DreamPRM-1.5: Unlocking the Potential of Each Instance for Multimodal Process Reward Model Training

arXiv:2509.05542v1 Announce Type: new Abstract: Training multimodal process reward models (PRMs) is challenged by distribution shifts and noisy data. We introduce DreamPRM-1.5, an instance-reweighted framework that adaptively adjusts the importance of each training example via bi-level optimization. We design two complementary strategies: Instance Table, effective for smaller datasets, and Instance Net, scalable to larger ones. Integrated into test-time scaling, DreamPRM-1.5 achieves 84.6 accuracy on the MMMU benchmark, surpassing GPT-5.

2025-09-09 04:30 GMT · 8 months ago arxiv.org

arXiv:2509.05542v1 Announce Type: new Abstract: Training multimodal process reward models (PRMs) is challenged by distribution shifts and noisy data. We introduce DreamPRM-1.5, an instance-reweighted framework that adaptively adjusts the importance of each training example via bi-level optimization. We design two complementary strategies: Instance Table, effective for smaller datasets, and Instance Net, scalable to larger ones. Integrated into test-time scaling, DreamPRM-1.5 achieves 84.6 accuracy on the MMMU benchmark, surpassing GPT-5.

Original: https://arxiv.org/abs/2509.05542