Open Models Almanac
Methodsoptimizationtraining stability

single mention · optimization

Safeguards against training drift and reward hacking

Safeguards intended to address training drift and reward hacking; the evidence does not specify their mechanisms.

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

Filed under optimization › training stability