Bayesian analysis of $α$-Starobinsky model with Planck, ACT and DESI data

2026-06-08T17:47:55Z

We present a joint Bayesian analysis to impose constraints on the generalized $α$-Starobinsky inflationary model using the high-precision cosmological datasets: Planck, CMB lensing from ACT DR6, and Baryon Acoustic Oscillations (BAO) from DESI DR2. For the parameter inference, we introduce an alternative sampling approach. Rather than imposing priors on the cosmological parameters of the inflationary potential $(V_0, \, α, \, N_*)$, we place priors directly on the primordial physical observables $(A_s,\, n_s,\, r)$ through analytical slow-roll consistency relations. Our pipeline internally maps these sampled observables to the corresponding $α$-Starobinsky parameters. These values are then passed to a modified version of $\tt{CLASS}$, which solves the exact inflationary dynamics numerically. This pipeline ensures that the final reported posteriors for the observables are computed exactly, completely free from the slow-roll approximation. Applying this methodology, we explore the viability of the $α$-Starobinsky model. We show that, when the full combined dataset is considered, the pure Starobinsky model (i.e., the canonical limit $α= 1$) shows signs of breaking down, since it requires a large number of $e$-folds of inflation after horizon crossing ($N_* > 60$) due to the shift in the scalar spectral index, $n_s$. In contrast, allowing the deformation parameter $α$ as a free parameter yields a clear $1σ$ preference for $\log_{10} α> 0$, present across all datasets. Notably, we also show that the addition of ACT DR6 lensing data introduces no significant impact on these primordial constraints, confirming that our robust posteriors are primarily driven by Planck and DESI measurements.

Choosing the right MCMC sampler: a systematic benchmark of gradient-free methods

2026-06-08T17:37:09Z

We present a set of metrics and methods for testing and comparing a range of modern gradient-free Markov Chain Monte Carlo (MCMC) samplers against the commonly used Metropolis-Hastings (MH) algorithm. The goal is to quantify key performance metrics, including sampler ergodicity, robustness and overall likelihood performance. To provide a controlled and interpretable testbed, we use the Rosenbrock function and Neal's funnel as representative unimodal cases, while Gaussian random likelihood landscapes in three, five, and eight dimensions serve as multimodal test scenarios. The samplers considered include affine-invariant moves from the literature, such as the stretch and walk moves, the differential evolution move, and the snooker move. We additionally introduce two novel variations: a modified stretch move that incorporates a Principal Component Analysis (PCA) transformation, and a hybrid blend move that combines features of both differential evolution and stretch dynamics. Beyond sampler evaluation, we demonstrate reconstructing likelihood landscapes from sampled points using a quadtree algorithm. Additionally, we explore the use of optimisation algorithms to refine the best parameter set in terms of its likelihood, and find consistent improvements in log-likelihood values, with the post-sampling gain becoming more significant in higher-dimensional problems. Our comparative results of sampler testing show that the differential evolution algorithm, when tuned to a target acceptance fraction of 25%, consistently outperforms all other samplers in terms of ergodicity, robustness, and likelihood performance.

PTA-Compatible Domain Walls at LISA and Taiji: Bayesian Reconstruction and Multiband Inference

2026-06-08T16:38:36Z

Domain walls provide an excellent fit to Pulsar Timing Array (PTA) data. A distinctive feature of the associated gravitational-wave (GW) spectrum is its ultraviolet (UV) tail, which can extend into the mHz band and thereby make cross-detection with space-based interferometers such as LISA and Taiji possible. In this work, we explore the PTA-compatible parameter space of domain-wall models and study both the reconstruction prospects at LISA and Taiji and the information gain achievable through joint PTA--LISA and PTA--Taiji analyses. We find that LISA and Taiji can probe an extended region of parameter space yielding an ultraviolet tail in the milli-Hz band, even in the presence of astrophysical foregrounds. Within the PTA-supported region, however, the genuinely informative regime is more restricted and concentrated toward the high-signal edge. In the space-based-detector-only analysis, the posterior is strongly degenerate in the underlying model parameters, although one principal parameter combination can be reconstructed with high precision where the signal is sufficiently strong. When PTA-informed priors are incorporated, the additional information gain is localized to the same high-signal region: there the space-based data produce a non-negligible posterior update and strongly suppress the catastrophic LISA(Taiji)-only degeneracy, while the final posterior often retains an axis ratio comparable to that of PTA alone. Our analysis is based on a 10-dimensional Bayesian inference with two domain-wall signal parameters and eight nuisance parameters describing instrumental noise and astrophysical foregrounds. We use 49 grid injections for the LISA-only and Taiji-only analyses and 65 injections from the PTA-supported region for the joint analyses, with Clough--Tocher interpolation used to construct posterior heat maps.

Stochastic constant-roll inflation beyond the hilltop with the spectral method

2026-06-08T16:07:28Z

Stochastic inflation can be used to study large inflationary perturbations. This paper presents such a study for a quadratic hilltop potential, corresponding to constant-roll inflation. I solve the perturbation distribution using the spectral method, with detailed solutions of the eigenvalues and eigenfunctions of the Fokker-Planck operator. Contrary to previous studies of stochastic constant-roll inflation, the solution allows trajectories that cross the hilltop and get stuck near a reflecting boundary on the other side, tunneling out slowly in a way dictated by the lowest eigensolution. Despite their rarity, these trajectories turn out to dominate the mean first-passage time. For this reason, I argue the mean does not properly describe the inflationary background. Using the median instead, I compute the distribution of the coarse-grained $ΔN$ distribution and show that its well-known exponential tail first flattens out and then forms a peak near a maximal $ΔN$ value. I argue similar intricacies arise in primordial black hole models.

Dynamical Systematics for Time Delay Lenses and the Impact on the Hubble Constant

2026-06-08T15:52:14Z

While time-delay lenses can independently probe $H_0$, the estimates are degenerate with the convergence of the lens near the Einstein radius. Velocity dispersions, $σ$, can be used to break the degeneracy, with uncertainties $ΔH/H_0 \propto Δσ^2/σ^2$ ultimately limited by systematic uncertainties in the kinematic measurements - measuring $H_0$ to 2\% requires $Δσ^2/σ^2 < 2\%$. Here we explore a broad range of potential systematic uncertainties affecting eight time-delay lenses used in cosmological analyses. We find that: (1) The characterization of the PSF in both absolute scale and shape is important, with biases in $Δσ^2/σ^2$ up to $1$-$5\%$ for ground-based observations. Small miscenterings of the lens are less important. (2) The difference between the measured velocity dispersion and the mean square velocity needed for the Jeans equations is important, with up to $Δσ^2/σ^2 \sim 2$-$6\%$. (3) The choice of anisotropy models is important with maximum changes of $Δσ^2/σ^2 \sim2$-$18\%$. Biases may be minimized by using models that reproduce the $h_4$ velocity moments typical of early-type galaxies. (4) Small differences between the true stellar mass distribution and the model light profile matter ($Δσ^2/σ^2 \sim 1$-$10\%$), with radial color gradients further complicating the problem. The Jeans equations for mixed stellar populations imply that the correct profile is a population line equivalent width weighting corresponding to no broad band filter profile. Finally, the homogeneity of the early-type galaxy population means that many dynamically related parameters must be marginalized over the lens sample as a whole and not over individual lenses.

Neutrino mass limits and decaying dark matter: background evolution versus perturbations

2026-06-08T15:46:11Z

We revisit cosmological neutrino mass bounds when a fraction of dark matter is allowed to decay to massless dark radiation. By compensating the late-time increase in the matter density induced by neutrinos becoming non-relativistic, decaying dark matter (DDM) can render datasets solely sensitive to the background density effectively insensitive to neutrino masses. Using data from baryonic acoustic oscillations (BAO) and Type Ia supernovae together with a distance prior from the cosmic microwave background (CMB), we find that neutrino masses as large as ${\cal O}(1\,\mathrm{eV})$ are allowed without degrading the fit. Moreover, the combination of BAO data with the CMB distance prior yields a preference for a non-zero DDM fraction, and alleviates the need for dynamical dark energy with phantom crossing. However, the degeneracy introduced by DDM is decisively broken once perturbation observables are included. Incorporating the full $\textit{Planck}$ CMB likelihood, and in particular CMB lensing, restores strong constraints on the neutrino mass in the DDM scenario, $\sum m_ν\lesssim 0.079\,\mathrm{eV}$. In contrast, neutrino mass constraints in a smooth dark energy model described by the Chevallier-Polarski-Linder parametrization become merely $\sim 25\%$ stronger compared to background-only analyses. Our results highlight the essential role of structure-growth measurements in assessing extensions of the dark sector and to obtain robust cosmological neutrino mass bounds.

UV-complete and stable Quintom Dark Energy models in the light of DESI DR2

2026-06-08T15:43:30Z

We propose that Quintom dark energy, the simplest framework allowing crossing of the cosmological-constant boundary, admits a natural UV completion in a 5D anisotropic orbifold lattice: the Non-Perturbative Gauge-Higgs Unification (NPGHU) model. In this setup, a bulk 5D SU(2) gauge field projects on the 4D boundary to a complex scalar and a U(1) gauge field, identified with the dynamical dark-energy sector, while the Standard Model and dark matter remain localized in four dimensions. At late times, bulk-induced dimension-6 higher-derivative operators generate both physical and phantom scalar and gauge degrees of freedom. We show that the resulting 4D effective action is a modified Quintom model whose background equation of state can naturally realize Quintom-B behavior. A crucial contribution arises from the massive gauge ghost, allowing an excellent fit to DESI data with negligible fine-tuning, unlike standard Quintom scenarios. We further show that the inherited properties of the NPGHU construction e.g. absence of fundamental ghost instabilities, absence of potential terms and a finite low-energy cutoff $Λ$ associated with approximate Lorentz invariance, play a central role in the consistency of the effective theory under linear perturbations and vacuum decay. For the most natural regime, $Λ\approx {\cal O}(10)H_0$, the model remains robust despite the presence of IR phantom modes. Our results provide a natural and predictive framework in which Quintom dark energy can be consistently embedded in a fundamental theory.

Pantheon+ supernovae corrected for progenitor age indicate the universe is decelerating

2026-06-08T15:41:19Z

We examine the impact of progenitor age-dependent luminosity evolution of Type Ia supernovae on a cosmographic measurement of the deceleration parameter $q_0$. Our recent redshift tomographic analysis showed that locally $q_0$ has a strong dipole anisotropy aligned approximately with the bulk flow, and only a small monopole component remains at distances exceeding a few hundred Mpc. Applying redshift-dependent corrections for progenitor age to the Pantheon+ catalogue, we find that this shifts the monopole component of $q_0$ to positive values (i.e. deceleration), while leaving the local dipole component essentially unchanged.

Phenomenology of bubble size distributions in a first-order phase transition

2026-06-08T15:39:05Z

In a cosmological first-order phase transition (FOPT), the true and false vacuum bubble radius distributions are not expected to be monochromatic, as is usually assumed. Consequently, Fermi balls (FBs) and primordial black holes (PBHs) produced in a dark FOPT will have extended mass distributions. We show how gravitational wave (GW), microlensing and Hawking evaporation signals for extended bubble radius/mass distributions deviate from the case of monochromatic distributions. The peak of the GW spectrum is shifted to lower frequencies, and the spectrum is broadened at frequencies below the peak frequency. Thus, the radius distribution of true vacuum bubbles introduces another uncertainty in the evaluation of the GW spectrum from a FOPT. The extragalactic gamma-ray signal at AMEGO-X/e-ASTROGAM from PBH evaporation may evince a break in the power-law spectrum between 5 MeV and 10 MeV for an extended PBH mass distribution. Optical microlensing surveys may observe PBH mass distributions with average masses below $10^{-10} M_\odot$, which is not possible for monochromatic mass distributions. This expands the FOPT parameter space that can be explored with microlensing.

A differentiable forward model for weakly perturbed stellar streams: substructure forecasts from density and kinematics spectra

2026-06-08T15:29:00Z

Stellar streams are a promising way to gravitationally detect low-mass substructure, since their low dynamical temperature makes them retain the imprint of weak gravitational perturbations. We develop a fast, differentiable forward model for perturbed stellar streams in the diffusion regime, where the stream is heated by many small velocity kicks rather than by a few strong encounters. The substructure population enters only through its power spectrum, so the computational cost is insensitive to the number of perturbers, and alternative dark matter models and/or baryonic perturbers can be explored by changing this single input. We validate the simulations against analytical predictions, then forecast the sensitivity of a GD-1-like stream to the substructure power spectrum, adding to the stream density the full kinematics, both proper motions and the radial velocity. Kinematic information tightens the constraints by a factor of $\sim 3$-$5$ relative to density alone, improving the precision on the dark matter free-streaming cutoff scale from $\sim 1.2$ dex to $\sim 0.25$ dex at a fiducial value of $M_{\rm hm} = 10^6 M_\odot$ for a $5$ Gyr stream. A single well-measured stream could thus constrain dark matter competitively with current limits from strong lensing and satellite counts.

Big Bang revisited

2026-06-08T15:12:01Z

The Friedmann cosmological solution of the standard Einstein gravitational field equation has a curvature singularity at a moment in time known as the Big Bang. It has been suggested that this Big Bang curvature singularity can be eliminated by use of a degenerate spacetime metric. This proposal was the main topic of our talk at the Workshop, but, here, we also discuss the possible appearance of CPT-conjugated worlds and the conjectured relevance of an extended version of Einstein's field equation.

Spectral distortion anisotropies from photon to dark photon conversions

2026-06-08T13:45:46Z

Dark photons are a gauge boson of a hypothetical dark sector, representing one of the most-studied minimal extensions of the Standard Model, with wide-ranging theoretical and observational implications. Here, we consider scenarios in which an initially unpopulated dark photon sector is populated via resonant photon to dark photon conversions. This process leads to observable spectral distortions in the cosmic microwave background (CMB), that can be used to constrain these models. We extend previous spectral distortion studies of the monopole spectrum to anisotropic spectral distortions, using the newly developed Frequency Hierarchy (FH) framework of CosmoTherm. We illustrate the physics by presenting detailed computations of the photon transfer functions and distortion cross power spectra throughout the dark photon parameter space. We find that the dark photon mass explicitly controls the shape (i.e., multipole-dependence) of the signal power spectra, while the overall amplitude of the signal is determined by the kinetic mixing parameter of the model. Using these results, we place complementary limits on the minimal dark photon model using data from Planck, finding that the constraints are only marginally weaker than those obtained with COBE/FIRAS data for the average (monopole) distortion. In addition, we compute the corrections to the standard temperature field, arguing that conversions at redshifts larger than $2\times 10^6$ may add iso-curvature type perturbations, which could lead to novel constraints in regimes where distortion anisotropies thermalize.

Les Houches on Dark Universe 2025: Elements of cosmology beyond FLRW

2026-06-08T13:41:00Z

Modern cosmology is based on the cosmological principle, which states that the Universe is statistically homogeneous and isotropic. When applied in its strict -- rather than statistical -- sense, the cosmological principle leads to the Friedmann--Lemaître--Robertson--Walker (FLRW) model, which serves as background spacetime. This background is used to predict: (1) the dynamics of cosmic expansion; and (2) the kinematics of light propagation through the Universe, which dictates the interpretation of cosmological observations. In this lecture, we shall discuss the performance of the FLRW model for those purposes, and present some results on the so-called backreaction and fitting problems.

Primordial Black Holes from Slow Phase Transitions with Delayed Reheating: A Peak-Theory Approach

2026-06-08T13:39:21Z

We study the possibility of significant PBH production from a slow first-order phase transition with delayed reheating. Since delayed reheating results in an early matter-dominated phase between percolation and reheating, we developed a peak-theoretic approach to PBH formation during this phase based on the non-Gaussian distribution of overdensity arising from the transition. To obtain the collapse probability, we performed large-scale Monte Carlo simulations and employed the hoop-conjecture criterion. We include tidal-torque terms to investigate the initial spin of the PBHs and find that the average spin parameter is $\mathcal{O}(10^{-3})$. Furthermore, we obtain an emergent overdensity threshold for collapse that depends on the phase transition properties and reheating efficiency. We find that the resulting PBH abundance is extremely sensitive to the reheating efficiency, with order-unity changes in efficiency leading to variations of many orders of magnitude in the collapse fraction. We identify regions of parameter space where the resulting PBHs can account for the entirety of the dark matter abundance. Finally, we also constrain the phase transition and reheating properties from current data on (non-)observations of PBHs.

Bayesian Deep Gaussian Processes for Correlated Functional Data: A Case Study in Cosmological Matter Power Spectra

2026-06-08T13:08:43Z

Understanding the structure of our universe and the distribution of matter is an area of active research. As cosmological surveys grow in complexity, the development of emulators to efficiently and effectively predict matter power spectra is essential. We are particularly motivated by the Mira-Titan Universe simulation suite that, for a specified cosmological parameterization (termed a "cosmology"), provides multiple response curves of various fidelities, including correlated functional realizations. Our objective is two-fold. First, we estimate the underlying matter power spectra, with appropriate uncertainty quantification (UQ), from all of the provided curves. To this end, we propose a novel Bayesian deep Gaussian process (DGP) hierarchical model which synthesizes all the simulation information to estimate the underlying matter power spectra while providing effective UQ. Our model extends previous work on Bayesian DGPs from scalar responses to correlated functional outputs. Second, we leverage our predicted power spectra from various cosmologies in order to accurately predict the entire matter power spectra for an unobserved cosmology. For this task, we use basis function representations of the functional spectra to train a separate Gaussian process emulator. Our method performs well in synthetic exercises and against the benchmark cosmological emulator (Cosmic Emu).