NewA calmer, faster ZeroHour — security assessments on articles, story timelines and search from any page.

#kl-regularization

0 stories this week · 2 this month · 2 in 6 months · mostly AI

Stories per day · 30 days

Relatedllm2 on-policy-distillation2 reinforcement-learning2 jolt1 knowledge-distillation1 policy-optimization1 self-distillation1

2 stories, newest first · one card per story

Hugging Face daily papers17d ago

Rethinking Relative Policy Differences Transfer of On-Policy Distillation

Relative policy differences, not expert strength alone, determine useful on-policy distillation transfer.

on-policy-distillation policy-optimization

1

arXiv cs.AI / cs.LG / cs.CL18d ago

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.

jolt on-policy-distillation

4