GRAM: A Generative Foundation Reward Model for Reward Generalization
Paper
•
2506.14175
•
Published
•
1
GRAM
Generative Foundation Reward Models for Reward Generalization
-
-
NiuTrans/GRAM-Qwen3-14B-RewardModel
15B • Updated • 9 • 3 -
NiuTrans/GRAM-Qwen3-8B-RewardModel
8B • Updated • 9 • 4 -
NiuTrans/GRAM-Qwen3-4B-RewardModel
4B • Updated • 14 • 2 -
NiuTrans/GRAM-LLaMA3.2-3B-RewardModel
3B • Updated • 10 • 3 -
NiuTrans/GRAM-Qwen3-1.7B-RewardModel
2B • Updated • 8 • 6 -
NiuTrans/GRAM-fine-tuning-65k
Viewer • Updated • 65.8k • 78 • 2 -
NiuTrans/GRAM-pre-training-566k
Viewer • Updated • 566k • 76 • 3