VideoJev: System One Model for Video Understanding and Interaction
Visual agents must turn video observations into timely decisions to track progress and choose what to do next. Building on the emerging System One paradigm, we introduce VideoJev, a family of models that cast video understanding and interaction as structured decision tasks. We adapt pretrained multimodal large language models (MLLMs) by replacing autoregressive answer decoding
Reading in Plain English mode — toggle below to switch
Visual agents must turn video observations into timely decisions to track progress and choose what to do next. Building on the emerging System One paradigm, we introduce VideoJev, a family of models that cast video understanding and interaction as structured decision tasks. We adapt pretrained multimodal large language models (MLLMs) by replacing autoregressive answer decoding
VideoJev introduces a novel approach to video understanding and interaction by integrating System One thinking into MLLMs, enabling more efficient and timely decision-making for visual agents. This shift from autoregressive decoding to structured decision tasks offers a pathway to building more responsive and intelligent AI systems that can directly act on video observations.