Comment by nlarew

The frontier labs are not "fine-tuning", they're doing massive scale RL post-training