LLaVA-CoT

Researchers propose Data-DPO, a target model-oriented data selection method for supervised fine-tuning that uses one-step probing to capture target-model-aware data preferences and combines them with …

Researchers propose MASS, a hierarchical data selection method that uses manifold coverage and sparse feature coverage to improve LLM post-training efficiency. In experiments on Vision Flan and LLaVA-…