Bayesian Validation of a Structured Λ(z) Cosmological Model Inspired by Selector-Time Theory Using Pantheon+ Data

This work presents a Bayesian validation of a cosmological model with a variable effective cosmological term Λ(z), inspired by Selector-Time Theory (STT), using the Pantheon+ Type Ia supernova dataset. The central objective is to evaluate whether a minimal and controlled modulation of the cosmological sector can produce a statistically robust improvement relative to the standard flat ΛCDM model when analyzed under full Bayesian inference. The proposed model introduces a bounded functional modulation based on the variable x(z)=z/(1+z), ensuring smooth, monotonic, and stable behavior across the redshift domain. The analysis combines deterministic optimization with full Markov Chain Monte Carlo (MCMC) sampling using the Pantheon+ covariance matrix. Results show that the refined model remains favored relative to ΛCDM both at the deterministic level (Δχ² ≈ -6.42) and at the level of posterior medians (Δχ² ≈ -5.86), indicating that the improvement is not a pointwise artifact but persists under full probabilistic treatment. The posterior structure reveals non-trivial parameter correlations, suggesting internal compensatory mechanisms within the cosmological sector. These findings support the hypothesis that a structured and controlled modulation of Λ(z) can produce measurable and consistent improvements in observational fits. The model is interpreted as a minimal phenomenological extension rather than a complete cosmological derivation from STT, providing a testable framework for further investigation.