Skill-RM: Unifying Heterogeneous Evaluation Criteria via Agent Skill

Skill-RM: Unifying Heterogeneous Evaluation Criteria via Agent Skill

Reward models (RMs) provide critical feedback signals for LLM post-training, notably in reinforced fine-tuning (RFT) and reinforcement learning (RL) pipelines. However, current reward evaluation relies on heterogeneous criteria — Skill-RM proposes unifying them via agent skill.

https:/...

https://arxiv.org/abs/2606.03980v1

#Skill #Unifying #Heterogeneo...