Login / Signup
HAF-RM: A Hybrid Alignment Framework for Reward Model Training.
Shujun Liu
Xiaoyu Shen
Yuhang Lai
Siyuan Wang
Shengbin Yue
Zengfeng Huang
Xuanjing Huang
Zhongyu Wei
Published in:
CoRR (2024)
Keyphrases
</>
probabilistic model
theoretical framework
formal model
computational model
high level
unified model
training algorithm
conceptual framework
theoretical foundation
bayesian framework
experimental data
theoretical analysis
management system
prior knowledge
hybrid model
generic model
data sets