Perplexity and NVIDIA researchers introduce JOLT learning method
The available page is a cookie notice, so the method's results and release status are not shown.
1
Stories per day · 30 days
1 RSS
Relatedreinforcement-learning2kl-regularization1llm1nvidia1on-policy-distillation1perplexity1self-distillation1
2 stories, newest first · one card per story
Stories tagged #jolt
A Good Self-Teacher Meets the Student Where They Are: Joint On-Policy Learning and Teaching
JOLT jointly trains a privileged self-teacher and student so on-policy distillation matches the student's ability.