Awesome World Models / arXiv

Memory Helps, but Confabulation Misleads: Understanding Streaming Events in Videos with MLLMs

作者：作者信息待提取 · 单位：Munich Center for Machine Learning, Germany · 来源：General Approaches to World Models · 发布日期：2025 · 分类：5. Building World Models by Bridging Language and Vision Intelligence

语言先验其他模态记忆视频生成
新标签打开 AlphaXiv新标签打开 Project
Memory Helps, but Confabulation Misleads: Understanding Streaming Events in Videos with MLLMs figure
AlphaXiv 中文概览（可滚动查看）