Ψ-0.6: Adaptive Streams of Thought in a Unified Vision-Language World Model（v1）
核对原文摘录 · 5 段，支持 6 项资料
原阅读方向 / 输出什么
The model predicts images, text, camera pose, depth, optical flow, and point tracks as tokens with a shared transformer.一句话认识
the model can translate the specified displacement into an optical-flow prediction, generate the resulting scene, and answer using its own prediction.输入什么
Images, camera pose, depth, optical flow, point tracks, and text are serialized as tokens in a sequence processed by our 8B-parameter autoregressive transformer.动作是否影响预测
optical flow and point tracks already function as proxy actions: they specify how parts of a visual scene should move, but do not directly actuate the world.材料说明的限制
Our current adaptive-stream experiments optimize downstream task performance on visual question answering rather than physical control.