Joint Bayesian 4PL Models for Thermal Load Sensitivity // Fits joint Bayesian four-parameter logistic (4PL) models to thermal- tolerance proportion data, extracts the classical thermal load sensitivity quantities (z, CTmax at 1 hour, T_crit) with full posterior uncertainty, and predicts heat-injury accumulation and survival under fluctuating temperature regimes with optional Sharpe-Schoolfield repair. Models are fitted with 'Stan' via the 'brms' package. Implements the framework described in Noble, Arnold, Nakagawa and Pottier (in preparation).