Fetching the paper…

Task Preference Optimization: Improving Multimodal Large Language Models with Vision Task Alignment · Around