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

#precision

0 stories this week · 1 this month · 1 in 6 months · mostly AI

Stories per day · 30 days

Relatedalignment1 reinforcement-learning1 reward-hacking1 rubrics1

1 stories, newest first · one card per story

Hugging Face daily papers19d ago

FORK reward adds precision to curb rubric-based RL reward hacking

Against the base policy, win rate rises from 10.6% to 68.0% on RubricHub-Medical and to 74.1% on HealthBench.

reward-hacking reinforcement-learning

1