Open Models Almanac
Methodspost-trainingreinforcement learning algorithm

single mention · post-training

Importance weighting for asynchronous policy lag

Importance-weights training to correct policy lag and numerical mismatch between inference and trainer policies.

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

Also called importance weighting.

Filed under post-training › reinforcement learning algorithm