Concept
Bayesian Regret Bound
Within the framework of Bayesian decision theory, an upper bound measuring the performance gap between approximate inference (such as amortized median probability models) and ideal Bayesian posterior inference. It indicates the conditions under which amortized selection can approach the ideal Bayesian decision.
Timeline (last 90 days)
Oct 16
The SelectPFN research team developed Bayesian regret bounds for amortized selection, quantifying the gap between amortized inference and ideal Bayesian inference for the amortized median probability model.
Unverified50%