We generate on-policy rollouts using the corresponding backbone checkpoint.
specific method · filed under post-training
On-policy rollout generation
Generates rollouts using the policy or corresponding backbone checkpoint.
The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See rollout & RL infrastructure for the methods filed beside it.
Also called on-policy rollouts.
Documented in
Evidence
2 quotes from 2 sources, one card per model, each opening with its strongest.
How sources treat it
used 2
usedunclearin Nemotron 3.5 Super and Nemotron 3.5 LightningNVIDIA
This technique uses strong teacher models to guide training on the model's own generated attempts (on-policy rollouts)
usedoptimizationin Nemotron 3 UltraNVIDIA