MaLiang-Harness：可编程图像与视频生成框架
原标题：Paper page - MaLiang-Harness: A Programmable Path to Image and Video Generation
AI 摘要
该论文提出 MaLiang-Harness 框架，将 MLLM 驱动的图像与视频生成组织为持续的构建、检查与修订过程，并定义 Program-to-Visual (P2V) gap 来描述程序正确执行但视觉结果不符合要求的问题。框架包含 PEG、TGP、REV 三项机制，在 MaLiang-IBench 与 MaLiang-VBench 上评测了 11 个闭源 MLLM，其中 GPT-6-Astra 在两个基准上均达到 100% 生成成功率。研究还发现通用能力评分与视觉生成表现之间存在不匹配，项目代码已在 GitHub 开源。
正文节选
MaLiang-Harness: A Programmable Path to Image and Video Generation Abstract Executable programs offer explicit control over how images and videos are constructed, but generating runnable code is only the beginning of visual creation. A program can execute correctly while violating the requested composition, appearance, or motion. We define this discrepancy as the Program-to-Visual (P2V) gap and introduce MaLiang-Harness, a unified framework for organizing MLLM-driven visual generation into a per