Open Models Almanac
Methodspost-trainingrollout & RL infrastructure

single mention · post-training

Multi-harness rollouts

Runs the same prompts through several agent harnesses to encourage behavior that transfers across scaffolds.

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 multi-harness rollout in RL.

Filed under post-training › rollout & RL infrastructure