The longitudinal spin transfer, $G_{1L}$, describes the hadronization of polarized partons. It thus was usually measured in polarized reactions or high-energy collisions where weak interaction dominates. In a series of works, we propose the dihadron polarization correlation as a novel probe of this quantity. Such an observable does not require the fragmenting partons to be polarized and therefore can be measured in the currently available experimental facilities, such as $e^+e^-$, $pp$ and $ep$ colliders. We make quantitative predictions for these experiments. In light of the data already harvested, the experimental investigation of this observable provides more opportunity for the quantitative study of the longitudinal spin transfer. In particular, the measurements in unpolarized $pp$ collisions can significantly constrain the fragmentation function of a circularly polarized gluon. This proceeding briefly summarizes the key concept.