Should Learners that implement '$loglik()' have a "loglik" property?

[mb706]

If so, it is not documented here:

mlr3/R/Learner.R

Lines 62 to 64 in d72be13

	#' * `loglik(...)`: Extracts the log-likelihood (c.f. [stats::logLik()]).
	#' This can be used in measures like [mlr_measures_aic] or [mlr_measures_bic].
	#'

If not, then the AutoTuner should be modified:
https://github.com/mlr-org/mlr3tuning/blob/6eae837fd9c8bbabe4029b243a2677b83047fe6f/R/AutoTuner.R#L244-L246

[berndbischl]

sloglik, AIC and BIC are basically "feature creep" -- which we can clearly see from this issue here.
where we now discuss, to produce EVEN MORE infrastructure what pretty nobody needs who uses mlr3. nearly no supported model has "loglik", no one wants to use insample metrics like AIC when using mlr3 with ML models.

and with custom code this would even still be possible