引导式测验 · 中等 等级
OpenAI Sora 测验
查看 Sora 的视频生成架构、记录的限制和出处,以及 Sora 产品的当前可用性状态。
问题 1 的 8
OpenAI 的 Sora 主要生成什么?
预览每个问题8 问题
当您参加测验时,就会出现答案和解释。
- OpenAI 的 Sora 主要生成什么?
- 文字描述中的视频剪辑
- 歌词中的音乐曲目
- 可 3D 打印的模型
- 数据的电子表格
- 哪种底层架构最能描述 Sora?
- 一个简单的基于规则的动画引擎
- 在时空斑块上运行的扩散变压器
- 仅用于文本的循环神经网络
- 卷积图像分类器
- 类似于令牌的单元 Sora 进程被称为什么?
- 像素
- 每秒帧数
- 时空补丁
- 词嵌入
- OpenAI在描述Sora为“世界模拟器”的一步时是什么意思?
- It proves the model has a complete causal model of physics
- It guarantees every generated action is physically correct
- It means the model can control real-world robots safely
- Video training can teach the model some patterns of how scenes and objects behave
- Sora 的已知限制是什么?
- 它可能会与精确的物理和对象一致性作斗争
- 只能输出黑白视频
- 它不能产生任何运动
- 它仅适用于一种固定的宽高比
- What provenance feature did OpenAI say accompanied Sora-generated videos in its product?
- A blockchain transaction for every frame
- C2PA metadata, with visible watermarks in the product experience
- A spoken disclaimer embedded in the audio
- A cryptographic signature proving every scene is factual
- When did OpenAI’s deprecations page say the Videos API and Sora 2 API models shut down?
- April 26, 2026
- February 13, 2026
- September 24, 2026
- October 23, 2026
- Which added capability did OpenAI describe for Sora 2 compared with earlier video-generation models?
- Automatic legal fact-checking of scenes
- Guaranteed exact object counts
- Real-time control of physical vehicles
- Synchronized generated audio with video