einstein field equations (22). More specifically, the The modified self-tuning action in Eq. (21) opens the residual cosmological constant for the region following door to a multiverse interpretation. More specifically, N 9 we may consider a Cosmos that is divided into several for this specific scenario. Eq. (31) defines a linear re- spacetime patches, each an own universe. This essen- lationship between quantities evaluated in the region tially corresponds to the late-time scenario predicted by and quantities evaluated in the region . Here ∆Λ is de-N Eternal Inflation [50]. Some of them may see a dynami- fined as in Eq. (6) but restricted to theU region . Hence, cal scalar field ϕ whereas others may have a frozen scalar Eq. (31) implies that the difference between theU averaged field. This is similar to the inflationary scenario dis- traces of the matter stress-energy tensor in each region cussed in Sec. IVB, however, with the manifold now not must be given by ∆Λ, which corresponds to the average split timelike but spacelike. Note that we already per- of the product of the derivative of the coupling σ′ with form a spacetime division of the background Universe the flux of the three-form over . With Eq. (6), we can by adopting the observable Universe for in Sec. II. recast Eq. (31) into the form U Self-tuning would operate in universes withM fixed scalar 4 αβγρ field dynamics, preventing vacuum energy contributions d xε Fαβγρ = from freely gravitating and proportioning a residual cos- ZU σ′(M 2) V mological constant self-consistent with the magnitude of P d4x√ g τ U d4x√ h τ , the accelerated expansion observed in this patch. Gen- 12 − − V − ZU N ZN erally, we could consider at least two types, or levels, of (32) multiverses [51]: which constrains the three-form flux in . We note that U • Type-I: Different regions will have different initial for more general configurations this relationship is more conditions with varying configurations of the mat- complicated, and we leave it to future work to explore ter content. The dynamics of the scalar field is multiverse interpretations of self-tuning in more detail therefore different across the different regions but (also see discussions in Refs. [12, 13]). the fundamental constants are fixed.

• Type-II: Fundamental constants such as the D. Life, the Universe and Everything Planck mass may also vary across the different regions. A multiverse picture coupled with the anthropic principle offers an interesting approach to understanding the We shall briefly inspect a simple multiverse picture for material configuration of our Cosmos or even the fun- self-tuning. For this, consider a simple set up where we damental constants and laws of Nature. To understand have only two disconnected regions. Let us denote by the emergence of life, we must understand fundamen- the region of the spacetime manifold where the tal physics, but to understand fundamental physics, we U 2 M condition (ϕ MP ) 0 holds in Eq. (21) such that must understand the emergence of life. As discussed in the self-tuning− mechanism≥ operates in . In the comple- Sec. II D, the self-tuning mechanism shows a natural dis- mentary region, denoted by = U , we shall have play of anthropic aspects, where we have used a prior 2 N M\U (ϕ MP ) < 0 such that the scalar field is dynamical. on our location in the cosmic history to estimate the Note− that the region may also well represent the col- observed current fractional energy in the cosmological N lection of a set of many other spacetime patches different constant ΩΛ. In the following, we shall perform a more from . Action (21) can now be split into two parts, involved yet still rudimentary estimate for the formation whichU in general will have different metrics. Let us call probability of conscious observers in the cosmic history of hµν the metric in the region and gµν the metric in the our Universe. We will infer this from the star formation self-tuning region . They areN determined by the corre- rate, the evolution of cosmic metallicty, and, depending spondent EinsteinU equations in each region. For simplic- on these, the formation probability of terrestrial planets ity, we shall furthermore assume that the matter content as potential hosts of intelligent life. is the same in both patches and that Λvac N = Λvac U . We first need to compute the star formation history. The results of Sec. IVB are formally valid| for this new| For this we follow Ref. [52], where we also adopt the same set up but they are given a new interpretation here. In cosmological parameters. Specifically, we use Ωm = 0.3, particular, by solving the dynamics with the constraint Ωb = 0.045, and h = 0.7 for the fractional energy den- equation (29), we can find a model that allows for re- sities in total matter and baryons and the dimensionless gions of dynamical ϕ while also canceling vacuum energy Hubble parameter, respectively. Furthermore, we adopt contributions in those regions. the solar metallicity Z⊙ =0.02, where Z more generally Let us consider some further simplifications. For self- denotes the metallicity in the gas and newly formed stars. 2 tuning in a Type-I multiverse, the limit ϕ MP must We use R =0.27 for the return fraction, quantifying the also hold in . Hence, if is a frozen region,−→ we must mass fraction of each generation of stars that is returned N 2 N have ϕ N = MP . This implies that the quotient in to the interstellar and intergalactic medium. Finally, we Eq. (30|) must be unity such that we obtain the simple adopt y = 0.019 for the net metal yield, the fractional relation mass of new heavy elements created and deposited in the interstellar and intergalactic medium by each generation τ = τ +2M 2∆Λ (31) of stars. h iN h iU P 10

1.0 1.0 Star Formation

Terrestrial Planet Formation Today 0.8 0.8 Emergence of Intelligent Life

0.6 0.6 Recombination

0.4 0.4

Today Normalised Probability Normalised Probability 0.2 0.2 Emergence of Intelligent Life

0.0 0.0 0 10 20 30 40 50 40 41 42 43 44

Age @GyrD log10HdU rpL

FIG. 1. Normalised probabilities for star formation, formation of terrestrial planets, and emergence of intelligent life. Left: Prob- abilities as a function of the age of our Universe. Right: Probability of the emergence of life as a function of a dimensionless measure of cosmic time, here the orders of magnitude diﬀerence between the evolving size of the Universe to the particle horizon normalised to the size of the proton as the basic building block of atoms, molecules, and eventually life.

The star formation rate is empirically given by [52] life. More specifically, this compares the maximal Comp- ton wavelength of a particle that could still be resolved 2.7 dρ∗ 0.015(1 R)(1 + z) within the horizon to the proton size. We show the emer- = − M yr−1Mpc−3 , (33) dt 1 + [(1 + z)/2.9]5.6 ⊙ gence probability for life as a function of this quantity, corresponding to the ratio of the diameter to the parti- where ρ∗ is the cosmological energy density in stars, in- cle horizon dU to the proton radius rp in Fig. 1. The cluding remnants. The cosmic metallicity is determined observer formation probability peaks close to 42 orders by Zb(z)= yρ∗(z)/ρb, where ρb is the total baryonic mat- of magnitude difference between the two. An interest- ter density. The formation of terrestrial planets can be ing question would now be to determine this number (or modelled following Ref. [53] with a formation probability that of other dimensionless measures of cosmic time) for −4 −3 −4 of PFTP k(Z), where k(Z) = (Z 10 )/(10 10 ) different universes with different cosmological and fun- ∝ − − interpolates between 0 and 1. For simplicity we shall damental parameters. We leave it to future work to in- adopt Z Zb. Finally, for the time span between the vestigate whether the emergence of life in the multiverse ≈ formation of a terrestrial planet and the emergence of always peaks around 42 or if this singles out a special conscious life forms as observers, we shall adopt 4.5 bil- case, and hence determine whether this number in some lion years, motivated by our own evolution on Earth. sense may indeed qualify well as the answer to the Ulti- Combining these relations, we can determine the observer mate Question of Life, the Universe and Everything [54]. formation probability, or the emergence probability of intelligent life, throughout cosmic history. We show our results in Fig. 1. An important conclusion that can be drawn from these results is that from an anthropic stand- V. GRAVITON LOOPS point there is no coincidence or Why Now? problem as we live close to the observer formation peak. Finally, we shall conclude our exploration of novel as- Finally, we may ask ourselves the question whether pects of the self-tuning mechanism in this paper with an life may form in other patches of the multiverse with analysis of the effect of graviton loops on the self-tuning different configurations of cosmological or even funda- property. It has been shown in Refs. [12, 13] that the mental parameters. In that case, rather than working self-tuning mechanism is a consistent semi-classical for- with timescales that differ between different universes, malism, very similar to what is the case for the seques- we should opt for a sensible dimensionless parameter to tering scenario (see Refs. [10, 11]). In the semi-classical describe the evolution of a universe and its relation to regime, one considers the gravitational field as classical, the formation of life. There are many options available. while the matter sector is treated quantum-mechanically. One choice one can take is to quantify this evolution in Then, quantum corrections give contributions to the cos- terms of a comparison of the evolving size of the cosmo- mological constant. However, if we also consider quan- logical particle horizon to the size of the proton as the tum effects in gravity, we will also find corrections to basic building block of atoms, molecules, and eventually the gravitational coupling. In particular, 1PI matter and 11 graviton loops modify the cosmological constant, while which includes the RGB term coupled to a local field θ(x). pure gravity loops are responsible for the corrections to The topological sector SA of this theory consists of the 4 the gravitational coupling. An exhaustive analysis of the term σ(Λ/µ ) Fµνρσ as well as two more coupled three- different effects of the quantum corrections and the ul- form gauge-fields, one coupled via a function of θ(x), i.e., 2 traviolet sensitivity of sequestering, both in the Jordan σˆ(θ) Fˆµνρσ , and the other via a function of κ (x), i.e., and Einstein frames, is presented in Ref. [55]. In the 2 2 ˜ 2 σ˜(κ (x)/MP )Fµνρσ . The functions κ (x), Λ(x), and θ(x) case of the sequestering model, it has been shown that are again auxiliary rather than ordinary fields. The vari- to cope with graviton loops the local model must be ex- ations are now performed with respect to gµν , the three tended [56]. We shall briefly review this discussion and gauge fields as well as the three auxiliary fields. Recall explore its extension to the self-tuning mechanism. that the auxiliary fields become rigid parameters after integrating out the three-form sectors. The additional variations with respect to θ(x) and the associated three- A. The Gauss-Bonnet invariant form yield and additional constraint on the Gauss-Bonnet invariant in terms of the three-form fluxes, RGB = The fact that loops involving graviton virtual lines are 4 ′ ′ 4 αβγρ h i µ (ˆσ cˆ)/(σ c), where c = M d x√ gε Fαβγρ and explicitly Planck mass dependent contributions invali- c−ˆ is defined analogously. Using this additional− constraint, dates the vacuum energy sequestering in its original for- it has been shown that vacuumR energy contributions can mulation [11, 56]. Recall that we are required to perform be completely sequestered from gravity, the loops with a variation with respect to the gravitational coupling, internal gravitons inclusively [56]. 2 2 i.e., MP (or κ in Ref. [56]). Similarly, the self-tuning in its original form is expected to fail in the presence of The addition of a coupled Gauss-Bonnet invariant is graviton loops. In sequestering, the variation with re- the minimal modification of the sequestering formalism 2 that also consistently accounts for the pure gravity loops. spect to MP in the global limit is such that R is fixed by a radiatively stable ratio of the two three-formh i fluxes Similarly, this term can be invoked in the self-tuning introduced in the formalism, which is just a linear re- mechanism to prevent the gravitation of vacuum en- lationship between R and a combination of the three- ergy beyond the semi-classical limit. Hence, the most form fluxes. For self-tuning,h i this instead involves a single straightforward extension of the original formulation of three-form flux. As was shown in Ref. [56] any other self-tuning to accomplish that is given by the contribu- curvature invariant that is not constructed purely out of tion of θRGB to the gravitational Lagrangian and the the Weyl tensor and the traceless part of the Ricci tensor modification of the topological sector of the form turns out to take care of the sequestering of the graviton loops. This yields a constraint that is a polynomial in the difference Λ T /4. Moreover, when the varia- 1 4 µνρσ ˆ − h i 2 SA = d xǫ σ(ϕ) Fµνρσ +ˆσ(θ) Fµνρσ . tional constraint is not explicitly dependent on MP , all 4! M of the Planck mass dependent contributions cancel from Z h i (35) the residual energy-momentum tensor, at least to leading with variations with respect to ϕ, θ, the two coupled order [55, 56]. three-forms, and the metric. The combination of the The minimal candidate to accomplish this task is the two resulting constraints is crucial for preventing vacuum Gauss-Bonnet curvature invariant, defined by RGB = energy contributions from gravitating, with the residual 2 2 2 Rµνρσ 2 Rµν + R . The contribution of this term to cosmological constant now also being radiatively stable the action− is trivial in four dimensions as it can be re- against pure gravity loop corrections. cast as a total derivative that we can integrate to zero. In this sense, the addition of RGB to the action merely introduces a topological modification that does not affect the local physics at finite wavelength. The fact that (the square of) the Ricci scalar appears explicitly in its definition is what ensures that a constraint is obtained B. Gauss-Bonnet term in Horndeski gravity that picks the correct counterterms to sequester all the next-order curvature corrections, i.e., the leading-order graviton loops. Hence, the general local sequestering ac- It is worthwhile to ask whether the sequestering or tion reads [56] self-tuning mechanisms in the presence of graviton loops that as described in Sec. V A require three or two sets κ2(x) S = d4x√ g R + θ(x) R of scalar fields, three-forms, and couplings, respectively, − 2 GB ZM " could not be simplified to one set only. In this respect it is interesting to note that the coupled Gauss-Bonnet term µν Λ(x)+ m(g , Ψm) , can be cast as a Horndeski scalar-tensor theory [34, 40]. − L # More specifically, let us consider the modification of the (34) action (1) by the addition of a non-minimally coupled 12

Gauss-Bonnet term, the quantum effects would then be canceled. The addition of the non-minimally coupled RGB term may then be viewed like adding zero, just as for sequestering in 4 ϕ S = d x√ g R V (ϕ) Eq. (34). Finally, by integrating out the three-form sec- M − " 2 − 2 Z tor and setting ϕ MP , the effective field equations reduce to those of General→ Relativity. + θ(ϕ) RGB + m(gµν , Ψm) , L # (36) VI. CONCLUSIONS where the coupling θ(ϕ) shall be an arbitrary smooth function of ϕ. It can be shown that the new term in the The simple additional variation of the General Rel- Lagrangian density, θ(ϕ) RGB, is dynamically equivalent ativistic Einstein-Hilbert action with respect to the to the following choice of Horndeski terms [40], quadratic Planck mass was recently proposed as a remedy to the Cosmological Constant Problem, and a local (4) 2 scalar-tensor formalism was subsequently developed as GB, 2 =8 θ X (3 ln X) , (37) L − counterpart to the originally global framework. In this (3) , =4 θ X(7 ln X)ϕ , (38) LGB 3 − paper, we have explored and illuminated further aspects (2) GB, 4 =4 θ X(2 ln X)R of the self-tuning mechanism arising from the Planck L − mass variation. (2) 2 ;µν +4θ (1 ln X) (ϕ) + ϕ;µν ϕ , We have first presented a brief review of both the local − h i (39) and global self-tuning formalisms as well as the differ- (1) ;µν ent aspects of the mechanism that have previously been , = 4 θ ln X Gµν ϕ LGB 5 − studied such as the degravitation of the Standard Model 2 (1) 1 3 ν α µ ;µν vacuum energy, the remedy of the coincidence problem, θ (φ) +2ϕ ϕ ϕ 3ϕ µν ϕ ϕ , − 3 X ;µ ;ν ;α − ; and the embedding of the local formalism into Horndeski h (40)i gravity. Previously, the formalism has only been studied in the where we have defined θ(n) ∂nθ/∂ϕn. Furthermore, Jordan frame, where matter fields follow geodesics of ≡ since the sum of Horndeski theories remains a Horndeski the metric but the metric satisfies a modified Einstein theory, Eq. (36) can always be cast in the form of Eq. (9) equation. Here, we have shown that self-tuning operates for any Horndeski generalization of the scalar-tensor sec- equivalently in the conformally related Einstein frame, tor in Eq. (36) [34, 40, 57, 58]. where metrics satisfy the Einstein field equations but do Importantly, however, some extra care must be taken not describe the geodesics of matter fields. In this pro- in the inclusion of the Gauss-Bonnet invariant when cess, we have also further clarified and highlighted the the topological sector of Eq. (2) is present. The con- similarities and differences between the self-tuning and straint ∂µ ϕ(x) = 0 it imposes forces the kinetic term sequestering mechanisms. In brief, self-tuning requires 2 X = (1/2) (∂µϕ) to vanish such that the terms with one less additional variation of an extra field in the ac- ln X or− 1/X may spoil the viability of the theory. How- tion to operate. Specifically, in the global approach these ever, cases do exist where the terms GB, i vanish despite are the Planck mass in the Jordan frame or the matter L 2 ∂µ ϕ(x) 0. Whereas terms like X ln X and X ln X coupling λ in Einstein frame. In the local formalism, self- vanish trivially→ in this limit, others involve the product of tuning operates with the variation of the scalar field and ln X and 1/X with terms that contain second covariant the coupled three-form gauge field. Unlike for seques- derivatives of the field. One way to achieve the vanishing tering, the cosmological constant is not separately var- of , i is then to impose an extra condition on the sec- ied, and the attached three-form in the local case is not LGB ond derivatives such that ∂µ ∂ν ϕ(x) 0. Explicit exam- required. But despite introducing less auxiliary fields, → ples with GB, i 0 can then be found, for instance, by self-tuning embeds the effective field equations of seques- adopting theL unitary→ gauge or assuming spherical sym- tering for specific forms of the Planck mass or scalar metry. field dependence of the vacuum energy and the classi- Hence, we conclude that a Horndeski scalar-tensor the- cal counterterm. Importantly, unlike sequestering, self- ory of a single scalar field amended with the coupled tuning generally degravitates the vacuum energy not by topological sector of Eq. (2) can indeed be sufficient to its cancelation between the geometric and material sides maintain the self-tuning mechanism in the presence of of the Einstein field equations, which only occurs in the graviton loops. Importantly, however, this is provided limit it effectively recovers the sequestering mechanism, that quantum gravity effects arise in the form of a non- but by the self-tuning of the classical counterterm due minimally coupled higher-order curvature term, or, in to the constraint equation imposed from the extra varia- other words, that they have a global dependency on the tions of the action. gravitational coupling. For those certain cases in which We have then extended the self-tuning mechanism to the scalar field and the coupling are of adequate form, in principle incorporate a dynamical inflationary epoch 13 at early times. Specifically, by introducing a step func- building blocks for atoms, molecules, and eventually life. tion in the three-form sector we allowed the scalar field For our Universe we found a peak at approximately 42 2 ϕ to behave dynamically in a regime ϕ

[1] A. G. Riess et al. (Supernova Search Team), [7] M. Henneaux and C. Teitelboim, The cos- Observational evidence from supernovae for mological constant and general covariance, an accelerating universe and a cosmolog- Physics Letters B 222, 195 (1989). ical constant, Astron. J. 116, 1009 (1998), [8] J. D. Barrow and D. J. Shaw, New solu- arXiv:astro-ph/9805201 [astro-ph]. tion of the cosmological constant problems, [2] S. Perlmutter et al. (Supernova Cosmology Project), Phys. Rev. Lett. 106, 101302 (2011). Measurements of Ω and Λ from 42 high red- [9] D. J. Shaw and J. D. Barrow, Testable solution of shift supernovae, Astrophys. J. 517, 565 (1999), the cosmological constant and coincidence problems, arXiv:astro-ph/9812133 [astro-ph]. Phys. Rev. D 83, 043518 (2011). [3] S. Weinberg, The Cosmological Constant Problem, [10] N. Kaloper and A. Padilla, Sequestering the stan- Rev. Mod. Phys. 61, 1 (1989), [,569(1988)]. dard model vacuum energy, Phys. Rev. Lett. 112, [4] J. Martin, Everything You Always Wanted 10.1103/PhysRevLett.112.091304 (2014). To Know About The Cosmological Con- [11] N. Kaloper, A. Padilla, D. Stefanyszyn, and stant Problem (But Were Afraid To Ask), G. Zahariade, Manifestly local theory of vac- Comptes Rendus Physique 13, 566 (2012), uum energy sequestering, Phys. Rev. Lett. 116, arXiv:1205.3365 [astro-ph.CO]. 10.1103/PhysRevLett.116.051302 (2016). [5] W. G. Unruh, Unimodular theory of canonical quantum [12] L. Lombriser, On the cosmological constant problem, gravity, Phys. Rev. D 40, 1048 (1989). Physics Letters B 797, 134804 (2019). [6] M. Henneaux and C. Teitelboim, The cos- [13] D. Sobral Blanco and L. Lombriser, Local self- mological constant as a canonical variable, tuning mechanism for the cosmological constant, Physics Letters B 143, 415 (1984). Phys. Rev. D 102, 043506 (2020). 14

[14] K. Koyama, Cosmological Tests of Modified [32] R. Bousso and J. Polchinski, Quantiza- Gravity, Rept. Prog. Phys. 79, 046902 (2016), tion of four-form fluxes and dynamical neu- arXiv:1504.04623 [astro-ph.CO]. tralization of the cosmological constant, [15] A. Joyce, L. Lombriser, and F. Schmidt, Journal of High Energy Physics 2000, 006 (2000). Dark Energy Versus Modified Gravity, [33] G. W. Horndeski, Second-order scalar-tensor Ann. Rev. Nucl. Part. Sci. 66, 95 (2016), field equations in a four-dimensional space, arXiv:1601.06133 [astro-ph.CO]. International Journal of Theoretical Physics 10, 363 (1974). [16] M. Ishak, Testing General Relativity in [34] S. Jana, C. Dalang, and L. Lombriser, Horn- Cosmology, Living Rev. Rel. 22, 1 (2019), deski theories and beyond from higher dimen- arXiv:1806.10122 [astro-ph.CO]. sions, Class. Quant. Grav. 38, 025003 (2021), [17] N. Aghanim et al. (Planck), Planck 2018 re- arXiv:2007.06907 [gr-qc]. sults. VI. Cosmological parameters (2018), [35] A. Davidson and S. Rubin, Zero cosmologi- arXiv:1807.06209 [astro-ph.CO]. cal constant from normalized general relativity, [18] T. Tröster et al. (KiDS), KiDS-1000 Cosmol- Classical and Quantum Gravity 26, 235019 (2009). ogy: constraints beyond flat ΛCDM (2020), [36] A. Davidson and S. Rubin, Normalized general relativ- arXiv:2010.16416 [astro-ph.CO]. ity: Nonclosed universe and a zero cosmological constant, [19] B. P. Abbott et al. (LIGO Scientific, Virgo, Fermi-GBM, Phys. Rev. D 89, 024036 (2014). INTEGRAL), Gravitational Waves and Gamma-rays [37] A. A. Tseytlin, Duality-symmetric string the- from a Binary Neutron Star Merger: GW170817 ory and the cosmological-constant problem, and GRB 170817A, Astrophys. J. 848, L13 (2017), Phys. Rev. Lett. 66, 545 (1991). arXiv:1710.05834 [astro-ph.HE]. [38] N. Arkani-Hamed, S. Dimopoulos, G. Dvali, and [20] L. Lombriser and A. Taylor, Breaking a G. Gabadadze, Nonlocal modification of grav- Dark Degeneracy with Gravitational Waves, ity and the cosmological constant problem (2002), JCAP 1603 (03), 031, arXiv:1509.08458 [astro-ph.CO]. arXiv:hep-th/0209227. [21] L. Lombriser and N. A. Lima, Chal- [39] G. Gabadadze and S. Yu, Dark energy and lenges to Self-Acceleration in Modified Grav- inflation with a volume normalized action, ity from Gravitational Waves and Large- Phys. Rev. D 98, 024047 (2018). Scale Structure, Phys. Lett. B765, 382 (2017), [40] T. Kobayashi, M. Yamaguchi, and J. Yokoyama, arXiv:1602.07670 [astro-ph.CO]. Generalized G-Inflation: —Inflation with the [22] S. Weinberg, Anthropic bound on the cos- Most General Second-Order Field Equations—, mological constant, Phys. Rev. Lett. 59, Progress of Theoretical Physics 126, 511 (2011). https://doi.org/10.1103/PhysRevLett.59.2607 (1987). [41] D. Langlois, Degenerate higher-order scalar-tensor [23] L. Lombriser and V. Smer-Barreto, Is there (dhost) theories (2017), arXiv:1707.03625 [gr-qc]. another coincidence problem at the reioniza- [42] L. Lombriser, Late-time acceleration by a resid- tion epoch?, Phys. Rev. D 96, 123505 (2017), ual cosmological constant from sequestering vac- arXiv:1707.03388 [astro-ph.CO]. uum energy in ultimate collapsed structures, [24] T. Padmanabhan, Gravity and quantum theory: Do- Journal of Cosmology and Astroparticle Physics 2019 (09), 065. mains of conflict and contact, International Journal [43] L. M. Krauss and G. D. Starkman, Life, the universe, and of Modern Physics D 29, 10.1142/S0218271820300013 nothing: Life and death in an ever-expanding universe, (2020). The Astrophysical Journal 531, 22 (2000). [25] S. Nobbenhuis, Categorizing different approaches to [44] L. M. Krauss and R. J. Scherrer, The return the cosmological constant problem, Found. Phys. 36, of a static universe and the end of cosmology, https://doi.org/10.1007/s10701-005-9042-8 (2006). General Relativity and Gravitation 39, 1545 (2007). [26] J. Polchinski, The Cosmological Constant and the String [45] F. C. Adams, M. T. Busha, A. E. Evrard, and Landscape, in The Quantum Structure of Space and R. H. Wechsler, The asymptotic structure of space-time, Time: Proceedings of the 23rd Solvay Conference on International Journal of Modern Physics D 12, 1743 (2003). Physics. Brussels, Belgium. 1 - 3 December 2005 (2006) [46] E. Gaztañaga, The cosmological constant arXiv:hep-th/0603249 [hep-th]. and the size of the physical universe (2019), [27] R. Bousso, TASI Lectures on the Cosmological Constant arXiv:1904.08218 [physics.gen-ph]. (2008), arXiv:0708.4231 [hep-th]. [47] E. Gaztañaga, Homogeneity and the causal boundary [28] A. Padilla, Lectures on the Cosmological Constant Prob- (2019), arXiv:1911.13199 [gr-qc]. lem (2015), arXiv:1502.05296 [hep-th]. [48] E. Gaztañaga, The size of our causal universe, [29] H. N. A. Aurilia and P. K. Townsend, Hid- Monthly Notices of the Royal Astronomical Society 494, 2766 (2020). den constants: The θ parameter of QCD and [49] P. Fosalba and E. Gaztañaga, Explaining cosmological the cosmological constant of n = 8 supergravity, anisotropy: evidence for causal horizons from cmb data, Nuclear Physics B 176, 509 (1980). Monthly Notices of the Royal Astronomical Society 504, 5840–5862 (2021). [30] J. D. Brown and C. Teitelboim, Dynamical neutraliza- [50] A. H. Guth, Eternal inflation and its implications, Jour- tion of the cosmological constant, Physics Letters B 195, nal of Physics A: Mathematical and Theoretical 40, https://doi.org/10.1016/0370-2693(87)91190-7 (1987). 10.1088/1751-8113/40/25/s25 (2007). [31] J. D. Brown and C. Teitelboim, Neutraliza- [51] M. Tegmark, Parallel universes (2003), tion of the cosmological constant by mem- arXiv:astro-ph/0302131 [astro-ph]. brane creation, Nuclear Physics B 297, [52] P. Madau and M. Dickinson, Cosmic Star Formation https://doi.org/10.1016/0550-3213(88)90559-7 (1988). History, Ann. Rev. Astron. Astrophys. 52, 415 (2014), arXiv:1403.0007 [astro-ph.CO]. 15

[53] E. Zackrisson, P. Calissendorﬀ, J. Gonzalez, A. Ben- 10.1103/PhysRevLett.118.061303 (2017). son, A. Johansen, and M. Janson, Terrestrial planets [57] J. M. Armaleo, J. Osorio Morales, and O. San- across space and time, Astrophys. J. 833, 214 (2016), tillan, Gauss–Bonnet models with cosmological con- arXiv:1602.00690 [astro-ph.GA]. stant and non zero spatial curvature in D = 4, [54] D. Adams, The Hitchhiker’s Guide to the Galaxy (1979). Eur. Phys. J. C 78, 85 (2018), arXiv:1711.09484 [gr-qc]. [55] M. G. Lawrence and D. Seery, Ultra- [58] C. van de Bruck and C. Longden, Einstein-gauss- violet sensitivity of the cosmological se- bonnet gravity with extra dimensions (2018), quester, Phys. Rev. D 103, 064018 (2021), arXiv:1809.00920 [gr-qc]. arXiv:2007.07058 [hep-th]. [56] N. Kaloper and A. Padilla, Vacuum energy sequestering and graviton loops, Phys. Rev. Lett. 118,