Papers

This paper investigates the geometric and cohomological properties of non-Kähler SYZ mirror symmetry for dual torus fibrations over solvmanifolds in the sense of Lau, Tseng and Yau. We are mainly concerned with three questions: (a) How the Lau-Tseng-Yau notion of non-Kähler SYZ is related to the mapping of supersymmetric branes between symplectic and complex sides; (b) Finding explicit non-Kähler SYZ mirror pairs determined purely by Lie-theoretic data; (c) better understand the cohomological correspondence in the Lau-Tseng-Yau framework (given by a Fourier-Mukai transform), especially concerning the role of Tseng-Yau cohomology. We prove that the Fourier-Mukai transform introduced by Lau-Tseng-Yau exchanges type-A supersymmetric cycles, which are given by special Lagrangian sections equipped with flat $\mathrm{U}(1)$ connections, with type-B cycles, corresponding to line bundles whose connections satisfy the deformed Hermitian-Yang-Mills (dHYM) equation. We provide pure Lie-theoretic criteria for the existence of non-Kähler SYZ mirror pairs whose base manifolds are solvmanifolds. Applying these criteria, we construct new explicit families of mirror pairs from almost abelian and generalized Heisenberg Lie groups, and provide a complete classification of such pairs arising from nilpotent Lie groups. To contextualize the role of the Tseng-Yau cohomology, we link it to noncommutative geometry. We introduce the Tseng-Yau and Bott-Chern mirror bicomplexes. We show that (some of) their enclosed cohomologies reduce to the primitive Tseng-Yau and Bott-Chern cohomologies and that for basic forms they are isomorphic under the Fourier-Mukai transform. As a last contribution, we discuss how to explicitly compute the Tseng-Yau and the Bott-Chern cohomology for the non-Kähler SYZ mirror pairs constructed here.

We derive an exact solitary wave solution for the $\PTb$-symmetric nonlinear Dirac equation with a scalar-scalar interaction. We consider a power-law nonlinearity of the form $|\bar{\Psi}\,\Psi|^{k}\,\Psi$ for positive values of $k$. The system's energy is conserved despite the presence of a gain-loss term, which is quantified by the parameter $\Lambda$. We show that the $\PTb$-transition point is defined by the solution's existence condition and is independent of the nonlinearity exponent $k$. Furthermore, momentum is conserved, although neither the canonical momentum nor the charge is a conserved quantity. A notable result is that the stationary solution, obtained from the continuity equations, exhibits nonzero momentum in its rest frame. We also derive a moving soliton solution, where the gain-loss parameter allows the soliton's velocity to be precisely chosen so that the moving soliton possesses zero momentum. Finally, we establish that the presence of a gain-loss mechanism and higher-order nonlinearity restrict the stability domain of the solutions.

Fibre inflation is one of the most attractive models realized in the type IIB orientifold compactification. It is embedded in the framework of L(arge) V(olume) S(cenarios) using a class of compactifying Calabi-Yau (CY) threefolds having K3-fibration. The standard single-field fibre inflation is driven by a fibre modulus which needs to travel a trans-Planckian distance of the order of ${\cal O}(5-8)$M$_p$ in the effective moduli space. The global embedding attempts using concrete CY orientifold setups have shown that Kähler cone conditions can generically induce some significantly tight bounds on the inflaton range, especially in the presence of a Swiss-Cheese structure via an exceptional rigid divisor in the CY threefold. Such field range bounds usually obstruct the inflationary plateau, leading to insufficient number of efolds during the inflationary dynamics. In this context, we review our recent work about the possibility of assisting multiple fibre moduli such that the burden of traveling the required trans-Planckian distance could be shared by multiple fields, and successful inflation could be realized before hitting (or being too close to) their respective individual Kähler cone boundaries.

We review the state-of-the-art knowledge of IR singularities in multileg QCD amplitudes, identifying the key reasons for the remarkable simplicity of the soft anomalous dimension. We then present a novel strategy to compute this quantity using a lightcone expansion of correlators of semi-infinite Wilson lines by the Method of Regions. Recently, this strategy allowed us to determine the three-loop soft anomalous dimension for amplitudes consisting of a single massive coloured particle with any number of massless ones. It opens the way to computing this quantity for amplitudes involving two heavy particles at three loops and potentially going to higher loop orders.

We analyze a class of product geometries of the form $\mathbb{R}^{1,1}\times \Sigma_2$ supported by electric, magnetic, or dyonic flux in the Einstein-Maxwell-$\Lambda$ theory. These spacetimes belong to a unified family of direct products $(dS_2,\mathbb{R}^{1,1},AdS_2)\times \Sigma_2$ distinguished solely by the sign of the Lorentzian curvature of the two-dimensional factor. We focus on the critical configuration for which the Lorentzian curvature vanishes. At this balance point between the cosmological curvature and the Maxwell stress, the longitudinal geometry becomes exactly flat while the transverse sphere radius is fixed algebraically by the conserved flux. We refer to this geometry as the critical Maxwell flux string: a homogeneous flux-supported geometry curved only in the transverse directions and invariant along a two-dimensional null worldvolume. It represents the algebraic midpoint interpolating between the Nariai $(dS_2\times S^2)$ and Bertotti-Robinson $(AdS_2\times S^2)$ spacetimes. A qualitative structural change occurs precisely at this midpoint. The spacetime is Petrov type-D with constant scalar curvature invariants, placing it in the degenerate Kundt/CSI class. Because the curvature structure reduces any polynomial rank-two tensor to a linear combination of the metric and the Maxwell stress tensor, the same configuration solves a broad class of algebraic higher-curvature gravity theories. In this sense, the critical flux string and its aligned deformations constitute almost universal solutions.

Making use of the known off-shell formulations for massless higher-spin ${\cal N}=1$ supermultiplets in four dimensions, gauge-invariant off-shell actions for massive higher-spin ${\cal N}=2$ supermultiplets in three dimensions (3D) are derived by Kaluza-Klein reduction in superspace. To illustrate the formalism, we also construct, for the first time, massive gauge-invariant 3D ${\cal N}=2$ supersymmetric counterparts of the linearised actions for the old and new minimal supergravity theories. Our off-shell ${\cal N}=2$ supermultiplets carry a non-zero central charge, and are formulated in 3D ${\cal N}=2$ central charge superspace. The models can be reduced to 3D ${\cal N}=1$ superspace, by integrating out two Grassmann variables, and then consistent reality conditions on the superfields can be imposed. As a result, only two supercharges remain unbroken.

Motivated by the first oscillation results from JUNO, we study the phenomenological viability of texture zeros in the Dirac neutrino mass matrix. The improved precision on the solar mixing angle $\sin^2{\theta_{12}}$ and the solar mass-squared difference $\Delta m_{21}^2$ provide a stringent probe for scrutinizing predictive texture zero frameworks. We perform a systematic scan of the allowed parameter space for two-zero textures, identifying sharp correlations among oscillation observables arising from the reduced parameter space. Our analysis reveals that current JUNO measurements impose stringent constraints on the viable texture structures. In particular, although textures $C$, $A_2$, and $A_1$ were previously viable, current JUNO data strongly disfavor $C$, leaving only textures $A_2$ and $A_1$ compatible with the data. These findings underscore the remarkable sensitivity of Dirac texture zero scenarios to the solar sector.

We develop a systematic Hamiltonian formulation for a gravitating topological matter system in three-dimensional spacetime, coupling a scalar matter field and a two-form gauge field to first-order Hilbert–Palatini gravity. We perform the Dirac–Bergmann analysis, obtaining the full structure of the constraints, classifying them into first- and second-class ones, and computing their Poisson bracket algebra. Furthermore, we write down the explicit expression for the Hamiltonian generator of gauge symmetries on the full set of canonical variables, containing the exact number of gauge parameters. A mapping of these parameters then shows that the gauge transformations reproduce on-shell the spacetime diffeomorphism and local Poincaré symmetries. The symmetry structure of the coupled model is thereby established on a Cauchy surface without boundary. Our canonical analysis further shows that the first-class constraints obey three reducibility conditions. We trace them back to an off-shell identity of Ricci type and to the redundancy of the gauge parameters carried by the two-form field. In the extended phase-space they close only modulo the second-class sector. Once they are taken into account, the model carries no local degrees of freedom. The fundamental symplectic structure on the reduced phase-space is fixed by the explicit computation of the Dirac brackets. Both these and the Hamiltonian gauge generator are obtained in the extended phase-space, with no canonical variable eliminated beforehand. Finally, we show that on a surface with boundary the improved generator built from the reducible parameters collapses to a single boundary integral, which vanishes identically.

We investigate symmetry-resolved entanglement in non-relativistic quantum field theories, including complex Lifshitz scalar chains and Lifshitz fermionic models. Using charged moments and the correlator method, we compute symmetry-resolved Renyi and von Neumann entropies and analyze their dependence on subsystem size, charge, mass, and the dynamical exponent z. Our results reveal distinct features of non-relativistic entanglement. In Lifshitz scalar theories, approximate equipartition among charge sectors emerges in the large-z regime, with configurational entropy dominating, whereas Lifshitz fermionic models exhibit genuine equipartition only in the relativistic limit, with fluctuation entropy prevailing. These findings highlight a rich interplay between conserved charges, subsystem size, mass, and dynamical scaling, and provide a framework to explore operationally accessible entanglement in non-relativistic systems. Our study offers insights relevant to experimental platforms such as cold atom setups and mesoscopic systems, where particle-number-resolved measurements can probe symmetry-resolved entanglement.

Despite the growing application of large language models (LLMs) to theoretical physics, there has been little academic exploration of how domain-specific physics reasoning ability develops during training. To investigate this, we perform the first academic comparison of reinforcement learning and supervised fine-tuning for LLMs on verifiable problem-solving tasks in theoretical physics. Because the open-source verifiable training data required to develop such capabilities is scarce, we built a robust data generation pipeline that can both create synthetic problems and make existing human-authored problems suitable for model training. Selecting quantum field theory (QFT) as our primary domain, we generated over 2,500 synthetic problems alongside a curated collection of human-adapted problems sourced from arXiv and standard pedagogical resources. We conduct both reinforcement learning (RL) and supervised fine-tuning (SFT) experiments, benchmarking performance gains as well as generalization to other physics domains. Although both methods improve over the base model on every dataset, SFT gains are concentrated on in-distribution synthetic tasks, while RL generalizes better, outperforming SFT on the human-adapted datasets and TPBench. We perform an extensive analysis of model chain-of-thought (CoT) reasoning before and after fine-tuning to understand how reasoning errors evolve during RL and SFT. Finally, we publicly release our data pipeline, verifiable QFT training data, and ~200M tokens of QFT reasoning traces.

We perform a systematic study of invariant operators in the three-Higgs-doublet model (3HDM). We compute the closed-form multigraded Hilbert series associated with the global symmetry group of the theory. In addition, we construct explicit expressions for the corresponding invariant operators up to cubic order in the couplings. As a phenomenological application, we further organize the invariants up to quadratic order in the field content into CP-even and CP-odd combinations. For the special case of invariants involving no $\mathbf{27}$ representations, our results are complete to all orders.

In this article we present a systematic observational verification of the ghost-free string-inspired $f(R,\mathcal{G})$ model, where the Gauss-Bonnet invariant is non-minimally coupled to an auxiliary scalar field $\chi$ through the coupling function $h(\chi)$. Previous studies confirmed the theoretical viability of this framework using phenomenological parameter choices. In this work, for the first time, a systematic comparison with observational data from Planck 2018 and the Atacama Comsology Telescope is carried out via a Bayesian MCMC analysis using the Cobaya code. We explore an extended set of sixteen models constructed from four types of the Hubble parameter combined with power-law, exponential, hybrid, and inverse logarithmic coupling functions $h(\chi)$. The hybrid coupling $h(\chi) = \gamma e^{b_1\chi}\chi^{b_2}$, introduced in this context, allows for interpolation between the power-law and exponential forms, providing additional flexibility in controlling the Gauss-Bonnet contribution at different stages of inflation. All sixteen models reproduce the red spectral tilt of scalar perturbations consistent with CMB observations, yielding $n_s \approx 0.97$ at $N = 60$ e-folds. We find that the preference for the dataset is systematically determined by the choice of Hubble parametrization rather than by the coupling function. The parameter $\mu\approx0.1$ remains stable in all configurations, suggesting its fundamental role within the ghost-free formalism.

We study the BCS–BEC crossover of the large $N$ attractive Fermi-Hubbard model on a one-dimensional lattice using the mean field approximation in the presence of an imaginary chemical potential. We show that the crossover is governed by three parameters. The imaginary chemical potential $i\theta$, the temperature via a thermal kernel $g(\beta E_k,\beta\theta)$ and the parameter $\delta_u$ whose sign controls the weak and strong coupling regimes. At the unitarity point ($U=U_c$), we find a thermal window $\phi=\beta\theta=2\pi/3,4\pi/3$ where the gap vanishes while the fermion number $N_f$, which quantifies the balance between particle-like and hole-like excitations, has a local maximum/minimum. Inside this thermal window BCS and BEC physics are await changes in the coupling to be selected as the dominant regime. We expect that our results will unveil a better understanding of pairing correlations in lattice many-body physics.

Ordinary, s-wave superconductors have been recognized as being topological phases of matter, in which the dynamical gauge field implies less understood global features. Using the tools of topological field theories and generalized symmetries, we provide an updated description of these systems. At very low energies, the Higgs model reduces to the BF theory, which exhibits topological order. Furthermore, the gauge field must be a spin$_c$ connection, to describe the spin of fermions forming Cooper pairs. Gauging implies that superconductors are inherently bosonic systems, yet they are endowed with a gravito-magnetic anomaly that is the remnant of their fermionic origin. We recognize that this anomaly is related to the Gaiotto-Kapustin-Thorngren bosonization, achieved via gauging fermion parity $(-1)^F$, now included in the gauge dynamics. This anomaly characterizes gauged electronic matter in great generality in three and four spacetime dimensions, forbidding trivial massive phases at low energy. It holds beyond the validity of the Higgs model, nd in other kinds of superconductors as well. It also appears in the nontrivial massless phase of three-dimensional electrodynamics, recently understood.

We study at zero temperature a microscopic quantum spin-1 model on the fuzzy sphere that realizes the $O(2)$ Wilson-Fisher conformal field theory (CFT) in $(2+1)$-dimensional spacetime at a quantum critical point. Here, we use the fuzzy-sphere regularization as it preserves the full spatial $SO(3)$ rotational symmetry of the CFT, enabling the state-operator correspondence that maps energy eigenstates directly to CFT operators. Using exact diagonalization (ED) and matrix product state (MPS) techniques combined with conformal perturbation theory (CPT), we extract conformal data including scaling dimensions and operator product expansion (OPE) coefficients. We identify 48 primary operators and their descendants, organized by the conserved $O(2)$ charge $S^{z}$ and spatial angular momentum $L$. Our numerical results for the scaling dimensions of the lowest primary operators show good agreement with conformal bootstrap predictions. We verify predictions from the large charge expansion, which provides systematic predictions for operators carrying large $U(1)$ charge, connecting the Goldstone mode physics in the ordered phase to phonon primaries at the critical point.

We provide several consistency checks of confining dynamics in a recently conjectured holographic dual of a four-dimensional ${\cal N}=1$ supersymmetric gauge theory that flows from a conformal manifold in the UV to a finite set of isolated, fully gapped vacua in the IR. This is obtained by considering D3-branes at the conifold singularity in the presence of an O7-plane, leading to a background where all supergravity fields have a non-trivial profile. We compute holographically the expectation value of a Wilson loop in the fundamental representation and show that it obeys an area law. We then construct the domain walls which interpolate between different vacua in terms of D5-branes wrapping a compact three-cycle of the internal manifold. Their dynamics is governed by the (2+1)-dimensional ${\cal N}=1$ Yang-Mills-Chern-Simons theory predicted by field theory arguments, that reduces in the deep infrared to a TQFT whose inflow action correctly reproduces the mixed anomaly of the four-dimensional theory. Finally, we argue that, unlike in previous models in the literature, axionic strings are unstable in this background. This implies that the corresponding massless axion that would couple to them is absent, in agreement with the fact that the vacua are fully gapped.

High temperature is usually expected to destroy order: as the Gibbs state approaches the infinite-temperature limit, it becomes an equal-weight ensemble over all states and the system is generically disordered. Recent works showed that entropic order can violate this expectation through coupling to bosons in classical lattice models and quantum field theories, where the ordered states have higher entropy. Here we present new analytic methods for constructing quantum lattice models that exhibit entropic orders. In particular, we construct quantum lattice models with continuous symmetry breaking at high temperature in 1+1 dimensions and clarify how entropic order can evade the Hohenberg-Mermin-Wagner theorems. We also construct high-temperature entropic $p+ip$ chiral topological superconducting states in 2+1 dimensions with temperature-independent anyon correlation functions. In addition, we obtain a broad family of high-temperature entropic non-chiral topological orders. We show that the entropic topological orders have strong higher form symmetries at high temperature unlike the conventional topological orders, and the symmetry is spontaneously broken. These results follow from two general constructions that couple a given lattice model with a low-temperature ordered phase either to ordered bosons or, for local commuting-projector Hamiltonians, to more general bosonic degrees of freedom.

We analyze the stability properties of a family of anti-de Sitter flux compactifications of the tachyon-free non-supersymmetric heterotic string in ten dimensions. In contrast with simpler such solutions, the solutions include two independent unbounded fluxes, leading to richer instability phenomena. In particular, when the two fluxes are sufficiently close in magnitude, the perturbative spectrum develops tachyonic modes, which can be projected out by an orbifold action. When the fluxes are far apart, tachyonic modes are absent, and the geometry displays inverse scale separation, where a factor of the internal manifold becomes parametrically larger than the anti-de Sitter factor. Still, non-perturbative instabilities in the form of brane nucleation are always available decay channels, and tend to drive the two fluxes closer together, eventually triggering the tachyonic instability when present.

We demonstrate that simple feed-forward neural networks (NNs) can accurately compute correlation functions of conformal field theories (CFTs) on a line. Strikingly, by optimising a NN solely on crossing symmetry and providing only the scaling dimension of the leading non-trivial operator and the correlator's value at a single "anchor point", we can reconstruct target physical correlators to within a few percent. We establish the robustness of this minimal-data approach across a broad class of theories and dimensions, including generalised free fields, contact and one-loop Witten diagrams in AdS$_2$, unitary and non-unitary 2d minimal models, the 3d Ising model, and half-BPS correlators in 4d $\mathcal{N}=4$ super-Yang-Mills theory, together with several thermal two-point functions, notably including those of the 3d Ising model. We argue that this remarkable alignment between NNs and CFTs stems from the spectral bias of gradient-based training, which heavily favours smooth functions. To ground this connection, we analyse the smoothness of conformal correlators using fractional Sobolev semi-norms, Chebyshev spectral decompositions, and a measure based on curvature. Finally, we establish the broader reconstructive power of this technique by extending it beyond the diagonal kinematics of the line.

On an unstable D-brane, the rolling of the tachyon away from the maximum of its potential is described by time-dependent solutions in string theory. Subsequent analysis leads to an understanding of physics around the tachyon vacuum in terms of an effective field theory. The classical solutions of this effective field theory in the late time limit are in one-to-one correspondence with configurations of non-rotating, non-interacting dust particles. In this work, we map this effective theory near the minimum of the potential to a consistent quantum description using collective field theory methods. The Hilbert space description we obtain in this way consists of a coherent state of particles at rest and excitations on top, i.e., a subsector of Klein-Gordon theory. This is in accordance with known results where one considers a decaying D-brane as providing a time-dependent source for closed strings, and the final closed string state produced is a coherent state. It also suggests, at the quantum level, an equivalence between the open string description and the closed string description regarding the decay of an unstable D-brane.

We develop a perturbative formulation of the Ricci flow in gravity. Following steps analogous to the gradient flow in QCD, we supplement the usual Feynman rules for perturbative gravity by flowed propagators and vertices as well as graviton flow lines which describe the evolution of gravity along the Ricci flow. By calculating vacuum expectation values of a number of independent operators at the two-loop level, we derive the required counterterms of the flowed action. Our results allow us to define a Ricci-flow based renormalization scheme for Newton's constant $G_N$. Studying its renormalization group behavior, we recover a non-Gaußian fixed point in accordance with well-known non-perturbative considerations

The SYK model has played an important role in recent developments in many-body quantum chaos. We study a spatially local generalisation of it: the SYK lattice. Starting from the nonlinear action of pseudo-Goldstone bosons that dominate its dynamics at low temperatures, in the long wavelength limit we reorganise this action as the effective field theory for fluctuating hydrodynamics, thereby showing how the hydrodynamic degrees of freedom embed into the microscopic description of the model. We compute the hydrodynamic effective action to high orders in the derivative expansion and determine all the corresponding transport coefficients. Hence this work derives hydrodynamics from the microscopic description of a strongly coupled quantum many-body system.

We propose that simple neural networks (NNs) trained on crossing symmetry can reconstruct conformal correlators restricted to a line to remarkable accuracy. The input is minimal: an external scaling dimension, a spectral gap, and the value of the correlator at a single point. We present evidence across a wide range of conformal theories and dimensions, for both four-point and thermal two-point functions. We attribute these observations to the spectral bias of gradient-based NN training, which appears to align with an intrinsic smoothness property of conformal field theory. This suggests a novel variational principle for conformal correlators and opens a path towards a powerful new computational framework for non-perturbative quantum field theory.

We describe the implementation and usage of `fermionic_amplitudes.m', a Mathematica package for the computation of tree amplitudes involving arbitrary numbers of gauge bosons and arbitrarily-charged massless fermions of (possibly) distinct flavours in pure (non-supersymmetric) gauge theory. These are given in terms of a basis of partial amplitudes involving distinct-flavoured fermions dressed by specific colour tensors. Distinct-flavour partial amplitudes are expressed as linear combinations of those involving only a single flavour, which may be evaluated as component amplitudes of (maximally) supersymmetric Yang-Mills theory. All relevant colour tensors can be realized as explicit, numeric arrays given any choice of charge generators (for any gauge theory – including $u_1$); from these, all colour contractions relevant to cross sections may be readily computed. The complete package and a notebook demonstrating its primary usage and functionality are included in this work's submission's ancillary files on the arXiv.

We show that the LIGO–Virgo–KAGRA verification of Hawking area law carries profound consequences for quantum gravity if such a law is postulated to hold exactly. The observed mergers can be produced in local Stelle gravity and in nonlocal quantum-gravity theories with entire or fractional form factors either by (i) singular Ricci-flat black holes or (ii) possibly regular classical black holes under very restrictive conditions: absence of $R^2$ and (Riemann)${}^2$ terms in the action, absence of extra real poles in the graviton propagator, and positivity of its spectral representation. To date, this is the strongest simplification of the ambiguities of this class of theories. We also prove that the classical standard entropy-area law holds as a consequence of Hawking area law and provide a rigorous realization of Barrow's fractal black holes otherwise.

We reevaluate the vacuum polarization functions for electroweak gauge bosons at three loops in QCD, employing state-of-the-art perturbative techniques. We apply these results to determine the ${\mathcal{O}}(\alpha \alpha_s^2)$ corrections to the electroweak radiative parameters $\Delta\rho$, $\Delta r$ and $\Delta \kappa$. We improve the accuracy of the calculation at this perturbative order, compared to the existing literature, and present some phenomenological implications of these results. We find a shift in the prediction of the $W$ boson mass, significant in view of the FCC precision targets. We improve the prediction of the $\overline{\mathrm{MS}}$ electric charge at $q^2=m_Z^2$ with the inclusion of these ${\mathcal{O}}(\alpha \alpha_s^2)$ corrections.

It is pointed out that the vertex operators of a chiral boson in 1+1 dimensions provide an explicit realization of dichotomic, bounded, Hermitian operators that saturate the Tsirelson bound of the Bell-CHSH inequality in the vacuum state.

In this review branes of string theory are described from three different perspectives: as endpoints of open string, as supergravity backgrounds with BPS properties, as dynamical objects with gauge invariant actions. Based on these descriptions various effects of brane interactions are reviewed: brane bound states, Hanany–Witten and Myers effects, supertubes. The review is based on the lecture course given at MIPT.

Bounds for the area of general closed marginally trapped surfaces (MTSs) are presented. They do not require any stability condition, and are determined by a constant that depends on a particular component of the Einstein tensor on the surface and another constant that governs the (in)stability of the MTS. When stability is imposed, the area bounds are refined. These bounds are realized in spacetimes exhibiting interesting generic properties: they possess marginally trapped tubes foliated by marginally trapped topological spheres containing a distinguished round sphere $\bar S$ with constant Gaussian curvature that saturates the area bound. This distinguished surface separates two distinct regions of the marginally trapped tubes: a dynamical horizon and a timelike membrane. The particular case where there is a positive cosmological constant leads to the well-known universal bound $4\pi/ \Lambda$ for spatially stable MTSs, and to the recently introduced `ultra-massive spacetimes'. These spacetimes are more extreme than black holes, as there is no event horizon and the entire exterior region undergoes unavoidable collapse. In this paper similar behaviour is found for non-positive $\Lambda$ if the energy-momentum content is powerful enough. The results may have implications for binary mergers and on accreting very compact objects.

We compare the correlation functions of inflationary perturbations computed either with quantum or classical dynamics. Even if they are enforced to agree at a specific time during inflation, classical and quantum correlations will differ at the end of inflation, provided that interactions are relevant. The difference between the results of the classical and quantum computations is exponentially sensitive to the number of e-folds elapsed from the time of agreement. We illustrate this finding with the tree-level bispectrum of the primordial curvature fluctuation and the one-loop power spectrum of tensor modes. We also show that classical evolution from a finite time does not imply the appearance of poles in the scalar bispectrum.