- Article
Recent years have witnessed performance improvements in Multimodal Large Language Models (MLLMs) on downstream natural image understanding tasks. However, when applied to the function graph reasoning task, which is highly information-dense and abunda...