1 paper
Di Feng, Kaixin Ma, Feng Nan +9
Multimodal large language models (MLLMs) are increasingly deployed in real-world, agentic settings where outputs must not only be correct, but also conform to predefined data schem…