Cosmological Attractors and Anisotropies in Two Measure Theories, Effective EYMH systems, and Off--Diagonal Inflation Models

Applying the anholonomic frame deformation method, we construct various classes of cosmological solutions for effective Einstein -- Yang-Mills -- Higgs, and two measure theories. The types of models considered are Freedman-Lema\^{i}tre-Robertson-Walker, Bianchi, Kasner and models with attractor configurations. The various regimes pertaining to plateau--type inflation, quadratic inflation, Starobinsky type and Higgs type inflation are presented.


Introduction
Over time, the Cosmological Constant Problem (CCP) has evolved from the "Old Cosmological Constant Problem" [1], where the concern was on why the observed vacuum energy density of the universe is exactly zero, to the present form pertaining to the evidence establishing the accelerating expansion of the universe [2]. One is therefore faced with the "New Cosmological Constant Problem" [3,4]. In other words, the problem has shifted from not why the CCP is exactly zero, but to why the vacuum energy density is so small. Various attempts to address the issue range from the conventional to the esoteric. Conventional field theoretic models are based on a single scalar field (quintessence) while the esoteric models involve tachyons, phantoms and K-essence. The latter may also admit multi scalar field configurations. Such models have also been supplemented further to take into consideration the recent observational data from Planck [5,6] and BICEP2 [7]. In all these models the inflationary paradigm [8,9] is the underlying theme, see also an opposite point of view in [10]. However, present data is insufficient to determine precisely what the initial conditions were that drove inflation. In addressing the present situation there are essentially two main approaches entertained. In one approach it is assumed that there is a basic mechanism driving to zero the vacuum energy but some "residual" interactions survive that slightly shift the vacuum energy density towards the presently observed small non-zero value. In the alternative approach it is assumed that the true vacuum energy will exactly be zero when the final state of the theory is reached and the present state pertaining to the small non zero vacuum energy density is the result of our universe having not reached that final state yet.
In this work, we will adapt the view point that the above two scenarios represent equally viable solutions to the CCP and both can be entertained naturally if one considers off-diagonal inhomogeneous cosmological solutions. Alternative constructs are also possible and are discussed in [9,11,12] in a different class of theories. As will be demonstrated, for certain well defined conditions, the models considered in this work can be treated as effective two measure theories (TMTs) studied in Refs. [13,14,15,16,17,18]. In these theories, the modified gravitational and matter field equations of TMTs generate effective Einstein -Yang-Mills -Higgs (EYMH) systems which can be solved in analytic form using geometric methods. The underlying principle of the geometric method is based on the anholonomic frame deformation method (AFDM) [19,20,21,22,23]. The main idea of the AFDM is to re-write equivalently Einstein equations, and various modifications of it, on a (pseudo) Riemannian manifold V in terms of an "auxiliary" linear connection D. This connection, together with the Levi-Civita (LC) connection ∇, is defined in a metric compatible form by a split metric structure g = {g αβ = [g ij , g ab ]}. In order to establish our notation, we take dim V = 4, with the conventional splitting of coordinates as 3 + 1, and the equivalent splitting as 2 + 2 respectively. The signature of the metric on V is taken to be (+, +, +, −). Indices i, j, k, ... take values 1, 2 while indices a, b, ... take values 3, 4 and the local coordinates are denoted by u α = (x i , y a ), or collectively as u = (x, y). 1 Quantities under consideration and with a left label (for instance, g D ) emphasize that the geometric object (D) is uniquely determined by g. Unless otherwise stated, Einstein's summation convention is assumed throughout with the caveat that upper and lower labels are omitted if this does not result in ambiguities. We emphasize that D contains nontrivial anholonomically induced torsion T relating to the underlying nonholonomic frame structure. Such a torsion field is completely defined by the metric and the nonholonomic (equivalently, anholonomic and/or non-integrable) distortion relations, when both the linear connections and the distortion tensor Z[T] are uniquely determined by certain well-defined geometric and/or physical principles. Physical models are constructed following the principle that all geometric constructions are adapted to a nonholonomic splitting with an associated nonlinear connection (N-connection) structure N = {N a i (u)} that splits into the Whitney sum consisting of the conventional horizontal (h) and vertical (v) components, where T V is the tangent bundle 2 . For such a splitting, all geometric constructions can be carried out equivalently with ∇ using the so-called canonical distinguished connection (d-connection), D. Here D is distinct from D. This linear connection is N-adapted, i.e. preserves under parallelism the N-connection splitting, and is uniquely determined (together with ∇) by the constraints g → ∇ : ∇g = 0; ∇ T = 0, the Levi-Civita connection; D : D g = 0; h T = 0, v T = 0, the canonical d-connection.
It is to be noted that in general, a d-connection D can equivalently split into the N-adapted horizontal (h) and vertical (v) components, respectively, as hD and vD, (or equivalently, as = ( h D, v D)). But such a splitting may not be compatible, (i.e., D g = 0, ) as it can carry arbitrary amount of torsion T, and hence is not subject to the aforementioned constraints depicted in (3). The advantage of the canonical d-connection D is that in this framework hatted Einstein equations result, Here the hatted Einstein tensor G and the effective source term Υ are defined in standard form following geometric methods and N-adapted variational calculus but for quantities (g, D) instead of the usual (g, ∇). The hatted Einstein equations decouple with respect to a class of N-adapted frames for various classes of metrics with one-Killing symmetry [21,22]. This allows us to integrate (4) in a very general form by generic off-diagonal metrics, metrics that otherwise can not be diagonalized in a finite spacetime region by coordinate transformations that are determined via a set of generating and integration functions depending on all spacetime coordinates and various types of commutative or noncommutative parameters 3 . Solutions thus determined describe various geometric and physical models in modified gravity theories with nontrivial nonholonomically induced torsion, T = 0, and generalized connections. As special cases, we extract LC-configurations and construct new classes of cosmological solutions in Einstein's gravity if we constrain the set of possible generating and integration functions to satisfy the following conditions, where metric g αβ (t) in the unprimed bases can be related to metric in the primed bases via frame transformations, i.e., g αβ (t) = e α ′ α e β ′ β g α ′ β ′ (t), where e α ′ α represents the tetrad frame field. For instance, g α ′ β ′ can be a Bianchi type metric, or a diagonalized homogeneous Friedmann-Lemaître-Robertson-Walker (FLRW) type metric. In general, g α ′ β ′ may not be a solution of any gravitational field equations but we shall always impose the constraint that it's nonholonomic deformation g αβ always is a solution of the hatted Einstein equation (4). In general, gravitational field equations (4) constitute a sophisticated system of nonlinear partial differential equations (PDEs) as opposed to the occurrence of ordinary differential equations (ODEs) in conventional general relativity. The AFDM, on the other hand, allows us to find new classes of solutions by decoupling the PDEs. We emphasize that in the AFDM approach advocated here, constraints of type (5) and/or (6) are to be imposed after the inhomogeneous g αβ (x i , y 3 , t) are constructed in general form. If the aforementioned constraints are imposed from the very beginning in order to transform PDEs into ODEs, a large class of generic off-diagonal and diagonal solutions will be compromised. The specific goal of this work is to apply the AFDM method and explicitly construct solutions in effective TMTs addressing attractors, acceleration, dark energy and dark matter effects in the new cosmological models.
This work is organized as follows. In section 2, we provide a brief introduction to the geometry of nonholonomic deformations in Einstein gravity and modifications that lead to effective TMTs. In such theories we shown how the gravitational and matter field equations can be decoupled and solved in very general offdiagonal forms for the canonical d-connection with constraints for LC-configurations. Section 3 is devoted to off-diagonal and diagonal cosmological solutions with small vacuum density. Also constructed and analysed are the off-diagonal inhomogeneous cosmological solutions with nonholonomically induced torsion. In section 4, we study the equivalence of effective TMTs with sources for nonlinear potentials and EYMH self-dual fields resulting in attractor type behaviour. In section 5, we analyze in explicit form how exact cosmological solutions with locally anisotropic attractor properties can be generated by deforming FLRW type diagonal metrics and off-diagonal Bianchi type cosmological models. Conclusions are presented in section 6.

Nonholonomic Deformations
For clarity, we elaborate upon our notation first. On a (pseudo) Riemannian manifold we prescribe an N-connection with horizontal(h) and vertical(v) decompositions (h and v splitting) (2)as (V, N). To this we associate structures of N-adapted local bases, e ν = (e i , e a ), and cobases, e µ = (e i , e a ), which are the following N-elongated partial derivatives and differentials, and e i = dx i , e a = dy a + N a i (u)dx i .
The frame basis e ν = (e i , e a ), satisfy the nonholonomy relations with nontrivial nonholonomy coefficients Such a basis is holonomic if and only if W γ αβ = 0. This is trivially satisfied in a coordinate basis if e α = ∂ α . As holonomic dual basis, we take e µ = du µ .
The geometric objects on V are defined with respect to the N-adapted frames (7), (8). These are referred to as distinguished objects or d-objects in short. A vector Y (u) ∈ T V is parameterized as a d-vector. Explicitly, Y = Y α e α = Y i e i +Y a e a , or Y = (hY, vY ), with hY = {Y i } and vY = {Y a }. Likewise, in this frame work, the coefficients of d-tensors, N-adapted differential forms, d-connections, and d-spinors are easily accommodated.
Any metric tensor g on V, defined as a second rank symmetric tensor, takes the following structure with respect to the dual local coordinate basis, Equivalently, g serves as the d-metric and in tensor product notation, is taken to be Linear connections on V are introduced in N-adapted and N-non adapted forms in the standard way. By definition, a d-connection D = (hD, vD) preserves under parallelism the N-connection splitting (2). Any dconnection D acts as covariant derivative operator, D X Y, for a d-vector Y in the direction of a d-vector X. With respect to N-adapted frames (7) and (8), we can compute the relevant quantities of interest in N-adapted coefficient form when D = {Γ γ αβ = (L i jk , L a bk , C i jc , C a bc )}. The coefficients Γ γ αβ are computed for the horizontal and vertical components of D eα e β := D α e β by substituting X for e α and Y for e β .
We compute the d-torsion T , the d-torsion nonmetricity Q, and the d-curvature R for any d-connection D from the following standard formulae, The N-adapted coefficients are correspondingly labeled as The Levi-Civita connection ∇ ( LC) and the canonical d-connection D defined by formulas (3) are also expressed in terms of the local N-adapted form. The coefficients of ) and are computed using the following formulae L i jk = 1 2 g ir (e k g jr + e j g kr − e r g jk ) , C a bc = 1 2 g ad (e c g bd + e b g cd − e d g bc ) By using the coefficients of ∇ = {Γ γ αβ }, written with respect to (7) and (8), we compute the coefficients of the distortion d-tensor Z γ αβ = Γ γ αβ − Γ γ αβ , which is the N-adapted coefficient formula for (1). We elaborate upon geometric and physical models in equivalent form by working with two metric compatible connections D and ∇ because all N-adapted coefficients for Z γ αβ = Γ γ αβ and Γ γ αβ are completely defined by the same metric structure g. The nontrivial d-torsions coefficients T γ αβ are computed by setting D = D in (13) and determined by the nonholonomy relations, Any (pseudo) Riemannian geometry is formulated on a nonholonomic manifold V using two equivalent geometric quantities, (g, ∇) or (g, N, D). In the "standard" method we take D → ∇ when ∇ T γ αβ = 0, ∇ Q γ αβ = 0, and ∇ R α βγδ is computed following formulae (14). For the "geometric variables" (g, N, D), using similar formulae, we compute D = D in standard form respectively the Riemann d-tensor R and the Ricci d-tensor Ric{= R βγ }. The nonsymmetric d-tensor R αβ of D is characterized by the following four h and v N-adapted coefficients and the "alternative" scalar curvature The Einstein d-tensor of D in hatted form is and is a nonholonomic distortion of the standard form, G αβ := R αβ − 1 2 g αβ R , that is computed from ∇. We solve the equations resulting from the constraints (5) and get solutions to a system of first order PDE equations Nonholonomic deformations of fundamental geometric objects on a pseudo-Riemannian manifold V with Nconnection 2+2 splitting are determined by the transforming of the fundamental geometric data (g,N, • D) → ( g, N, D), where the "prime" data (g,N, • D) may or not be a solution of certain gravitational field equations in a (modified) theory of gravity but the "target" data ( g, N, D) affirmatively define exact solutions of (4) with metrics parameterized in the form (11) and (21). The prime metric is parameterized as As an explicit example, we takeg to be a Friedman-Lemaître-Robertson-Walker (FLRW) type diagonal metric withN b i = 0. The target off-diagonal metric is of type (12) with e a taken as in (8). With additional parameterizations via the so-called gravitational "polarization" functions η α = (η i , η a ), the metric g takes the form In the special case in which η α → 1 and N a i =N a i , we get a trivial nonholonomic transformation (deformation). For the data ( g, D) the effective source for a scalar field φ and a gauge field Fǎ µν in modified gravitational interactions (4) is the energy-momentum tensor e T αβ where whereǎ is a internal group index. This tensor is constructed with respect to the N-adapted (co) frames (7), (8) following the same procedure as in Refs. [13,14,15,16,17,18], and Υ αβ = κ 2 e T αβ where κ is the gravitational constant. The explicit coordinate dependence for Υ is Elements with α = β are all taken to be zero. The effective nonlinear scalar potential e V is determined by two scalar potentials V (φ) and U (φ) as where M is a constant. The Einstein d-tensor G αβ is given in N-adapted form by formula (19). The resulting nonlinear system of PDEs can be integrated in explicit form for arbitrary parameterizations of type Υ β δ = diag[Υ α ]. 4 As a specific example, we take the TMT effective action studied in [13,14,16,17,18] for m L resulting in the energy-momentum tensor (22) and where R is the scalar curvature. Modified Einstein equations are derived in the light of LC-conditions (20). The energy-momentum tensor follows from variation in N-adapted form using the N-elongated partial derivatives and differentials, We consider a new 'scaled' d-metric g αβ where where e −2 σ is the scale factor determined in terms of the constant and potentials used in the effective potential e V (24). The function Φ = ε µναβ e µ A ναβ = ε µναβ ε abcd e µ ϕ a e ν ϕ b e α ϕ c e β ϕ d , with four scalar fields ϕ a , (a = 1, 2, 3, 4), defines the second measure in TMTs. The effective gravitational theory (25) with the source e T αβ (18) and re-scaling properties (26) is equivalent to the theory given by the following Fǎ µν Fǎ µν .
In the above, the N term 5 . It is a CP violating parameter and is determined to be very small from constraints from phenomenology. The non-Riemannian configuration is determined from the canonical d-connection Γ α βγ for g αβ .
Identifying the scalar indices as interior indices ("overline check") and varying (27) with respect to ϕǎ in N-adapted form, we obtain the equation A μ a e µ 1 L = 0.
The solution of this equation is e µ 1 L = 0, or 1 L = M = const. Thus for any M = 0, we obtain a spontaneous breaking of global scale invariance of the theory. This follows from the mismatch between the left hand side and the right hand side of the equation. If we fix M as an integration constant for the right hand side, the left hand side has a non-trivial transformation. In terms of the metric g αβ , the equation for the scalar filed becomes Not considering effective gauge interactions, i.e. for N = 0, we define the vacuum states for V + M = 0, where e V = 0 and d e V /dφ = 0 (it is also considered that d e V /dφ is finite and U = 0). We conclude that the basic feature of TMTs do not depend on the type of nonholonomic distributions on spacetimes if we work with metric compatible canonical d-connections or the LC connections. For both cases, we solve the 'old" cosmological constant problem, implying that the vacuum state with zero cosmological constant is achieved for different types of linear connections and without resort to fine tuning. Independently of wether we change the value of constant M, or add a constant to V, we still satisfy the conditions e V = 0 and d e V /dφ = 0 if V + M = 0. Here we also note that if we consider N = 0, it implies that an external source drives the scalar field away from such vacuum points and can be addressed in terms of instanton effects.
N-adapted variations with respect to g µν result in the equation where a =ǎ for this class of TMT theories. Additional constraints for LC-configurations when the equations (20) for the data (g, D[g]) are satisfied transform (31) into the system (17) in [15]. A small vacuum density determined by instantons was analyzed for LC-configurations of (30). It is a cumbersome task to find cosmological solutions of the system defined by equations (29) -(31). Nevertheless, it is possible to construct generic off-diagonal cosmological solutions for the systems of modified commutative and noncommutative Einstein -Yang -Mills -Higgs fields using the AFDM [22,27,19]. Our strategy is to find solutions for the theory (25) resulting in modified Einstein equations (4) with effective stress-energy tensor (22) and effective source (23). Metrics such as g αβ , in general, transform into g αβ for the theory (27) using N-adapted conformal transforms of type (26). We integrate in explicit form the equations (4) with a source (23) for the N-adapted coefficients of a metric g (21) parameterized in the form and supplementing with frame/coordinate transformations that satisfy the conditions h ⋄ a = 0, Υ 2,4 = 0. 6 For convenience, the partial derivatives ∂ α = ∂/∂u α are labeled as The nontrivial components of the Ricci and Einstein d-tensors are computed using the N-adapted coefficients of the canonical d-connection (15) for the metric ansatz (21) with data (32) for ω = 1 introduced respectively in (17), (18) and (19). Eventually, we arrive at the following system of nonlinear PDEs The torsionless (Levi-Civita, LC) conditions (5), (20), transform into The system of nonlinear PDE (33)-(36) posses an important decoupling property which admits step by step integration of such equations. To achieve this, first we introduce the coefficients where The coefficients serve as generating functions. For ∂ t h a = 0 and ∂ t ̟ = 0, 7 we rewrite the equations in the form The function ψ(x k ) are found by solving a two dimensional Poisson equation (40) for any prescribed source v Υ(x k ). The equations (39) and (41) convert any two functions to two others from a set of four, h a , ̟ and Υ. In one explicit form, h 3 and h 4 are determined for any prescribed ̟(x k , t) and Υ(x k , t). Once h a are determined, we integrate twice w.r.t t in (42) and find n i (x k , t). In the final step we solve for w i (x i , y a ) by solving a system of linear algebraic equations (43). The equation (44) is necessary to accommodate a nontrivial conformal (in the vertical "subspace") factor ω(x i , y a ) that depends on all four coordinates. For convenience, we shall use Ψ := e ̟ as our re-defined generating function. We conclude section 2 with the following remarks. We have shown that TMT theories as determined by actions of type (27) can be formulated in nonholonomic variables as effective EYMH systems with modified Einstein field equations (4). This allows one to apply the AFDM and decouple such systems of nonlinear PDEs in very general form and write them equivalently as systems of type (40)- (44). This procedure and the resulting equations provide important results for mathematical cosmology. For instance, by considering the coordinate y 4 = t to be time like, one can show that TMT theories and other modified gravity models can be integrated in general forms.

Off-Diagonal Cosmological Solutions with Small Vacuum Density
In this section we provide a series of examples of new classes of exact solutions of modified Einstein equations with (non) homogeneous cosmological configurations constructed by applying the AFDM. We emphasize that all solutions generated in this section will be for a TMT theory with sources (22) parameterized in the form (23), when the effective nonlinear scalar potential is taken in the form (24). In a similar form, we can construct solutions with effective sources for other types of modified gravity theories like in [26,49].
For any ∂ t ̟ = 0, ∂ t h a = 0 and Υ = 0, we write (41) and (39) as Using Ψ := e ̟ and introducing the first equation into the second in (45), we obtain the relation is an integration function. We use the first equation in (45) and compute Formulae for h a are expressed in a more convenient form by considering an effective cosmological constant Λ 0 = const = 0 and a re-defined generating function, Ψ →Ψ, subject to the condition where the integration function 0 h 3 (x k ) from (46) is formally introduced either in Ψ or equivalently in Υ.
Our final results are and hold for an effective cosmological constant Λ 0 = 0 so that re-definition of the generating functions, Ψ ←→ Ψ, are unambiguous where The functional in the formula for h 4 in (48) is interpreted as a re-defined source Υ → Ξ for a prescribed generating function . Such effective sources contain information on effective matter field contributions in modified gravity theories. We work with the generating quantities, (Ψ, v Λ) and [ Ψ, Λ 0 , Ξ] related via formulae (49) in terms of the prescribed effective cosmological constant Λ 0 . The numerical value of Λ 0 is fixed to meet present day constraints from cosmology. Using formulae h a (48), we compute the coefficients α i , β and γ from (38). This allows us to find solutions to equations (42) by integrating two times with respect to t, and (43), solving a system of linear algebraic equations for w i . As a result, the N-coefficients are expressed recurrently as functionals (an example of which is [ Ψ, Λ 0 , Ξ]) and are as follows, where 1 n k (x i ) and 2 n k (x i ), or 2 n k (x i ), are integration functions with possible re-definitions by coordinate transforms. After a tedious calculation for g a = ω 2 (x k , y a )h a that involves the vertical conformal factor ω(u α ) depending on all spacetime coordinates, the vertical metric h a (48) and the N-coefficients N a i (50) reveals the fact that the formulae for the Ricci d-tensor R αβ (17) are invariant if the first order PDE (44) are satisfied. For nontrivial ω, the solutions to the modified gravitational equations, (4) parameterized as a d-metric (21), do not posses in general any Killing symmetries and contain dependencies of ω on [ψ, h a , n i , w i ] with as many as six independent variables for g αβ .
Putting together the solutions for the 2-d Poisson equation (40) and the formulae for the coefficients (48), (50) we conclude as our final result that the system of nonlinear PDEs (40)-(43) for non-vacuum 4-d configurations for the data (g, N, D), and with Killing symmetry on ∂ 3 when ω = 1, integrates to the line element Such inhomogeneous cosmological solutions with nonholonomically induced torsion are determined by ψ(x k ), Ψ(x k , t), ω(x k , y 3 , t), Ξ(x k , t) that depend on the effective cosmological constant Λ 0 and integration functions 1 n k , 2 n k . Straightforward computations reveal that, in general, the nonholonomy coefficients W γ αβ (10) are non vanishing. Therefore the class of solutions (51) can not be diagonalized in N-adapted form unless supplemented with additional assumptions on generating/ integration functions and constants. The nontrivial coefficients of the canonical d-torsion (13) are also non vanishing. They are determined by introducing the coefficients of the d-metric into N-adapted formulas (15) and then into T γ αβ (16). Let us prove that the zero d-torsion conditions (37) for LC-configurations can be solved in explicit form by imposing additional constraints on d-metrics (51). For the n-coefficients, such conditions are satisfied if 2 n k (x i ) = 0 and ∂ i 1 n j (x k ) = ∂ j 1 n i (x k ). In N-adapted form, such coefficients do not depend on generating functions and sources but only on a corresponding class of integration functions, e.g., 1 n j (x k ) = ∂ i n(x k ), for any n(x k ). It is a more difficult task to find explicit solutions for the LC-conditions (37) involving variables w i (x k ). Such nonholonomic constraints can not be solved in explicit form for arbitrary data (Ψ, Υ), or arbitrary (Ψ, Ξ, Λ 0 ). We first use the property that (50)). This results in the expression . The second step is to restrict our construction to a subclass of variables when H =Ψ[Ψ] is a functional which allows us to generate LC-configurations in explicit form. By taking h 3 [Ψ] =Ψ 2 /4Λ 0 (48) as a necessary type of functional H =Ψ = ln | h 3 |, we satisfy the condition e i ln | h 3 | = 0 in (37). Next, we solve for the constraint on Substituting in this formula the generating function Ψ =Ψ gives and we deduce that (47) withΨ, we arrive at In order to prove this formula we have used (52) and e iΨ = 0. From the last two formulae, we obtain This is possible for either Υ = const, or if Υ can be expressed as a functional Υ(x i , t) =Υ[Ψ]. If such conditions are not satisfied, we can re-scale the generating functionΨ ←→ Ψ, wherě We consider a functional in the formula for h 4 (48) (as a re-defined source,Υ → Ξ), for a prescribed generating function Ψ, wheň for any effective cosmological constant Λ 0 in order to satisfy such conditions. If we introduce a functionǍ =Ǎ(x k , t) for which Summarizing the results, we conclude that the linear quadratic line element where ω is a solution of and defines generic off-diagonal cosmological solutions with zero nonholonomically induced torsion. Such inhomogeneous cosmological solutions are determined by the generating functions and effective sources ψ(x k ), Ψ(x k , t), ω(x k , y 3 , t), Ξ(x k , t), the parameter Λ 0 , and the integration functions 1 n i = ∂ i n(x k ) respectively. The main result of this section is the demonstration that TMT theories admit generic off-diagonal cosmological solutions of type (51), with nontrivial nonholonomically induced torson, or of type (54), for LC-configurations. Another fundamental physical result is the emergence of a nonlinear symmetry for generating functions, see formula (49), for cosmological solutions of such nonlinear systems which allows to transform arbitrary effective and matter fields sources into an effective cosmological constant Λ 0 treated as an integration parameter. The value of the integration parameter can be fixed by getting compatibility with observational cosmological data.

Time like parameterized off-diagonal cosmological solutions
In this section we consider a subclass of solutions pertaining to g αβ (x k , y 3 , t) extracted from either (51), that depend only on time like coordinate t. For applications in modern cosmology, we consider g α ′ β ′ (t) as certain offdiagonal deformations of the FLRW, or the Bianchi type Universes [21,26]. In explicit form, we construct physical models withǵ (21). The strategy is first to construct solutions for a class of generating functions and sources with spacetime dependent coordinates and then to restrict the integral varieties to configurations with dependencies only on the time like coordinate. This procedure requires that

Cosmological solutions for the effective EYMH systems and TMT
The effective gravitational theory (25) with source e T αβ (18) in TMTs describes a nonlinear parametrical interacting EYMH system where we interpret φ as a Higgs field that can carry internal indices and acquire vacuum expectation φ [0] , and couple to the gauge field A = A µ e µ with values in non-Abelian Lie algebra. On the premises defined by the nonholonomic V, the d-operator D µ is elongated additionally to accommodate the gauge potentials in the form D µ = D µ + ie[A µ , ], where the commutator [., .] signifies the non-Abelian structure. The gauge coupling is e and i 2 = −1. The gauge field A µ enters the covariant derivative D µ = e µ +ie[A µ , ] and the "curvature" where the boldface F βµ is used for N-adapted constructions. 8 With respect to N-adapted frames the nonholonomic EYMH equations, postulated either by following geo- 8 For standard gauge field models but on nonholonomic manifolds we can follow a variational principle for a gravitating non-Abelian SU(2) gauge field A = Aµe µ coupled to a triplet Higgs field φ. In such cases, the value φ [0] is the vacuum expectation of the Higgs field which determines the mass H M = √ λη, when λ is the constant of scalar field self-interaction with potential where the trace T r is taken on internal indices. In EYMH theory, the gravitational constant G, κ = 16πG, defines the Plank mass M P l = 1/ √ G and it is also the mass of gauge boson, W M = ev. In the literature, various versions of modified gravity and TMTs are elaborated upon with different types of nonlinear scalar and gauge fields. metric principles, or "derived" following an N-adapted variational calculus from (25), are the following, where the source (23) is determined by the stress-energy tensor The nonlinear potential e V (φ) is as in (59) for a TMT if it is taken in the form (24).
The system of nonlinear PDEs (56)-(58) posses a similar decoupling property as in (4) if plausible assumptions are made for gravitational and matter field interactions. To see this and construct new classes of modified EYMH equations we take the "prime" solution to be given by data for a diagonal d- (2) gauge field configurations, the diagonal ansatz for generating solutions can be written in the form where the coordinates and metric coefficients are parameterized respectively as u α = ( , and Λ is a cosmological constant. The function m(r) is interpreted as the total mass within the radius r for which m(r) = 0 defines an empty de Sitter space written in a static coordinate system with a cosmological horizon at r = r c = 3/Λ. The solution of (56) associated to the quadratic metric line element (61) is defined by a single magnetic potential ω(r), where τ 1 , τ 2 , τ 3 are Pauli matrices. The corresponding solution of (58) is given by Explicit values for the functions σ(r), q(r), ω(r), ̟(r) have been found in Ref. [28] for ansatz (61), (62) and ] define physical solutions with diagonal metrics depending only on the radial coordinate. A typical example is the well known diagonal Schwarzschild-de Sitter solution (56)-(58) that is given by and defines a black hole configuration inside a cosmological horizon because q(r) = 0 has two positive solutions and M < 1/3 √ Λ. The conditions for nonholonomic deformations of (61) are as follows. The "target" d-metric η g with nontrivial N-coefficients, for • g → g is parameterized as in (21). The gauge fields are deformed as where • A µ (x 1 ) is of the type (62) and η A µ (x i , y a ) are functions for which where s is a constant and ε βµ is the absolute antisymmetric tensor. The gauge field curvatures F βµ , • F βµ and η F βµ are computed by substituting (62) and (64) into (55). Any antisymmetric F βµ (65) is a solution of D µ ( |g|F µν ) = 0, i.e. determines η F βµ , η A µ , for any given • A µ , • F βµ . For nonholonomic modifications of scalar fields, we take This nonholonomic configuration of the nonlinear scalar field is non-trivial even with respect to N-adapted frames e V (φ) = 0 and F T βδ = 0, (59). For ansatz (21), the equations (66) are A nonholonomically deformed scalar (Higgs field depending in non-explicit form on two variables because of constraint (66)) modifies indirectly the off-diagonal components of the metric via n i , w i and the above conditions The effective gauge field F βµ (65) with the potential A µ (64) modified nonholonomically by φ and subject to the conditions (66) determine exact solutions of the system (31) if the spacetime metric is chosen to be in the form (21). The energy-momentum tensor is determined to be F T α β = −4s 2 δ α β [29]. Interacting gauge and Higgs fields, with respect to N-adapted frames, result in an effective cosmological constant s Λ = 8πs 2 which should be added to the respective source (23).
To conclude, a generic off-diagoanal ansatz g = [η i (21) and (effective) gauge-scalar configurations (A, φ) subject to conditions mentioned above define a decoupling of the nonlinear PDEs (56)-(58) if the sources (23) are transformed in the form This is in sharp contrast to the situation where with respect to coordinate frames, such systems of equations describe a very complex, nonlinearly coupled gravitational and gauge-scalar interactions.

Effective vacuum EYMH configurations in TMTs
The effects of off-diagonal gravitational, scalar and gauge matter fields result in driving the vacuum energy density to zero even when the effective source Υ α and cosmological constant Λ 0 are nontrivial. This is possible due to the contributions of effective self-dual gauge fields. Such an effect is discussed in [15] for instantons. If Υ β δ = 0 in (67), one imposes further nonholonomic rescaling Υ →Λ 0 when Λ 0 − 4s 2 = 0. We can generate a very large class of solutions in TMTs with effective EMYH interactions into nonholonomic vacuum configurations of modified Einstein gravity. In this section, we analyze a subclass of generic off-diagonal EYMH interactions which can be encoded as effective vacuum Einstein manifolds of various class and lead to solutions with nontrivial cosmological constant Λ 0 = 4s 2 . In general, such solutions depend parametrically on Λ 0 − 4s 2 and do not have a smooth limit from non-vacuum to vacuum models. Effects of this type exist both in commutative, noncommutative gauge gravity theories [27], Einstein gravity and its various modifications [19,22], and TMTs. Examples are provided in the following sections.
Einstein equations (40)-(44) corresponding to a system of nonlinear PDEs (56)-(58) with source Υ β δ (67) are, To derive self-consistent solutions of this system for Υ − 4s 2 = 0 we consider off-diagonal ansatz depending on all spacetime coordinates, where the coefficients of this target metric are defined by solutions of the the following equations, The coefficients β and α i are computed following formulae (38) for nonzero ∂ t ̟ and ∂ t h 3 . The coefficients h a , h 3 and w i are additionally subject to the zero-torsion conditions (5), (6) as in the form (20) where, for simplicity, we fix n i equal to a constant as a trivial solutions of (70). For equation (74), we can take ψ = 0, or consider a trivial 2-d Laplace equation with spacelike coordinates x k . There are two possibilities to satisfy the condition (75) and derive the corresponding off-diagonal solutions. In the first case we take h 3 = h 3 (x k ), when ∂ t h 3 = 0. This implies that the equation (75) has solutions with zero source for arbitrary function h 4 (x k , t) and arbitrary N-coefficients w i (x k , t) as follows from (38). For such vacuum LC-configurations, the functions h 4 and w i are general and should be constrained only by the conditions (20). This constrains substantially the class of admissible w i if h 3 depends only on x k (we can perform a similar analysis as in subsection 3). The corresponding quadratic line element is where we introduce a functionǍ =Ǎ(x k , t) for which w i = ∂ iǍ satisfies ∂ i w j = ∂ j w i in (37) and ω is a solution of In the second case a very different class of (off-) diagonal solutions result if we choose, after corresponding coordinate transformations, ̟ = ln ∂ t h 3 / |h 3 h 4 | = 0 ̟ = const and ∂ t ̟ = 0. For such configurations, we can consider ∂ t h 3 = 0 and solve (75) for with 0 h equals a non vanishing constant. Such v-metrics are generated by any f (x i , t) satisfying ∂ t f = 0, when where the signs are fixed in such a way that for N a i → 0 we obtain diagonal metrics with signature (+, +, +, −). The coefficients (38) for (76) became trivial if α i = β = 0, and w i (x k , t) is any functions solving (20). The last system of equations for the LC-conditions are equivalent to for any n i (x k ) when ∂ i n k = ∂ k n i . Constraints of type n k ∂ 3 h 3 = ∂ k h 3 have to be imposed for a nontrivial multiple h 3 depending on y 3 . The corresponding quadratic line element is where w i are taken to solve the conditions (80)with ∂ i w j = ∂ j w i and ω is a solution of We conclude that off-diagonal interactions in effective EYMH systems result in vanishing cosmological constant as is demonstrated in the general solutions (77) and (81) presented above for LC-configurations. Such constructions can be generalized to include inhomogeneous effective vacuum configurations with nontrivial nonholonomically induced torsion (16). Effects of this nature exist in TMTs when the analogous EYMH systems are described by an action with two measures (27) related to an action (25) via a N-adapted conformal transform (26). Additionally, a subclass of cosmological solutions satisfying the conditions (5) and (6) can be generated if, for instance, we restrict the generating functions in (81) to satisfy via frame/coordinate transforms , ω → 1 and the integration functions are changed into integration constants.

Examples of (off-)diagonal nonholonomic deformations of cosmological metrics
In this section, we present details on how AFDM are employed to construct a new class of inhomogeneous and anisotropic cosmological solutions with target d-metrics g (21) , with certain well defined limits for η α → 1, to a primed metric • g. These can be interpreted as conformal, frame or coordinate transformations of the well known metrics like FLRW, Bianchi, Kasner, or another metric corresponding to a particular cosmological solution [30,31,32,33].

Off-diagonal deformations of FLRW configurations in TMTs
We show how N-anholonomic FLRW deformations can be constructed to define three classes of generic offdiagonal cosmological solutions for modified EYMH systems in TMTs. Similar models for one measure theories are presented in in [21].
FLRW metrics: For convenience, we introduce the necessary notations to describe the primed standard FLRW metric, when written in the diagonal form, is with K = ±1, 0 and spherical coordinates x 1 = r, x 2 = θ, y 3 = ϕ, y 4 = t and For simplicity, we take K = 0 and choose Cartesian coordinates (x 1 = x, x 2 = z, y 3 = y, y 4 = t), when the coefficients of Fg are taken, respectively, in the form Fg 1 = Fg 2 = a 2 , Fh 3 = a 2 F h 4 = −1 and FN a i = 0. In this case, the nontrivial coefficients of the primed diagonal metric depend only on the time like coordinate t and takes the form, Here we also note that instead of FLRW we can consider any other 'primed' metric • g that can be a Bianchi, Kasner or a metric of a particular cosmological solution [31,34,35,36,37]. The metrics (82) and/or (83) define exact homogeneous cosmological solutions of equations (19) and (20) with source Υ αβ = κ 2 T αβ for a perfect fluid energy-momentum stress tensor, Here ρ and p are the proper energy density and pressure in the fluid rest frame. The Einstein equations corresponding to ansatz (82) take the form of two coupled nonlinear ODE (the Friedmann equations) and The Hubble constant H ≡ ∂ t a/a has the units of inverse time and is positive (negative) for an expanding (collapsing) universe. The equations (85) and (86) are related via the condition ∇ α T α β = 0, for which the considered diagonal homogeneous ansatz is written as Here we note that the strong energy conditions for matter, ρ + 3p ≥ 0, or equivalently, the equation of state, w = p/ρ ≥ −1/3, must be satisfied for an expanding universe.
Off-diagonal effective EYMH cosmological solutions of type 1: In this case the d-metric is of the type (51) with ∂ t h a = 0, ∂ t ̟ = 0 and Υ − 4s 2 = 0, when Fh 4 correspond to an effective cosmological constant Λ 0 − 4s 2 = 0 with re-defined generating functions, s Ψ ←→ s Ψ. The left label "s" emphasizes that such values encode contributions from effective gauge fields, where The functional in the formula for h 4 in (48) can be considered as a re-defined source, Υ − 4s 2 → s Ξ, for a prescribed generating function Ψ, when Υ − 4s 2 = ∂ t ( s Ξ)/∂ t ( s Ψ 2 ). Such effective sources contain information on effective EYMH interactions in TMTs. For convenience we work with a couple of generating data, ( s Ψ, v Λ − 4s 2 ) and [ s Ψ, Λ 0 − 4s 2 , s Ξ] related by formulae (87) for a prescribed effective cosmological constant Λ 0 and the parameter s for gauge fields. Such values have to be fixed in a form which is compatible with experimental/ observational data, and result in a small vacuum density. Summarizing the results for off-diagonal nonholonomic deformations of the prime metric (83), we get a quadratic line element where s ω(x, z, y, t) is a solution of (44) which for our data is written in the form For the N-connection coefficients, we have The function ψ(x, z) in (88) is a solution of (68), i.e. of ∂ 2 xx ψ + ∂ 2 zz ψ = 2(Υ − 4s 2 ).
To understand possible physical implications of d-metrics (88) it is more convenient to use the so-called polarization functions η α and η a i := N a i −N a i as in (21) and parameterize such solutions in the form where are determined by the above solutions for the coefficients of the target d-metric.
Solutions (89) describe general off-diagonal deformations of the FLRW metrics in TMTs encoding modified EYMH interactions. Such interactions may result in changing the topology and symmetries, and are characterized by inhomogeneous, locally anistropic configurations or non-perturbative effect. The problem of physical interpretation of such cosmological off-diagonal solutions is simplified to some extent if we consider small deformations with polarizations of the type η α ≈ 1 + χ α and η a i ≈ 0 + χ a i for small values |χ α | ≪ 1 and |χ a i | ≪ 1, by which we obtain small deformations of the FLRW universes by certain generalized two measure interactions and/or modified gravity theories with effective EYMH fields. Nevertheless, even in such cases the target configuration may encode nonlinear and nonholonomic parametric effects as results of re-scaling (87) of generating functions. This way we model nonlinear nonholonomic transformations of a FLRW universe into an effective and small-deformed one with small values of effective cosmological constant, nonlinear anisotropic processes and other effects of similar magnitude.
Off-diagonal cosmological solutions of type 2 and "losing" information on effective EYMH: This class of solutions are characterized by the condition ∂ t h 3 = 0. The equation (69) can be solved only if Υ−4s 2 = 0, i.e. when the contributions from effective YM fields compensate other (effective) modified gravity and/or matter field sources. We take the function w i (x k , t) as a solution of (71), or its equivalent (76), because the coefficients β and α i from (38) are zero. To find nontrivial values of n i we integrate (70) for ∂ t h 3 = 0 for any given h 3 and find n i = 1 n k x i + 2 n k x i h 4 dt. Also, we take g 1 = g 2 = e ψ(x k ) , with ψ(x k ) determined by (68) for a given source (Υ − 4s 2 ).
In summary, this class of solutions can be chosen to be defined by the ansatz for arbitrary generating functions h 4 (x k , t), w i (x k , t), 0 h 3 (x k ) and integration functions 1 n k x i and 2 n k x i . In general, such solutions carry nontrivial nonholonomically induced torsion (16). The conditions (20) constrains (90) to a subclass of LC-solutions resulting in the following equations for any w i (x k , t) and 0 h 3 (x k ). This class of constraints on solutions (90) do not involve the generating function h 4 (x k , t) but only the N-connection coefficients for a prescribed value 0 h 3 (x k ).
Another metric to consider is the prime FLRW metric as in (82) and/or (83) and repeat the constructions for the metric (89) but with the difference that we take ∂ t h 3 = 0. However, we study here another possibility, i.e., to begin with a prime metric which is not a solution of gravitational field equations and finally to generate off-diagonal cosmological solution with effective nontrivial nonholonomic vacuum configuration. Let us consider • g i = 1, • h 3 = 1, • h 4 (t) = −a −2 (t), which, by TMT with effective EYMH anisotropic and inhomogeneous nonlinear interactions, result in target d-metrics of the type (90). Using polarization functions we write with • w i (t) = 0 and • n i (t) = 0. Such cosmological solutions are constructed as nonholonomic deformations of a conformal transformation (with multiplication on factor a −2 (t)) of the FLRW metric (82). We work with polarization functions η 4 (x k , t) when The solutions are written in the form where The class of solutions (89) represent the off-diagonal deformations of the FLRW metrics in TMTs encoding effective gauge and scalar field interactions when the effective cosmological constant is fixed to be zero. We generate solutions with non-Killing symmetry for nontrivial v-conformal factors s ω(x, z, y, t) subject to the constraints Off-diagonal cosmological solutions of type 3 and effective matter fields interactions: Non-vacuum metrics with ∂ t h 3 = 0 and ∂ t h 4 = 0 are generated by taking the ansatz where is a solution of (33) for any given v Υ(x k ) − 4s 2 . The function h 3 (x k , t) is constrained to satisfy the equation (34) which for ∂ t h 4 = 0 leads to where the constant s 2 is introduced as an additional source in order to take into account possible contributions resulting from (anti) self-dual fields. The N-connection coefficients are The Levi-Civita configurations for solutions (93) are selected by the conditions (37) which, for this case, are satisfied if 2 n k x i = 0 and ∂ i 1 n k = ∂ k 1 n i , Such conditions are similar to (91) but for a different relation of v-coefficients of d-metrics to another type of generating function Ψ. They are always satisfied for cosmological solutions with Ψ = Ψ(t) or if Ψ = const (in the last case w i (x k , t) can be any functions as follows from (35) with zero β and α i , see (38)).
where (94) generates a d-metric or (95) which depends on parameter (Λ 0 − 4s 2 ) = 0. The singular case with Λ 0 = 4s 2 can be described by a d-metric (93) when h 3 is a solution of For such configurations, we lose information about Λ 0 and s 2 but certain encodings of matter field interactions are possible in the function ψ(x k ) if the right side source ∂ 2 xx ψ + ∂ 2 zz ψ = 2(Υ − 4s 2 ) is changed to the nontrivial case 2( v Υ − 4s 2 ) for an N-adapted and anisotropic source v Υ(x k ).
Finally, we emphasize that off-diagonal deformations of FLRW metrics in TMTs with effective EYMH interactions sources of the type Υ − 4s 2 can be used for driving to zero an effective cosmological constant or for modeling parametric transforms to configurations with small effective vacuum energy.

Nonhomogeneous EYMH effects in Bianchi cosmology in TMTs
Spatially homogeneous but anisotropic relativistic cosmological models were constructed following the Bianchi classification corresponding to symmetry properties of their spatial hypersurfaces [38,39,31]. Such cosmological metrics are parameterized by orthonormal tetrad (vierbein) bases e α ′′ = e α α ′′ ∂/∂u α , if and B e α ′′ , B e β ′′ = B w γ ′′ α ′′ β ′′ (t) B e γ ′′ , are satisfied and the 'structure constants' depend on time like variables, The values B w γ ′′ α ′′ β ′′ (t) are determined by some diagonal tensor, n τ ′′ γ ′′ , and vector, b α ′′ , fields used for the mentioned classification. Depending on parametrization of such tensor and vector objects, one constructs the so-called Bianchi universes which are either open or closed similar to the homogeneous and isotropic FLRW case. With nontrivial limits from observational cosmology, there exist the so -called Bianchi I, V, V II 0 , V II h and IX universes and their corresponding cosmologies.
The AFDM allows us to generalize any Bianchi metric B g α ′′ β ′′ (96) into locally anisotropic solutions. As the first step, we transform a set of coefficients B g αβ (t) into the prime metric using frame transformations, Bg αβ = B e α α ′′ B e β β ′′ B g αβ . One also needs to solve certain quadratic algebraic equations for B e α α ′′ in order to define frame coefficients depending on the coordinate t, and Bg αβ is parameterized as a prime metric We generalize these anisotropic homogeneous cosmological metrics to generic off-diagonal locally anisotropic and inhomogeneous configurations defining cosmological solutions in TMTs with effective EYMH interactions.
The target ansatz is considered to be of the type (21), , with prime data determined by coefficients of (98). We construct metrics g defining generic off-diagonal solutions of the nonholonomic EYMH system in TMTs, (68)-(72) with source (67), following the same procedure as in section 3. In terms of polarization functions, such solutions take the following form The off-diagonal deformations of Bianchi metrics determined by sources Υ − 4s 2 = 0, and Λ 0 − 4s 2 = 0, with ∂ t h a = 0, ∂ t ̟ = 0 are computed as are computed for an effective cosmological constant Λ 0 − 4s 2 = 0 with generating function We put the left label "B" in our formulae in order to emphasize that certain values contain information on prime metrics. For simplicity, we omit "s" even when gauge and Higgs fields contributions are there.
can be considered as a re-defined source, Υ − 4s 2 → B Ξ, for a prescribed generating function B Ψ for locally anisotropic and inhomogeneous Bianchi configurations, when . This allows to compute the N-connection coefficients which is constrained additionally to define LC-configurations following the procedure described in section 3.
The v-conformal factor B ω(x k , y 3 , t) is a solution of (44) with coefficients (100) when Having constructed an inhomogeneous locally anisotropic cosmological metric g(x k , t) (99), we consider additional assumptions on generating and integration functions when the coefficients are homogeneous but with nonholonomically deformed Bianchi symmetries. This is possible if we chose at the end "pure" time dependencies B Ψ(t), Υ(t), B h a (t), w i (t) and constant values B g k and B n i .

Kasner type metrics
Another class of anisotropic cosmological metrics is determined by the Kasner solution and various generalizations [40,41,42]. Such 4-d metrics are written in the form The constants p 1 , p 2 , p 3 define solutions of the vacuum Einstein equations if the following conditions are satisfied Following the anholonomic deformation method, we generalize such solutions to generic off-diagonal cosmological configurations as in section 4.2 when Υ = 4s 2 .
The data for a primary metric are taken asg 1 = 1,g 2 = t 2(p 3 −p 1 ) ,h 3 = t 2(p 2 −p 1 ) ,h 4 = −t −2p 1 andN a i = 0 with constants p 1 , p 2 and p 3 considered for (101). For simplicity, let us analyze solutions with p 3 = p 1 and consider an example when a Kasner universe is generalized to locally anisotropic configurations characterized with gravitational polarizations For h a = η a • h a and N a i = η a i + • N a i , the target metric is of type (81) generated for Υ = 4s 2 , where w i = w i (x i , t) are arbitrary functions and The coefficient h 4 is determined by h 3 following formula |h 4 | = 0 h ∂ t |h 3 | which holds true for η a for arbitrary generating function f x i , t if p 2 = p 1 . Additional constraints on f x i , t are needed if the last condition is not satisfied. In the limit of trivial polarizations, this d-metric results into a conformally transformed metric (with factor t 2p 1 ) of the Kasner solution (101). In general, such primed metrics are not a solution of the Einstein equations for the Levi-Civita connection but it is possible to chose gravitational polarizations that generate vacuum off-diagonal Einstein fields even when the conditions of type (102) are not satisfied.
To generate homogeneous but anisotropic solutions we eliminate dependencies on space coordinates and consider arbitrary w i = w i (t) and constant 1 n k and 2 n k , when For LC-configurations, we take 2 n k = 0 and impose constraints of type (80) on w i (t).
In a similar manner, we construct various nonholonomic deformations of the Kasner universes of types 1-3 and/or and generalize them to solutions of type (103).

Effective TMT Large Field Inflation with c α-Attractors
We consider a broad class of (off-) diagonal attractor solutions that arise naturally in (modified) gravity theories and TMTs and define what we imply by natural inflationary models. In this work, we study cosmological attractors as they are considered for cosmological models in Refs. [9,11,12]. The use of the word 'attractor' need to be clarified as a similar term is widely used in the theory of dynamical systems, for certain equilibrium configurations with critical points in the phase space, i. e., critical points which are stable. Our use of the word attractor solutions is in the same spirit as the authors of references [9,11,12]. What the authors of those works mean by cosmological attractors (see, for instance, Ref. [9]) can be stated in their own words: "Several large classes of theories have been found, all of which have the same observational predictions in the leading order in 1/N. We called these theories "cosmological attractors." In our approach, the use of the word "attractor" is similar but in a more general context for generic off-diagonal solutions. Certain configurations in our work are determined by solutions, in general, with nonholonomically induced torsion and can be restricted to LC-configurations. Under such assumptions, these configurations appear again in other models under consideration by us. We group all such models as having "cosmological attractor configurations" since the configurations are common to these class of models. It is implicit that such solutions satisfy the conditions for "standard" cosmological attractor configurations (in A. Linde and co-author sense) only for certain sub-classes of nonholonomic constraints when the models are determined by imposing constraints on the corresponding generating and integration functions and integration constants. For general nonholonomic constraints, such configurations do not define cosmological attractor configurations in the sense of the above mentioned original works [9,11,12] but positively can be considered to possess similar properties for small off-diagonal deformations (perturbations) of the metrics. The important point is that such models have the same observational predictions in the leading order of 1/N . In this section, we shall define and study cosmological attractor configurations for modified gravity theories in terms of a parameter c α that determines the curvature and cutoff. "Henceforth in the following, in order to make our manuscript more transparent, wherever we use the word "attractor" in the text , it will simply imply that certain class of theories and respective off-diagonal cosmological solutions are generated which, under specific conditions on the parameter space lead to the same observational predictions.

Nonholonomic conformal transforms and cosmological attractors
Attractor type configurations are possible to construct for a certain classes of nonlinear scalar potentials in (28). We use the term "configuration" because in that formula and in formula (29) there are considered Nelongated derivatives. The equations are written with respect to nonholonomic bases and for generalized Ricci scalar curvature. As such additional assumptions are necessary in order to extract a "standard " cosmological attractor considered, for instance, in [9]. To begin with, we take the effective potential (24), for an arbitrary function q and study the model with the lagrangian The equations (29) impose the condition 1 q L = M = const 9 . Attractor models are usually constructed in terms of two fields. In addition to φ(u µ ) we consider a second field χ(u µ ). The fields (φ, χ) are subject to additional nonholonomic constraints involving the generating function Ψ = e ̟ (39), some possible re-definitions (49) of effective matter field sources Υ and the effective cosmological constant Λ 0 .
The theory (105) is related to a class of models The Lagrange density 1 χ L posses a SO(1, 1) symmetry which is deformed by the term q 2 (φ/χ). In turn, the Lagrange density 1 q L may restore the SO(1, 1) symmetry at a critical point because for large φ there exist asymptotic limits, tanh φ → ±1 and q 2 (tanh φ) → const. The terms proportional to q 2 can be transformed into effective sources and cosmological constant via eventual re-scaling of generating functions. Employing selfduality for gauge field configurations with source Υ − 4s 2 , (67), and using the gauge (107) withφ = sinh φ and χ = cosh χ, we can approximate 1 χ L to Another important property of the Lagrange density 1 χ L is that for a fixed value q = q 0 there is local conformal invariance under N-adapted transforms Such a theory describes anti-gravity if φ 2 − χ 2 > 0, i.e. χ represents the cutoff for possible values of the scalar field φ. By identifying σ from (108) with σ in (26) The explicit construction depends on the type of generating functions, conformal transforms, effective sources, asymptotic limits and gauge conditions employed in our theory. This way we construct different toy TMT models with EYMHs which for data (φ, χ) possess attractor properties and the parameters defining such attractors encode off-diagonal gravitational and (effective) matter field interactions. It is a very difficult technical task to construct cosmological solutions in such theories. Nevertheless, transforming any variant (109) into an effective gravitational theory (25) with source e T αβ → q T αβ (18) corresponding to q V contributions, (104), the effective EYMH equations (56)-(58) can be integrated in very general forms following the AFDM. Their solutions depend on integration functions and integration constants. It is very surprising that the formulas (106)-(109) and their physical consequences are similar to those for holonomic models considered in Refs. [9,11,12]. Our solutions encoding cosmological attractor configurations were derived for a class of modified theories with generalized off-diagonal metrics and nonlinear and distinguished linear connections and contributions from EYMHs for different TMT modes. It is not obvious, that such nonlinear systems may have a similar cosmological attractor behavior like in the original works with diagonal solutions. Generic off-diagonal models can be elaborated following our geometric techniques with N-adapted nonholonomic variables and splitting of corresponding systems of nonlinear PDEs. In such variables, it is possible to generate new classes of inhomogeneous and anisotropic solutions. Our goal was to find such classes of nonholonomic constraints and subclasses of generating and integration functions, and constants, when solutions with "hat" values and TMT-EYMH contributions really preserve the main physical properties of cosmological attractors. This emphasizes the general importance of the results on cosmological attractors in the cited works due to A. Linde and co-authors. Our main conclusion is that for a corresponding class of nonholonomic constraints, a cosmological attractor configuration may "survive" for very general off-diagonal and matter source deformations, in various classes of TMT theories and effective Einstein like ones encoding modified gravity theories.

Effective interactions and cosmological attractors
We can fix different gauge conditions but obtain the same results. For instance, we can work with χ(x) = 1 instead of (107), and the scalar fieldφ. With respect to N-adapted Jordan frames, the total Lagrangian is We change the d-metric into a conformally equivalent metric with equivalent Einstein frame formulation in terms of E g µν , when E g µν = (1 −φ 2 ) J g µν . The Lagrangian 1 Equivalently, 1 E L transforms into 1 q L (105) if the scalar fields are re-defined as follows, In the theory 1 E L (110) there is an ultra violet (UV) cutoff Λ = 1, i.e. Λ = M p in terms of the Planck mass and if φ become greater than 1 we get a TMT theory with antigravity. Such models were studied in the literature [37,43] and other papers before the concept of cosmological attractors was introduced. Our main goal is to study how "nonholonomically deformed" cosmological attractors can be modelled by nonholonomic constraints, generating functions and effective sources in such a way that there are satisfied the criteria for "standard" cosmological attractors to emerge in the sense of Refs. [9,11,12]. We do not consider in this work similar constructions with nonholonomic variables but only emphasize that certain antigravity effects can be modelled by off-diagonal gravitational interactions and effective polarization of physical constants [23]. The previous formulae show that φ becomes infinitely large ifφ → 1 but the effects of the cutoff can be ignored if |φ| ≪ 1, whenφ ≈ φ. Excluding some very singular behaviour near the boundary of the moduli space, the asymptotic behaviour of q V (φ) (104) at large φ is universal. This universality exists for TMT models with effective EYMH interactions as follows from above equivalence (under well-defined conditions) of theories (109) and (25).
The goal of this section is to study cosmological effects of (in general, locally anisotropic and inhomogeneous) attractors parameterized by a constant c α O(1). Attractor configurations can be introduced in several inequivalent ways. We will generalize the constructions following [9] and analyze possible off-diagonal solutions determined by sources q V (φ) and c α-parameters.
We consider the Lagrangian which is given also in the Einstein frame as (110) but contains a cutoff c α. We label the scalar field asφ (instead ofφ) in order to emphasize that we shall analyze a special class of solutions with c α-dependence. We obtain a c α-attractor configuration by re-scaling the scalar field, which leads to effective theories of the type with a shifted cutoff position at Λ = √ c α.

Off-diagonal attractor type cosmological solutions
As alluded to in the previous subsection, the asymptotic behaviour of q V (φ) (104) at large φ is universal. This universality allows to construct various classes of generic off-diagonal cosmological metrics in modified models of gravity with effective EYMH interactions using the conformal factor transformation (26). This is possible even when the generating functions and sources are very different for different classes of effective matter field interactions with nonlinear scalar potentials. The goal of this section is to prove how q-terms of the type q V (φ, c α) for attractors are encoded in various classes of solutions studied in previous section. This holds for any where the left label "c" indicates that certain values refer to attractor configurations with c α-scale. The physical cosmological d-metric c g αβ is computed to be Having computed c g µν for the data [ c g µν , q V, M, U ] , we construct a corresponding TMT model when the second measure is taken to be Formulae (111) and (112) can be applied to generate solutions for the TMT system (29) -(31) if c g µν is known as an attractor cosmological metric (in general, nonhomogeneous and locally anisotropic) for effective EYMH interactions.

Off-diagonal effective EYMH cosmological attractor solutions of type 1
Using (89) and (111), we construct families of generic off-diagonal cosmological attractor configurations with metrics where the gravitational polarizations and N-connection coefficients are computed to be The parameter c α contributes to all data defining such nonholonomic deformations of FLRW primary metric because it is included in the effective source when Υ → c Υ with c Υ − 4s 2 = 0. The corresponding effective cosmological constant is labeled Λ c 0 and satisfies the condition Λ c 0 − 4s 2 = 0 (for the class of solutions of type 1). As a result, the generating functions is redefined to simplify the formulas, c Ψ ←→ c Ψ, with The information on q V (φ, c α) is also contained in the functional It is considered as a re-defined effective source, c Υ − 4s 2 → c Ξ, for a prescribed generating function c Ψ, for We express (113) as a d-metric (21) with coefficients relevant to the v-metric:  For the off-diagonal attractor N-connection coefficients, we compute The "vertical" conformal factor c ω(x, z, y, t) is a solution of (44) for which attractor data is written in the form The function c ψ(x, z) presented in the attractor's polarization functions is a solution of (68) when ∂ 2 xx c ψ + ∂ 2 zz c ψ = 2( c Υ − 4s 2 ). Finally, we conclude that the formulae for the coefficients of the d-metric (113) depend on the type of N-adapted frame and coordinate transforms necessary to fix observational data. The conformal factor e 2 c σ(u) encodes attractor parameters in a more direct form.

Generalized locally anisotropic Bianchi attractors
Sources with attractor potential q V (φ) (104) induce generic off-diagonal cosmological solutions, in general, with inhomogeneity and local anisotropy. For a target ansatz of type (21), we parameterize when the prime solution Bg is determined by coefficients of (98). Our purpose is to state the conditions when c g from above formula defines generic off-diagonal solutions with attractor properties in TMTs with effective EYMH interactions, i.e. of (68)-(72) with source (67) encoding an attractor potential. 10 We follow the same procedure as in sections 3 and 4.3.2 and write in terms of polarization functions, We use double left labeling with "B" and "c" in order to emphasize possible Bianchi anisotropic and attractor like behaviour of certain geometric/ physical objects. The off-diagonal deformations with ∂ t c h a = 0, ∂ t c ̟ = 0 are determined by The generating functions encode data on inhomogeneous locally anisotropic interactions, attractor configurations and EYMH sourses, which results in a re-defined source , for a prescribed generating function B c Ψ. The N-connection coefficients in (114) are computed . 10 It is supposed that the parameter c α contributes to all data defining nonholonomic deformations of a primary Biachi metric. This parameter is included into effective source when Υ → c Υ with c Υ − 4s 2 = 0 and the effective cosmological constant Λ c 0 is chosen to satisfy the condition Λ c 0 − 4s 2 = 0.
Following the procedure explained in section 3, we impose additional constraints and extract LC-configurations. Dependency on all spacetime coordinates are modelled via a v-conformal factor B c ω(x k , y 3 , t) (in indirect form, it also contain attractor properties) as a solution of (44) with attractor coefficients stated above when Restricting the class of generating functions, we extract homogeneous configurations but with anisotropies when parameterizations are of the type B c Ψ(t), c Υ(t), B c h a (t), B c w i (t) and constant values for B c g k and B c n i . Applying the AFDM, we generate off-diagonal cosmological attractor solutions of types 2 and 3 for the conventional and other families of inflation potentials, for instance, when we use q( φ/ [9]. We note that we have used a different system of notations and our approach is based on geometric methods which allows us to construct exact solutions of modified gravitational and matter field equations. For certain well defined conditions, we reproduce the results and "diagonal" models studied in (supersymmetric) models with dark matter and dark energy effects. Nevertheless, nonlinear parametric systems of PDEs corresponding to effective EYMH interactions in (modified) TMTs contain solutions at a richer level that were not analyzed and applied to modern cosmology. Even though the off-diagonal effects at large observational scales seem to be very small, the generic nonlinear character of cosmological solutions depending on space like coordinates result in new nonlinear physics described by re-scaling via generating functions and effective sources. Attractor type configurations offer alternative solutions of crucial importance for explaining the inflation scenarios in modern cosmology.

Cosmological implications of TMT nonholonomic attractor type configurations
Here we concentrate on observational consequences of generic off-diagonal solutions for the effective EYMH systems with attractor properties in TMTs. We have demonstrated that Lagrangians of type 1 q L (105) and 1 χ L (106) and their effective energy-momentum tensors are naturally included as sources (22) in action (27) with two measures, which result in a nonholonomic modification of Einstein gravity (25). Geometrically, we reproduce such effects via re-definition of generating functions (26) and fixing a cut off constant c α for attractor configurations, when the effective matter field interactions are modelled for a nonholonomic off-diagonal vacuum configuration with small effective cosmological constant and gravitational η-polarizations.
In general, proposing and observing physical realizations for solutions with arbitrary η-deformations of well known prime cosmological metrics (for instance, of FLRW, Bianchi or Kasner type ones) are difficult. Nevertheless, we have elaborated upon the large distance inflationary scenarios if η ≈ 1 + ε η when |ε η| ≪ 1. We note that such configurations encode nonlinear parametric effects even when the off-diagonal and inhomogeneous terms are not taken into consideration in order to explain certain observational data. Using the results of analysis for 1 q L (105) and LC-configurations [9], we conclude and speculate on such observational consequences: 1. TMTs and nonholonomic modifications of the EYMH theory contain inflationary model of the plateau-type and features of universal attractor property when n s = 1 − 2/N and r = 12 c α/N 2 .
2. For c α = 1 such models are related to cosmological scenarios with the Starobinsky type model and Higgs inflation [45,46,47,48]; we obtain an asymptotic theory for quadratic inflation with n s = 1−2/N, r = 8/N, for large cutoff c α.
3. Decreasing c α, we get a universal attractor property both for diagonal and off-diagonal configurations; there are many models which have the same values n s and r. This property is preserved for EYMH contributions, solitonic and/or gravitational waves for corresponding nonholonomic configurations. 4. In the limit of large c α, we have generated models of simplest chaotic inflation. We have shown that effective nonlinear potentials with a second attractor are other viable possibilities.

5.
For intermediate values of c α, the predictions interpolate between these two critical points, thus oscillating between the sweet spots of both Planck and BICEP2 [5,6,7].
With respect to the old and new cosmological problem, the issues 1-5 is analyzed in the context of TMTs when the constructions are naturally extended to include effective gauge field contributions which, in turn, modify nonlinearly the sources, effective cosmological constant and generating functions. Via conformal transforms, the attractor configurations are related to inhomogeneous and locally anisotropic solutions in modified gravity theories. It is not surprising that the cosmological attractor configurations with TMT and nonhlonomic modifications of the EYMH theory are described in diagonal limits by the same parametric data as for the holonomic attractor solutions [9,11,12]. We imposed such nonholonomic constraints and selected respective generating functions which reproduce this class of cosmological solutions. Nevertheless, the constants c α, n s , r encode contributions from modified gravity theories and off-diagonal gravitational and matter field interactions and result in different observational consequences.

Concluding Remarks
To mention a few, the most important physical solutions in modern gravity and cosmology theories pertaining to black holes, wormhole configurations, FLRW metrics, are constructed for diagonal metrics transforming the (modified) Einstein equations into certain nonlinear systems of second (or higher) order ODEs. The solutions generally depend on integration constants. Such constants are fixed following certain symmetry and other physical assumptions in order to explain and describe the experimental and observational data. There are also constructed more sophisticated classes of solutions, for instance, with off-diagonal rotating metrics with Killing, Lie type symmetries and solitonic hierarchies which provide important examples of nonlinear models of gravitational and matter interactions. Nevertheless, the bulk of such analytic and numerical methods of constructing exact solutions are based on certain assumptions where the corresponding nonlinear system of PDEs are transformed into ODEs. The solutions are parameterized via integration parameters, symmetry and physical constants. The main idea is to formulate an approach to simplify the equations and find solutions depending, for instance, on a radial or a time like variable. The drawback of this approach is that a number of nonlinear parametrical solutions are lost and thus unavailable for possible applications in cosmology and astrophysics.
The AFDM is presented as a geometric method for constructing general classes of off-diagonal metrics, auxiliary connections and adapted frames of reference when gravitational and matter field equations in various modified/ generalized gravity theories, including general relativity, are decoupled. This decoupling implies that the corresponding nonlinear system of PDEs splits into certain subclasses of equations which contain partial derivative depending only on one coordinate and relates only two unknown variables and/or generating functions. As a result, we can integrate such systems in very general off-diagonal forms when various classes of solutions are determined not only by integration constants but also by generating and integration functions, symmetry parameters and anholonomy relations. The solutions depend, in general, on all spacetime coordinates and can be with Killing or non-Killing symmetries, of different smooth classes, with singularities and nontrivial topology. We can make, for instance, certain approximations on the type of generating functions and effective source at the end, after a general form of solution has been constructed. This way we generate new classes of cosmological metrics which are homogeneous or inhomogeneous, and in general, with local anisotropies, which can not be found if one works from the very beginning with simplified ansatz and higher symmetries. Furthermore, the possibility to re-define the generating functions and sources via nonlinear frame transformations and parametric deformations allows one to entertain new classes of solutions and study various nonlinear physical effects.
In this paper, we studied in explicit form certain classes of modified gravity theories which can be modeled as TMTs with effective EYMH interactions. Possible scalar fields and corresponding nonlinear interaction potentials were chosen to select and reproduce attractor type solutions with cut off constants which seem to have fundamental implication in elaborating isotropic and anisotropic inflation scenarios in modern cosmology. In general, one can work with off-diagonal configurations and consider diagonal limits for minimal and/or nonminimal coupling constants. We proved that the decoupling property holds also in TMTs which results in the possibility of constructing various classes of off-diagonal cosmological solutions with small vacuum density. Such solutions describe spacetimes with nonholonomically induced torsion. Nevertheless we formulated well-defined criteria when additional nonholonomic constraints are introduced that allow to extract LC-configurations. We studied nonholonomic deformations of FLRW, Bianchi and Kasner type metrics encoding TMT effects and possible contributions of effective EYMH interactions.
We have shown that attractor type cosmological solutions with cut-off parameters can be derived by nonlinear re-definitions of generating functions and effective sources in TMT if a corresponding type of nonlinear scalar potential is chosen. In general, such attractor solutions are model independent and are constructed in explicit form to accommodate effective EYMH interactions. In this way various large scale inflationary models, with anisotropic expansion and parametric nonlinear processes can be realized.
For certain conditions, the gravitational and matter field equations of TMTs are expressed as effective Einstein equations with non-minimal coupling [18]. In this presentation, we proved that in nonholonomic Nadapted variables and for additional assumptions the constructions are generalized in such form that two measure configurations serve to encode massive gravity effects and nonlinear parametric off-diagonal interactions (see formulas (25)-(27)). In general, such a theory also has 4 extra degrees of freedom with the Boulware-Deser (BD) ghosts. This problem can be circumvented if one imposes additional constraints. We imposed nonholonomic constraints for constructing cosmological attractor configurations. This procedure constrains the extra dimension degrees of freedom and encodes the TMT and massive term contributions into certain subclasses of solutions for off-diagonal effective Einstein spaces (see similar constructions for ghost-free massive f (R) theories in Refs. [49]). We conclude that in our models the BD ghosts are absent for such special classes of nonholonomic configurations if generic off-diagonal cosmological solutions are constructed for effective Einstein equations of type (33)- (37).
There remain many open questions on how to provide viable explanations for the recent observational data from Plank and BICEP. In this work, we have shown that attractor configurations can be constructed in TMTs with effective gravitational and matter field equations. Such solutions provide a new background for investigating cosmological theories with anisotropies, inhomogeneities, dark energy and dark matter physics.