Speaker
Description
We present Kosmulator IDE, a modular extension of the Kosmulator code [1] designed to constrain interacting dark energy (IDE) models within a fast and flexible Markov Chain Monte Carlo (MCMC) environment. Kosmulator is a vectorised, Python-based inference tool developed for rapid testing of cosmological models at the background level, offering significant speed improvements compared to traditional Einstein–Boltzmann pipelines while maintaining statistical consistency with established codes.
The Kosmulator IDE module implements analytical results for eight phenomenological IDE models derived in recent work, each characterised by interaction terms of the form $Q=3H(\delta_{\rm dm} \rho_{\rm dm} + \delta_{\rm de} \rho_{\rm de})$ and related special cases [2]. These models admit closed-form solutions for the dark sector energy densities and the Hubble function $H(z)$, enabling efficient likelihood evaluation without requiring numerical integration of the background equations.
A key feature of this module is the incorporation of physically motivated priors defined directly as analytical constraints on parameter space, based on conditions required to avoid non-physical behaviour such as negative energy densities, imaginary solutions, future singularities (e.g. big rip scenarios), and dynamical instabilities (including so-called “doom factor” pathologies [3]). The implementation allows users to select between different “regimes,” as defined in [2], thereby explicitly controlling which theoretical behaviours are permitted during parameter inference.
This framework enables a systematic investigation of how posterior constraints, best-fit statistics (e.g. $\chi^2$, $\Delta$AIC), and model viability depend on the choice of priors and tolerated physical behaviour. In addition, Kosmulator IDE facilitates reproducibility by allowing researchers to directly recover and compare results from existing IDE studies within a unified computational environment.
More broadly, this work serves as a proof of concept demonstrating that Kosmulator can be readily extended beyond the standard $\Lambda$CDM paradigm to accommodate alternative cosmological models with minimal implementation effort. The framework’s modular design and support for equation-based priors position it as a powerful tool for rapidly screening theoretical models before deploying more computationally intensive pipelines. While current implementations are limited to background-level observables, ongoing developments aim to incorporate perturbations and full CMB likelihoods.
References
[1] Kosmulator: A Python framework for cosmological inference with MCMC
R. T. Hough, R. Rugg, S. Sahlu, and A. Abebe, arXiv:2602.08424 (2026)
[2] III. Interacting Dark Energy: Summary of models, Pathologies, and Constraints, M. van der Westhuizen, A. Abebe, and E. Di Valentino, Phys. Dark Univ. 50, 102121 (2025).
[3] Dark coupling, M. B. Gavela et al., JCAP 07, 034 (2009).
| Apply for student award at which level: | PhD |
|---|---|
| Consent on use of personal information: Abstract Submission | Yes, I ACCEPT |