Aligning Diffusion Language Models via Unpaired Preference Optimization
"人类偏好对齐" 相关文章
Multiplayer Nash Preference Optimization
加速近5倍！北大与字节团队提出BranchGRPO，用「树形分叉 + 剪枝」重塑扩散模型对齐
加速近5倍！北大与字节团队提出BranchGRPO，用「树形分叉 + 剪枝」重塑扩散模型对齐
Difficulty-Based Preference Data Selection by DPO Implicit Reward Gap
Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models
