1
Faithful Grounded Visual Reasoning via Learned Proxy-Tokens
提出通过可学习的代理令牌实现忠实的基础视觉推理，破解多模态大模型在视觉问答中的黑盒困境，增强可解释性。
arXiv:2606.23354v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have achieved remarkable success in Visual Question Answering…