Open Models Almanac
Methodspost-trainingreward modelling

single mention · post-training

Deterministic reward-checking chain

Applies ordered deterministic checks to each terminated rollout, with the first failing check determining reward.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See reward modelling for the methods filed beside it.

Also called deterministic chain of checkers.

Filed under post-training › reward modelling